NCAA Tournament Blog - NCAA Tournament Basketball Stats

2009-2010 NCAA Tournament Blog

Read the daily ramblings of Pete Tiernan as he discusses the strategies behind BracketScience.

Seasons: 2009-2010 2008-2009 2007-2008 2006-2007


Thu, Apr 8, 2010, 2:24 PM	By ptiernan	0 Comments...
The 2010 tourney is history--and it will no doubt live in all of our memories as one of the greatest dances of all time. Consider this: if this year is indeed the final tourney for the 64-team format, then you couldn't ask for better bookends to the era of the "perfect tournament." The 64-team bracket debuted in 1985 with the all-time shocker of eight-seed Villanova beating Georgetown, and ended with a Hoosieresque storyline and one of the most stirring sporting events ever. Try to top that, Mister 96-team Tourney Smartypants. Anyway, with Duke's win, all my head-scratching over why the stats models kept coming up Duke make perfect sense now. It's because they were supposed to win. In fact, I've had more people email this year than any other year thanking me for helping them win or get in the money for their pool. And--you know me--I'm very reluctant to make bold proclamations how about well these stats-based models will do. I like to approach the numbers with a healthy dose of skepticism. That said, the top models below likely had you smiling by the end of this dance: 1. Michelle's mix (Duke) - 1320, 99.9 percentile 2. Factors (Duke) - 1180, 98.8 percentile 3. Combopase (Duke) - 1120, 97.3 percentile 4. Final Four/champ (Duke) - 1110, 97.0 percentile 5. Pythag (Duke) - 1100, 96.6 percentile 6. Pythag/coach (Duke) - 960, 93.7 percentile 7. Pulse check (Kansas) - 830, 89.9 percentile 8. High seed chalk (Kansas) - 780, 86.7 percentile 9. Upset/tossup (Syracuse) - 670, 74.1 percentile 10. Seed matchup (Kansas) - 660, 72.2 percentile 11. Upset special (Georgetown) - 620, 61.8 percentile 12. Contend/pretend (Kansas) - 610, 57.5 percentile 13. Pete's gut (Kansas) - 580, 47.6 percentile 14. Keeper (Kentucky) - 560, 40.8 percentile I've compared the original 12 posted models against two unpublicized models--my wife Michelle's (at the tippy top of the list) and "high seed chalk" in the eighth position. This is what you would've scored had you went with the higher seed in all match-ups and the one seed with the higher margin from the Final Four on. For those curious about how Michelle managed her triple-pool-winning feat, read the guest blog below. Overall, this will go down in my tourney prognosticating history as the year when stats thoroughly thumped my instincts. The nine models based purely on numbers averaged in the 86.3 percentile; the three models for which I didn't use any strict statistical guidance averaged in the 50.1 percentile. Yowch. In fact, six of the nine stats-based models beat chalk, while just three fell short of chalk. Meanwhile my instinct-based picks ranked 11, 13 and 14--and all of them were well below chalk. Next year, I vow to listen to the siren song of stats...at least a little more than this year. Be on the lookout for my 2010 wrap-up blog, in which I'll count down the top ten stories from this year's tourney.
0 Comments...


Wed, Apr 7, 2010, 2:06 PM	By ptiernan	0 Comments...
I told my wife Michelle that if she won the bracketscience.com Bracket Madness pool, she had to write a blog explaining her process. (Don't worry: Michelle isn't eligible for any of the gifts). She not only won the pool on this site; she won our friends pool, her office pool and wound up in the 99.9th percentile of the ESPN Tourney Challenge. I bow to her superior soothsaying skills! Ten minutes ago, Michelle sent me her blog. Here's how she did it: __________ This is the first year that it's actually paid off to live with a bracketscientist!!! In addition to coming in first on the website, I won our "friends & family pool" as well as my office pool. My strategy for filling out my bracket included sharing a bottle of wine with Pete and grilling him on the team match-ups. I want you to know it took me 3 hours to fill out that bracket, which I would not have admitted had I lost! Here's how I filled it in: First, Pete was walking around the house saying that most of his models were coming up with Duke as the champion, so that was good enough for me, and I entered Duke as my champion. Second, I like the Upset/TossUp Model for the first round only, so I mostly went with those picks. On the few exceptions I made, it all evened out, so I scored exactly the same as that model. Loved Washington going two rounds! Don't you love to be the one in the office pool who picks a few of the upsets? Third, for subsequent rounds, I compared the team match-ups head-to-head using the excel spreadsheet that Pete posts to the website. I would hide all the other teams, so I could view the two teams together. I love that excel sheet, and I'm all about efficiency ratings, so I mainly looked at SOS (strength of schedule), OER (offensive efficiency rating), DER (defensive efficiency rating), and Pythag, with some talking with Pete about coaches. Fourth, I have an irrational dislike of Syracuse (they let me down one year), so I wanted them out of my bracket as soon as possible. I said to Pete, "Can Florida State beat them in the first round?" He said no. I said, "Can Butler beat them in the second round?" He said yes, so there you go! My new team aversion for next year is Georgetown - I had them going three rounds this year and they let me down. Looking forward to 2011....
0 Comments...


Mon, Apr 5, 2010, 11:57 PM	By ptiernan	0 Comments...
I didn't want to believe it...but all those stats models that kept coming up Duke were right. What a great game. My head wanted my stats models to be proven right; my heart was pulling for that last Hayward shot to go in. Get this: for the eighteenth time in 20 years, the six champs rules held yet again. Incredible!
0 Comments...

Most contests already settled...and I'm humbled by the winners
Sun, Apr 4, 2010, 11:10 AM	By ptiernan	2 Comments...
The results of the Final Four games have nearly settled all of the bracketscience.com contests. Here's where things stand this Sunday morning after Butler and Duke advanced: GREAT EIGHT CHALLENGE 1. bdance of Saline, MI (8GB Apple iPod nano) 2. llchile of Charlotte, NC (free lifetime membership) 3. buckeyenut61 of Arvada, CO (free lifetime membership) I know the leader of this pool. Barry and I frequent the same pub in quiet Saline, Michigan, just south of Ann Arbor. All things considered, Barry would much rather have lost this pool and seen his beloved Kentucky Wildcats cut down the nets. But he'll settle for bragging rights over me until the 2011 dance. BRACKET MADNESS 1. llchile of Charlotte, NC (4GB iPod shuffle) 2. mstollak of De Pere, WI (free lifetime membership) 3. IF BUTLER WINS: 58down of Drummonds, TN (free lifetime membership) 3. IF DUKE WINS: RandyT of Oklahoma City, OK (free lifetime membership) There's a little drama left in this contest. If Butler wins, 58down gets a lifetime membership. If Duke wins, it's RandyT. Actually, if you look at the standings, you'll find that the person who will win this contest if the Blue Devils prevail is one "mtiernan." That's my wife, Michelle! Talk about humility. She mopped up in pools this year, winning both her office and our local pools--and possibly this pool too. She's promised to write a blog on her strategy (I bow to the new bracketmaster). As mentioned before, her bracket is a unique blend of pulse check stats, ancient school grudges, Pete-badgering and wine. Who am I to argue. Of course, she can't win any of the prizes. But a one-year license to remind me of her bracket superiority is gift enough (for her) and punishment enough (for me). SECOND CHANCE DANCE 1. ccfdawg of Athens, GA (2GB Apple iPod shuffle) 2. JRRHOUSTON of Houston, TX (free lifetime membership) 3. Brian Dachel of Oak Creek, WI (free one-year membership) Some might be surprised to see three prizes here. The official rules call for just two. But when I put together the promo graphic for the home page, I had a brain fart and said "prizes for second and third." That little mistake works in Brian Dachel's favor. He gets a one-year membership. The real story of this contest, though, is another personal connection. The true leader of the Second Chance Dance is Jeff Smeenge. Of course, he's not allowed to win--because Jeff helps me with all the databasing and web app work on the site. He's the guy who figured out how to serve up the Bracketmaster research data. That's right: my wife...my business partner...and my pub buddy all have bragging rights over me this tourney. It's going to be a long, long year before I get redemption. And to add insult to injury, we're probably going to 96 teams--so my workload will get even more intense. Remind me again: why do I do this?
2 Comments...


Wed, Mar 31, 2010, 8:00 PM	By ptiernan	3 Comments...
It shouldn't come as a surprise that this year's Final Four teams have the fourth lowest average seed among the 26 semi-finalists of the 64-team era. And you can tell just by watching MSU, Butler and West Virginia (Duke gets a pass on this one) that this is a group of grinders. Even with the Blue Devils' high scoring average, this is the second lowest scoring group of Final Four contenders, with the second lowest scoring margin. I'm also not shocked that this is the second most inefficient Final Four group since I started keeping track of Pythag in 2004. Only the 2006 squads had a lower average Pythag. That makes sense, since there were no one seeds and George Mason's low Pythag. But with a low average Pythag of .9455, this year's Final Four are somewhere between Baylor and Purdue. Here's another interesting fact: not a single one of these teams--not even Duke (.9818)--have a higher pre-tourney Pythag than the average of the four 2008 one seeds that vied for the championship. One thing I am surprised about with this bunch is that it's among the most experienced Final Four groups in the last 26 years. The coaches have the third most average tourney trips (14.75 trips--even factoring in Stevens' three) and have been to the Elite Eight the fifth most times (five on average--and Stevens has none), In addition, the teams have the sixth most consecutive bids (8.75) of any Final Four field. And the last surprise for me is how reliant this year's Final Four is on its frontcourt. The four teams get just 47.5 percent of their points from guards. That percentage is the lowest since 2001, bucking a trend toward increasingly guard-dominant squads.
3 Comments...


Tue, Mar 30, 2010, 12:43 PM	By ptiernan	0 Comments...
After the Elite Eight, here's where the 12 models stand in the ESPN Tourney Challenge: 1. Pulse check (Kansas) - 830, 99.7 percentile 2. Factors (Duke) - 700, 91.4 percentile 3. Upset/tossup (Syracuse) - 670, 85.2 percentile 4. Seed matchup (Kansas) - 660, 82.6 percentile 5. Combopase (Duke) - 640, 76.5 percentile 6. Final Four/champ (Duke) - 630, 72.9 percentile 7. Pythag (Duke) - 620, 69.2 percentile 7. Upset special (Georgetown) - 620, 69.2 percentile 9. Contend/pretend (Kansas) - 610, 63.7 percentile 10. Pete's gut (Kansas) - 580, 51.7 percentile 11. Keeper (Kentucky) - 560, 43.7 percentile 12. Pythag/coach (Duke) - 480, 16.3 percentile Pulse Check looks impressive--and would probably have you at the top of your pool right now--but there's the little matter of its picking Kansas in the finals. Actually, Factors is the healthiest model on the list, since it has Duke to win it all. Remember that hand-wringing I was doing when so many stats-based models kept coming up Blue Devils? Maybe I should've listened to the numbers. In fact, my wife did--and she's in the 99.8 percentile on the ESPN Tourney Challenge, with 840 points and Duke to win. She's winning her work pool, our friend pool--and is in second in the bracketscience.com pool. She promised that if she wins, she'll write a guest blog that explains her complicated system using pulse check stats, incessant questioning of her husband, wine and past prejudices ("Pete...tell me who can beat Syracuse? Gonzaga!? Butler?! Anybody! I'm not picking them in another pool again!") Michelle is always amazed that I don't listen to my own advice. I've wrangled with the numbers enough to know that they can only take you so far. Sometimes going off on my own works to my advantage; other times it hurts me. This year, it definitely hurt. The models above that are based on nothing more than my preferences average in the 54.9 percentile. The models where I'm ruled by some sort of statistical guidance average in the 73.1 percentile--a 33.2% improvement over my hunches. Yeesh!
0 Comments...


Sun, Mar 28, 2010, 9:15 PM	By ptiernan	2 Comments...
Michigan State's win over Tennessee and Duke's defeat of Baylor removed any chance that the 2010 dance would be the craziest of all time. The Madometer settled below 18.0% after the Blue Devils' win. This year is still tied for third with 1999 as the third wackiest tourney (1986 and 1990 were the most unpredictable), and it can't get any crazier than 17.9% Chitwoodin' mad. Speaking of Chitwood, expect to hear a lot of allusions to Hoosiers and Jimmy Chitwood throughout the week. Butler is an Indiana school--and actually plays in Hinkle Fieldhouse, the infamous site of the Hickory Huskers' victory over South Bend Central. (And, yes, its baskets are ten feet high--just like they are in your local high school gym. Gene Hackman had Verle measure.) Throughout the week, I'll write blogs about this Final Four group--and lessons from the 2010 tourney. Lesson number one: when the numbers say Duke will win (five of nine stats models had the Blue Devils), don't go against your models. Lesson number two: there is something to stat that teams which didn't go to the previous dance won't be cutting down the nets. All season long, I pointed that out in relation to Kentucky. Then I promptly picked them in my Keeper bracket. Admittedly, I was trying to be a contrarian against my stats-based results, but I got what I deserved.
2 Comments...


Sun, Mar 28, 2010, 1:08 PM	By ptiernan	0 Comments...
With five Power conference teams still in the running for the 2010 tourney championship, it's too early to say exactly where the Big Six will land in terms of their PASE values, but one thing is clear: the Big East is going down, no matter what happens from here on in. Before the 2010 dance, the Big East held a narrow +.142 to +.136 PASE lead over the ACC. But even if West Virginia wins the tourney and Duke loses today, the best the Big East can do is a +.096 PASE while the worst the ACC will have is a +.116 PASE. That's because the Big East has massively tanked this year. Seven of their eight teams underachieved against seed expectationsóbig time. Not counting the Mountaineers, the other Big East teams were 9.2 games below seed-projected wins, or an average of 1.3 games per team. That's hugely disappointing. I don't remember a conference ever falling so short of expectations. In fact, there's a good chance that the Big East could fall even below the SEC in the conference PASE race. Before the 2010 dance, the Big East held a solid +.142 to +.095 PASE advantage over the SEC. If the Mountaineers fail to win the tourney and Tennessee bags one more win, the SEC will overtake the Big East in the PASE race. Amazing. Meanwhile, there's little drama among the three bottom-dwelling Big Six power conferences. The Pac-10, which was so maligned going into the dance, will add to its PASE. They figure to add +.020 to their underachieving -.167 PASE. The Big 12 will also stay anchored in the fifth position. No matter what Baylor does, the conference will either lose -.012 off their -.073 PASE or add +.011ónot enough to catch the Big Ten. Speaking of the Big Ten, there's a little more drama in their situation. If Izzo can somehow coax a championship out of his wounded squad, he'll push the Big Ten into overachievement territoryóby a miniscule hundredth-of-a-game margin. And if the Spartans lose today, the Big Ten will still add +.008 to their pre-tourney -.029 PASE. That's what happens when you have the all-time overachieving coach in your conference. We already talked about how impressive Izzo's record is. After beating Northern Iowa, it got even more amazing. Right now, he's the best performing coach, with an astounding +.864 PASE. Even if he loses today, his PASE will rise to +.941. But if he wins, his PASE will eclipse one-game-per-tourney seed overachievement, climbing to +1.018. And if Michigan State cuts down the nets in 2010? How does a +1.172 PASE grab you? For a coach with 13 tourney trips, to have a PASE that high is absolutely astounding.
0 Comments...


Wed, Mar 24, 2010, 7:23 PM	By ptiernan	3 Comments...
With all the big upsets in the 2010 tourney, several coaches will see their overall PASE ratings make correspondingly big movesóboth positively and negatively. The biggest PASE decliners will be John Thompson III, Kevin Stallings, Jay Wright, Bill Self and Mike Brey. Before the Hoyas tanked against Ohio, Thompson III had a PASE of +.216, good enough for 27th place out of 75 coaches with at least four tourney trips. He'll see his PASE drop to -.117 underachievement (the exact number will depend on what the average wins per seed end up being after the dance). Based on this year's rankings, Thompson III would fall to 54th place among veteran tourney coaches. That's a drop of 27 positions. Ouch! Kevin Stallings and Jay Wright can expect similarly dramatic declines. Vanderbilt's upset loss to Murray State will send Stallings' PASE spiraling from slight overachievement of +.004 to solid underachievement of -.250. And his ranking will suffer tooódropping from 43rd to 64th. Nova' loss to St. Mary's will send Wright's PASE from +.436 to around +.202. He'll fall from his lofty 13th place position all the way to 29th out of 75 veteran tourney teams. Then there's Bill Self and Mike Brey. Northern Iowa's stunning upset of Kansas takes .223 off Self's PASE. Where he once had the 24th best PASE at +.237, he'll be down near 43rd at +.014. Consider this: old Bill has now been victimized in 3v14, 4v13 and second-round 1v9 matchups. That's some serious underachievement. Brey was just the 50th best performing veteran coach, with an underachieving -.093 PASE. And then Notre Dame bit the dust against Old Dominion in round one. Now his PASE is -.234 and he'll rank around 63rd out of 75 coaches. Those are the coaches who are likely to suffer the worst PASE declines. Who figures to gain the most from the 2010 dance? Well, for nine veteran coaches, the jury is still out. Coach K, Boeheim and Calipari still need to win two games to add to their PASE. And Matta and Huggins need to win one. Matt Painter will already build on his +.123 PASE, even if the Boilermakers lose their next game to Duke. Two wins is +.520 games above expectations for four seeds. Bruce Pearl will get an even bigger bump. Before the dance, Pearl was a -.102 PASE underachiever. If he didn't win another game, he'd still wind up as a slight +.024 PASE overachiever. The same goes for Lorenzo Romar. Before this year's tourney, Romar ranked 70th out of 75 coaches, with a dreadful PASE of -.458. Even if the Huskies bow out to West Virginia, his PASE stands to rise to -.128. That's still underachievement territory. But it would up his ranking 16 positions to 54th. And that brings us to the amazing case of Tom Izzo. He's already the best PASE overachiever out of 75 coaches with at least four tourney trips. And that impressive +.864 PASE will increaseóno matter what happens to the Spartans from here on in. That's because winning two games as a five seed qualifies this year as +.870 PASE overperformance. Think about this: Izzo has beaten seed expectations 10 times in his 13 tourney trips. That works out to a 76.9% seed overachievement rate, better than every other veteran tourney coachóexcept in-state rival John Beilein, who has overachieved in four of his five tourney trips. Izzo will also widen his PASE lead over his nearest competitors. Second-place coach Steve Fisher will see his PASE drop after a first-round loss with his Aztecs, as will fourth-place Billy Donovan, sixth-place Rick Pitino, seventh-place Mike Anderson (even though he won one game) and ninth-place Tom Penders. Third-place Beilein, fifth-place Larranaga, eighth-place Mike Davis and tenth-place Sean Miller won't see their PASE values change (except for minor variances due to new wins per seed values), because they missed the dance.
3 Comments...


Mon, Mar 22, 2010, 7:11 PM	By ptiernan	2 Comments...
After the first round, here are the results of the 12 bracket models we ran in the ESPN Tourney Challenge (champ pick in parentheses): 1. Pulse check (Kansas) - 470, 99.4 percentile 2. Upset/tossup (Syracuse) - 430, 92.9 percentile 2. Final Four/champ (Duke) - 430, 92.9 percentile 4. Pythag (Duke) - 420, 88.2 percentile 5. Contend/pretend (Kansas) - 410, 82.3 percentile 6. Keeper (Kentucky) - 400, 74.6 percentile 6. Combopase (Duke) - 400, 74.6 percentile 8. Upset special - 380, 55.4 percentile 8. Pete's gut - 380, 55.4 percentile 8. Seed matchup - 380, 55.4 percentile 8. Factors - 380, 55.4 percentile 12. Pythag/coach - 320, 8.5 percentile If you were in a pool of less than 100 people, it's likely that Pulse Check would put you in the top spot. In our Bracket Madness contest, it would be good for a third place tie. Alas, that bracket is doomed because it lands on Kansas to win. The healthiest of these brackets are the two second-place models, which have Syracuse and Duke cutting down the nets. All things considered, sticking with the stats-based models was a better way to go this year--at least so far. The models driven purely by numbers and none of my hunches average a 72.2 percentile. Those in which I exercised my own "judgment" were in the 61.8 percentile--and that doesn't even include the disaster "crazy" picks I played in Bracket Madness. Including that, my hunches would be down around a 46.9 percentile.
2 Comments...


Mon, Mar 22, 2010, 6:05 PM	By ptiernan	0 Comments...
This year's Sweet 16 is significantly different from the 25 that have come before it in the 64-team era. For one thing, the starting units are older by graduating class than any since 1985. For another, the teams are hotter. No other field has entered the dance winning more games in their last ten than the 2010 Sweet 16. Finally, they're led by less experienced coaches, having been to the dance two years and the Elite Eight trip one year fewer than average. Here is how the 2010 Sweet 16 compares in 13 key stats to the Sweet 16 in the modern tourney era (rank against top year in parentheses): 2010 average seed: 5.00 (5th lowest, 1990 and 1999 - 5.50) 2010 average coach tourney trips: 6.02 (2nd fewest, 1986 - 6.00 years) 2010 average coach Elite Eight trips: 1.14 (2nd fewest, 1999 - 1.13 trips) 2010 average consecutive bids: 3.36 (3rd fewest, 2006 - 3.25 bids) 2010 average winning percentage: .821 (2nd highest, 2008 - .823) 2010 average wins in last 10 games: 7.47 (tied for most, 2008) 2010 average winning streak: 3.56 games (4th highest, 2007 and 2008 - 3.94) 2010 average points scored: 75.5 (tied 4th lowest, 1985 - 74.5) 2010 average points allowed: 64.1 (3rd lowest, 2008 - 62.9) 2010 average scoring margin: 11.4 (tied 15th lowest, 1985 - 9.1) 2010 average percent of points from starters: 78.9 (tied 12th highest, 2006 - 82.4) 2010 average percent of points from guards: 50.8 (7th highest, 2004 - 59.9( 2010 average age by class (1 = FR, 2=SO, 3=JR, 4=SR): 3.20 (oldest, 1997 second most - 3.10)
0 Comments...


Mon, Mar 22, 2010, 9:09 AM	By ptiernan	5 Comments...
After the first two rounds of the tourneyóand all the carnage of upsetsóthe leaders in our two contests are doing surprisingly well. In the Great Eight, here are the top five: 1. Nykes, 49 points 2. Buckeyenut61, 49 points 3. Samsanch, 44 points 4. Fun0147, 43 points 5. Bdance, 43 points A couple of interesting coincidences among these leaders. Nykes is from Bloomfield Hills, where I went to high school. And Bdance lives in my hometown, and frequents my favorite pub. (I owe you a bourbon, Barry!) As for my ranking, I'm smack dab in the middle of the pack, 106 out of 212, with just Cornell and Washington left. In the Bracket Madness contest, these are the top six: 1. Hehateme1979, 48 points 2. RandyT, 48 points 3. Ilchile, 47 points 4. Mstollak, 47 points 5. Anwatkins, 47 points 6. Teflonmike, 47 points I'd have to look back, but I think Hehateme1979 was in the running last year for one of the contests. Isn't that right, Hehateme? And where did I wind up? Well, if you'll recall, I vowed to submit a high risk/high reward bracket that reflected what I figured would be a crazy dance. So I advanced all the 11 seeds, three of the 12 seeds and two of the 13 seedsóto the Sweet 16! And I had BYU in the Final Four! What the hell was I thinking? To say that the strategy failed is an understatement. I'm lastódead last out of 193 peopleóin the pool. That takes a special gift. I was right about the unpredictability, wrong about where it would come from. I take solace in the fact that I didn't play this bracket in my friends' poolóand am third there with 44 points. And I didn't embarrass myself on CBSSports.com either, offering two stats-driven brackets in the mid-90s percentile and one in the mid-80s. (I should stick with what the numbers say, ) But dead last in the pool on my own website? Yeesh, Where's that Second Chance Dance button?
5 Comments...


Sun, Mar 21, 2010, 7:41 PM	By ptiernan	1 Comments...
Did you notice the Madometer on the home page. It's shot up at 19.6 percent Chitwoodin' madness. That's higher than the Madometer reading for the most unpredictable dance of all time--1986, which deviated from perfect high-seed sanity by 18.8 percent. Consider this: if the high seeds in every matchup won out for the rest of the tourney (not likely), this dance would still be 14.1 percent madness--above the average Madometer reading of 13.7 percent. Here's another indication of how crazy 2010 has been: there have already been nine upsets, above the average of 8.3 per tourney. Here are all the surprises to date (remember: we only count an upset if the seed gap is four or more): - 14 Ohio over 3 Georgetown - 13 Murray State over 4 Vanderbilt - 12 Cornell over 5 Temple - 11 Washington over 6 Marquette - 11 Old Dominion over 6 Notre Dame - 12 Cornell over 4 Wisconsin - 11 Washington over 3 New Mexico - 10 St. Mary's over 2 Villanova - 9 Northern Iowa over 1 Kansas Next round, every game except the Duke/Purdue tilt has upset potential...and beyond that, who knows? All we need is five more shockers to break the record for the most upsets in a single tourney. That record--13--occurred in 1985, 1986, 1990 and 2002. Should be a fun weekend coming up!
1 Comments...


Sat, Mar 20, 2010, 7:10 PM	By ptiernan	5 Comments...
Unbelievable! Kansas falls to Northern Iowa! Before this game, nine seeds were just 3-51 against top seeds in round one. Considering that 45 percent of America had the Jayhawks winning the tournament, this is the ultimate bracket buster. Way back in late December, here's what I said about Kansas: "I don't know why I'm so skeptical of the Jayhawks. But nothing I've seen so far changes my mind that this team won't be cutting down the nets in March. Kansas has yet to play a tough road game. Sure, UCLA was supposed to be hard, but the Bruins are just 5-7óand that game wasn't exactly a laugher. The Jayhawks also struggled with Memphis on a not-so-neutral St. Louis court. I'll be very interested to see how Kansas handles two big upcoming road games against Temple and Tennessee. If they get by those opponents in convincing fashion, I may have to readjust my prediction." I readjusted my prediction, even though the Jayhawks lost to Tennessee. I shouldn't have.
5 Comments...


Sat, Mar 20, 2010, 3:48 PM	By ptiernan	5 Comments...
I used to hate listening to Bill Raftery call a game. Now I can't get enough of him. Why? Because I write down all the crazy stuff he says...and he says so many bewildering things that I'm on the edge of my seat the whole game. Unfortunately, with all the flipping around from game to game, I'm not getting my fill of Raftery (more cowbell!), but there's still no lack of crazy comments. When he isn't riffing on classic catch-phrase clustering ("A little lingerie on the deck with the blow by...head and shoulder...shimmy and shake. Philly style! Goodness!), his head-scratchers fall into two categories: 1) bewildering nonsensical gibberish, and 2) unintentional (I think) "that's-what-she-said" statements. Here are my five favorite gems of gibberish: 5. They don't pack the tack. 4. How about this charm--and a little wit as well. Everybody knows...take it easy big fella! 3. Raindrops keep falling on my pillow! 2. There's a little tavernly discard... 1. A little pre-mature at the time...but some early onions. Sticking it, Mickey! And then, that's she said: 5. Villanova gets a toe job successfully. 4. Such a load! That's said affectionately, Vern. 3. Squeeze him. Check him out! 2. I would ride that guy! 1. He has stroked it admirably. And, finally, one last...well...um... "This will put you to sleep. The little kiss and the footwork. He can dance as well. Use the derriere! Phew!"
5 Comments...


Sat, Mar 20, 2010, 11:23 AM	By ptiernan	2 Comments...
After the first round, here are the results of the 12 bracket models we ran in the ESPN Tourney Challenge: 1. Pulse check - 250, 96.6 percentile 1. Upset/tossup - 250, 96.6 percentile 3. Upset special - 240, 90.1 percentile 3. Pythag - 240, 90.1 percentile 5. Final Four/champ - 230, 77.9 percentile 6. Pete's gut - 220, 59.8 percentile 6. Seed matchup - 220, 59.8 percentile 6. Keeper - 220, 59.8 percentile 6. Combopase - 220, 59.8 percentile 10. Contend/pretend - 210, 37.7 percentile 11. Factors - 200, 20.6 percentile 12. Pythag/coach - 180, 4.0 percentile Pulse Check and Upset/Tossup are performing the best--but neither would likely put you at the top of a pool with 50 people or more. To give you an idea of where these models might rank, 20 people of the 193 that played Bracket Madness have better brackets, with WorldCup82 and Ilchile leading the way at 27 points. I decided not to play any of my bracket models in Bracket Madness--and am paying for it dearly. I rank 180th out of 193, with 20 points. To add major insult to injury, my wife Michelle is tied to second with 26 points. She says her system involve a complicated strategy of spending three hours with the Excel sheet, focusing on the pulse check data, peppering me with questions...and having a couple glasses of wine. Why didn't I think of that?
2 Comments...


Fri, Mar 19, 2010, 12:21 PM	By ptiernan	3 Comments...
After the first day of craziness, here's how the 12 models rank: 1. Upset/tossup - 120, 96.9 percentile 1. Pulse check - 120, 96.9 percentile 1. Pythag - 120, 96.9 percentile 4. Upset special - 110, 87.4 percentile 4. Final Four/champ - 110, 87.4 percentile 6. Pete's gut - 100, 66.7 percentile 6. Seed matchup - 100, 66.7 percentile 6. Keeper - 100, 66.7 percentile 9. Contend/pretend - 90, 38.3 percentile 10. Combopase - 80, 13.2 percentile 11. Pythag/coach - 60, 0.4 percentile 11. Factors - 60, 0.4 percentile It's still early days. Last year, my gut bracket was 16 for 16 the first day...and that bracket wound up in the 55th percentile. There are some interesting developments here: the two brackets that focus on upsets are near the top--not surprising, considering the low-seed uprising yesterday. But "upset special" usually tanks--and it has Georgetown going far. Yikes. Pythag and Pulse Check are holding their own too...but when I factored in coaching to the Pythag numbers--tanksville. We'll see where these all rate after today's nuttiness.
3 Comments...


Thu, Mar 18, 2010, 9:34 PM	By ptiernan	3 Comments...
We're only 12 games into the 2010 tourney and it's already more unpredictable than the entire 2007 dance. That's right: 12 games have resulted in more craziness than all 63 games in 2007. Amazing. In fact, right now, the Madometer reads 35.4 percent Chitwoodin' madness. Remember: 1986 was the most insane tourney ever--and that only pushed the Madometer to 18.8 percent. If you're a Big East fan, you're feeling particularly upset about the evening. Three seed Georgetown and six seeds Notre Dame and Marquette all were shocked tonight. And two seed Villanova was taken to the brink by Robert Morris. I think we've all heard so many so-called experts tell us how dominant the Big East is, that the conference might've gotten more credit than it deserved. Georgetown, Villanova, and Notre Dame looked surprisingly flat. Marquette scrapped and played well, but lost to an underrated Washington squad. And haven't we been hearing all season about how weak the Pac-10 is? I've said it once, and I'll say it again: nobody knows anything. I can't wait to see how high the Madometer shoots up by the end of the night. If you'll notice, I designed it on the home page just to go up to 20. I think I'm going to have to redesign it!
3 Comments...


Wed, Mar 17, 2010, 3:48 PM	By ptiernan	5 Comments...
I finally got the model monkey off my back and found the time to compare this year's tourney field to all the past tourney teams of the 64-team era. The findings provide more evidence that 2010 will be one crazy dance. Take a look at this chart comparing 2010 to the historical averages for 16 key stats: The 2010 field comes into the tourney winning more games in their last 10 than any other field since 1985. The two closest years were 1988 and 2000. Guess what happened in those years? In '88, sixth-seeded Kansas cut down the nets (quick: find a worthy six seed with an All-American big man...Notre Dame?). In 2000, fifth-seeded Florida and two eight seeds, Wisconsin and North Carolina, reached the Final Four. So...the 2010 equivalent would be something like Michigan State, Texas and Gonzaga getting to the semis. Madness! Need more reason to feel squeamish about your chalky bracket? The 2010 tourney field has the second best winning percentage of any field since 1985. The only year when the winning rate was better was 1998. That year, Kentucky cut down the nets. They were a two seed...and the rest of the Final Four was filled out by a one (NC), a three (Stanford) and a grind-it-out four (Utah). Let's see: how do you feel about Kansas, Kansas State, Baylor and Wisconsin--with the Wildcats winning it all? Here are a few other things that jump out. The 2010 tourney field owns the third highest scoring margin, eclipsed only by--you guessed it--1988 and 2000. Whoa. And the coaches in this field own one of the lowest Elite Eight advancement rates of any tourney. The lowest? 1998. Now hang on just a second! Finally, this is the highest scoring tourney field since 2003. That was the year Syracuse won the tourney despite not going to the previous dance. The only other time that happened was 1986--the most unpredictable tourney ever, according to Madometer readings. Hmm. (At least that bodes well for "missed-last-dance" Kentucky.) So the 2010 tourney field would seem to compare most closely with 1988, 1998 and 2000. If you average their Madometer readings, you'll find that they were 14.7 percent insane. That's above the average Chitwoodin' madness of 13.7 percent. Can you tell I'm a little worked up?
5 Comments...


Wed, Mar 17, 2010, 2:20 PM	By ptiernan	0 Comments...
Seems like the last thing I get around to doing during Tourney Forecasting week is our very own bracketscience.com contests. I love Great Eight. It's such a refreshing break from bracket pondering--and I get to go out on a limb and pick some Cinderellas (not that my Keeper bracket isn't chock full of those). So far, I've narrowed down to 24 teams. I'm getting there. We're also debuting a bracket challenge of our own this year: Bracketmadness. Don't forget to submit your picks for each game. Fair warning: I reserve the right to play a completely oddball bracket in this contest. Rest assured, though. I'll report out on the progress of the 12 bracket models as the tourney progresses. I just loaded them all into the ESPN Tourney Challenge.
0 Comments...


Wed, Mar 17, 2010, 12:43 PM	By ptiernan	3 Comments...
I just posted my twelfth and final bracket--the one I'll play for keeps in local pool. As you'll see, it steps right into the deep end. BYU over Syracuse in the West? Baylor over Duke? Kentucky finally getting some love? I intentionally went contrarian this year; just got tired of seeing too much Duke and Kansas. Danger, Will Robinson...
3 Comments...


Wed, Mar 17, 2010, 10:16 AM	By ptiernan	7 Comments...
The Bracket Doctor is in...and he feels your pain. He knows how much you're anguishing over Cornell and Temple. He knows that you can't find anyone you like to advance in the West region. And he's painfully aware that Duke's road to Indy seems too easy. Let's talk it out tonight for an hour at 7:30pm ET on the Fan Chat under "Play Time." I'll also be IM-able on the Facebook fan page throughout the day in between radio interviews. Check out http://www.facebook.com/BracketScience and gimme a ping. You may need to "Friend" Pete Tiernan before we can link up in IM. Get your questions ready for tonight!
7 Comments...


Tue, Mar 16, 2010, 8:26 PM	By ptiernan	10 Comments...
I just posted the Pulse Check Stats model...and--what do you know?--it didn't pick Duke. Okay, so it did like Kansas, the other team my stats seem to fixate on. I get a chance tomorrow to rectify all these biases in the numbers with my Keeper bracket. I don't know yet if I'll have Kentucky cut down the nets, but I'll certainly give them more respect than my models do. I know that they didn't go to the previous dance, and that's general a big sign of a pre-mature exit, but this is a special team. Sort of like that Syracuse squad of 2003 that cut down the nets with a younger starting unit and a freshman phenom. Sound familiar? Sometime before noon, I'll also post a "chat" schedule. You can get to chat on the site by clicking on "Play Time" then "Fan Chat." More on that soon.
10 Comments...


Tue, Mar 16, 2010, 7:25 PM	By ptiernan	2 Comments...
This is the third and final installment of my pulse check. Unlike the stretch run and champ week reviews, this is based on the actual teams in the tourney field. The data also forms the basis of the last statistical model I'll post within the hour. I review each team against 14 key stats that foretell strong tourney performance. Here are what the columns mean in the chart below: • 09 - Did they go to last year's dance? (An "N" gets a red disqualifier) • CY - How many tourney trips has the coach made? (Should be three or more) • E8 - How many times has the coach been to the Elite Eight? (At least once) • CA - What's their conference affiliation? (Should be "B" for Big Six) • AP# - What's their AP rating? (The top 15 avoid a red disqualifier) • SOS - What's their strength of schedule (Anything worse than 40 is bad) • PYT - What's their possession-based Pythag ranking (Should be top 15) • OE - Where do they rate for offensive efficiency? (Must be 25 or better) • DE - Where do they rate for defensive efficiency? (Must be 25 or better) • PPG - What's their point scoring average (Should be 75 points per game or more) • MAR - What's their average scoring margin (Needs to be better than 10 points) • FG% - What's their field goal percentage? (Anything lower than .450 is a warning sign) • G% - What percentage of points do they get from guards? (Too reliant--higher than 60%--and they're at risk of underachievement) • MO - How's their current momentum (I gave any team with four losses in the last 10 or two in a row a minus) • DQ - A tally of all the red highlights for each team For the upcoming bracket, I'll use the number of disqualifiers to advance teams. Generally speaking, the team with fewer disqualifiers moves on...except when we reach the Sweet 16. Then, I do a different analysis for teams within two disqualifiers of each other. I'll explain in the model. And here's the chart, with team write-ups to follow:
2 Comments...


Tue, Mar 16, 2010, 2:13 PM	By ptiernan	4 Comments...
I'm putting together the new "Contender/Pretender" bracket model, so I had to divde the teams into the two categories by seed, according to the guidance provided in the "Contender/Pretender" piece under "Feature Articles." For your reference, here's who the numbers say are contenders and pretenders at their respective seed positions: I'll use this information, plus seed matchup guidance in situations where contenders play contenders or pretenders play pretenders to arrive at a bracket. Early indications are that it won't be Duke in the finals. That's a relief.
4 Comments...


Tue, Mar 16, 2010, 12:24 AM	By ptiernan	5 Comments...
I've completed nine of the 12 models I committed to cranking out. I could sure use a Kramerica-like intern for these couple of days. There are still three models left: - Model 8 based on Contender/Pretender rules - Model 11 based on the Pulse Check disqualifiers I've been tracking in the blog - Model 12...the combination of all this heavy-duty number crunching with my gut instincts and a little sentimentality thrown in. What the hey! One development that has emerged with the first nine models I've cranked out is that Duke keeps coming up as the winner. If you take away my Gut Instinct bracket, which goes with Kansas, and the Upset Special, which takes a stab at the Hoyas, five of seven numbers-based brackets come up Dukie. I've never had that kind of alignment with all the models...and, frankly, I'm skeptical. Here's my guess why it's happening: 1. Duke has the best Pythag numbers of any team in the dance 2. Duke's PASE numbers are solid...so it never slips in any model factoring in PASE 3. In Duke's region, second-seeded Villanova's poor momentum numbers often disqualify it from taking on the Blue Devils in the Elite Eight. 4. The fact that Kentucky didn't go to the previous dance and has relatively weak Pythag numbers for a top seed gives Duke a consistent edge in the Final Four head-to-head matchup. That's all well and good...but what am I going to do about the bracket I ultimately decide to play in my pools. I've yet to decide. Do I have faith in all the models I've created and stick with Duke--or go contrary to what stats say, recognizing the potential craziness of this dance--and pick a longer shot. At this point, I'm not even sure.
5 Comments...


Mon, Mar 15, 2010, 6:46 PM	By ptiernan	0 Comments...
I've posted the updated Excel sheet under "2010 Resources." In addition to all the other stats you received in the first version, this one includes the new RPI and SOS data, as well as identifying the teams with All-Americans. At about 6:00pm ET, the US Basketball Writers Association announced their two teams. Here they are: FIRST TEAM Sherron Collins Wesley Johnson Scottie Reynolds Evan Turner John Wall SECOND TEAM Cole Aldrich James Anderson DeMarcus Cousins Luke Harangody Jon Scheyer
0 Comments...


Mon, Mar 15, 2010, 5:22 PM	By ptiernan	0 Comments...
Not that anyone cares, but the problem was some hard-coded quotation marks in our HTML. Actually, user bamilus discovered it minutes before we did. Thanks, man.
0 Comments...


Mon, Mar 15, 2010, 4:57 PM	By ptiernan	1 Comments...
We threw in the "download PDF" feature for the models late yesterday and thought it all worked fine. Now we're finding that it has some problems with certain browsers. (Works fine on my Mac!) We're investigating and will come up with some solution--even if it's posting a zip file with all the PDF's together. Answers soon...
1 Comments...


Mon, Mar 15, 2010, 3:11 PM	By ptiernan	1 Comments...
What would Mad Monday be without one model glitch. Long-time member Don Northrop pointed out that, while I wrote about Syracuse and Kentucky in the Finals of the Upset/Tossup model, I had Duke etched into the graphic. I fixed that. So if you've downloaded the PDF, you should re-grab.
1 Comments...


Mon, Mar 15, 2010, 11:17 AM	By ptiernan	6 Comments...
About a month back, I started comparing the top 32 ranked teams based on Ken Pomeroy's efficiency data to the average of the top 32 ranked teams over the previous six tourneys. What I found was that the higher performing teams this year were weaker on average than previous elite teams, while the lesser squads were better. I promised back then that I'd do the analysis on the actual teams that made the dance--and extend the comparison out to 52 teams. That would cover the squads that are theoretically seeded one through 13. Here's what I discovered" Pretty dang telling. The four highest ranked teams (those falling into the column with the orange "1") are comparable to the the average top four from the 2004 through 2009 tourneys. But then the two, three, four and five "seed" groupings are considerably weaker than in past dances. The six and seven grouping are pretty much the same--and then we see a dramatic shift. From the 29th team all the way through the 49th team, theoretically the set of squads seeded eight through 12, this year's field is stronger than past fields. That's particularly true for the 10, 11 and 12 seed groupings. Everyone has been saying that this dance will be more wide open than the recent tourneys. This chart quantifies it. If 10-12 seeds are stronger than 2-6 seeds, what does that say about opening round matchups like the 5v12 and 6v11 tilts...or second round tilts like 2v7, 2v10, 4v12, 5v13 and 3v11. Can't wait to find out.
6 Comments...


Sun, Mar 14, 2010, 9:55 PM	By ptiernan	6 Comments...
I just posted the preliminary Excel sheet. It has all the stats you'll need except the RPI and SOS values (waiting for the NCAA update) and the All-American information. That's coming tomorrow, when the US Basketball Writers Association announces their picks. If you absolutely need the Excel sheet tonight, go ahead and grab it. But the final sheet with all the stats (and any fixes I might come across) will get posted tomorrow. As for the models, I'm starting to post them. My "gut" model will be up soon--and it's a snoozy affair...a few upsets in the first two rounds, but landing on Kansas, Kentucky, Kansas State and Baylor (my only flyer). It will be interesting to see what the statistically-based models come up with. One new feature: I'm creating printable PDFs for all the models to help those of you who've asked for some way to print these off.
6 Comments...


Sat, Mar 13, 2010, 3:24 PM	By ptiernan	8 Comments...
Believe it or not, most bracket pools are won by someone who gets between 48 and 54 correct picks out of the 63 tourney games, depending on the size of the pool. This works out to a 76 to 85 percent accuracy rate. If that seems lower than you expected, remember that inaccurate picks in early rounds reduce the number of later-round games in which you even have a chance to be right. So three out of four ain't too bad. If 48 out of 63 is the benchmark of winning bracket performance, what constitutes average tourney prediction success? When you consider that even the most clueless bracketeer could adopt a strategy of advancing the higher seed in all matchups, then that should stand as the baseline for tourney prognostication. In this case, every top seed would reach the Final Four. To determine winners in the semis and finals, it would be reasonable to advance the team with the higher victory margin, since PASE stats show that this is the best tourney performance indicator. I went back over the last 25 tourneys of the 64-team era and figured out how well the "higher seed/bigger margin" strategy would do. Here's the year-by-year breakdown: The numbers show that you should be able to predict 41 or 42 out of the 63 tourney games correctly without even thinking about it. In fact, the baseline bracket picking strategy would've correctly identified eight of the 25 champions in the modern tourney era, including North Carolina last year, Kansas in 2009, Florida in 2007 and North Carolina in 2005. Heck, if you had used the higher seed/bigger margin strategy in 2008, you would've correctly pegged the Final Four, called both the finalistsóand picked the winner. Of course, this kind of late-round accuracy is an anomaly. Before 2008, all four top seeds had never reached the Final Four. And let's face it: picking your bracket solely by seeding isn't exactly a winning strategy. Using this approach would likely put you in contention for smaller pools in only two of the 25 yearsóin 2007 (49 correct answers) and in 1993 (48 correct answers). There's no denying that seeding is the single best determinant of a team's tourney fate. It doesn't take a bracket scientist to figure out that the one, two, three and four seeds have accounted for 88 of the 100 Final Four teams and won 23 of 25 tourneys in the 64-team era. The real the question is: which factors beyond seeding help improve your ability to predict tourney outcomes above the baseline accuracy of 66 percent? That will be my goal with the 12 bracket models I unveil late Sunday night and throughout Monday.
8 Comments...


Fri, Mar 12, 2010, 12:12 AM	By ptiernan	2 Comments...
I said it might happen. Villanova needed some convincing conference tourney wins to keep its scoring margin above 10 points a game--and remain as one of the five teams that meets all the six criteria of 17 of the last 19 champs. It didn't happen. The Wildcats' average scoring margin stands at 9.7 point a game heading into the dance. And that means there are only four teams now that have all the credentials of nearly every tourney champ. That would be Kansas, Duke, Syracuse and Maryland. The Terps will likely fall out too...unless they grab a top-four seed. Regardless, this will be the fewest teams to meet champ critera since I've been tracking the stats. To put things in perspective, last year at this time, seven teams had the six qualities of nearly every tourney champion (and North Carolina was one of them).
2 Comments...


Tue, Mar 9, 2010, 10:32 PM	By ptiernan	3 Comments...
With just four days until Selection Sunday, I thought I'd resurrect the review I did about three weeks ago, evaluating the top teams based on 12 key stats that foretell tourney success or failure. For those who don't want to dig back through past blogs, here's the deal: I take a look at all the AP Top 25 teams and any team off that list that has a top 25 possession-based Pythag ranking. Then I slot them into categoriesófive favorites, five challengers, five contenders, five scrappers, five sleepers, five pretenders and the two remaining teams, which I call omissions. It goes without saying that this is a dynamic list. As champ week unfolds, I'll no doubt have different opinions. But at this moment, with 4 days, 17 hours, 40 minutes and 10 seconds until Selection Sunday, this is where my head's at. I've reviewed 32 teams through the infamous "eye test," as well as 12 key stats that foretell tourney success or failure. In the chart below, a red highlight means that a team fails to meet a basic criterion for overachievement in the dance. The last column, "DQ," tallies up all the disqualifiers. Here's what each column means and the requirement: • The first column tracks whether the team has risen or fallen since the last review (a red down arrow means at least a two-group drop; a green up arrow is a two-group rise; black arrows are one-group moves) • 09? - Did they go to last year's dance? • CY - How many tourney trips has the coach made? (Should be three or more) • E8 - How many times has the coach been to the Elite Eight? (At least once) • CO - What's their conference affiliation? (Should be "B" for Big Six) • AP# - What's their AP rating? (The top 15 avoid a red disqualifier) • SOS - What's their strength of schedule (Anything lower than 40 is bad) • PYT - What's their possession-based Pythag ranking (Should be top 15) • OE - Where do they rate for offensive efficiency? (Must be 25 or better) • DE - Where do they rate for defensive efficiency? (Must be 25 or better) • PPG - What's their point scoring average (Should be 77 points per game or better) • MAR - What's their average scoring margin (Needs to be better than 10 points) • MO - How's their current momentum (I gave any team with four losses in the last 10 a minus, three losses earned a zero, two or fewer a plus). And here's the chart, with write-ups on each group to follow (they're considerably more abbreviated than the last time): FAVORITES The only change in this group from the stretch-run review is that I elevated Ohio State into this group and dropped Villanova. With Turner back, the Buckeyes' offense seems much more efficient. The only big chink in OSU's armor is a soft SOS. Kansas, Duke and Syracuse still have the inside track. And I've kept Kentucky in the upper echelon, despite the nettlesome fact that only two of 25 teams have won the championship after not making the previous dance. CHALLENGERS Two new teams (besides the falling Wildcats) enter the ranks of challengersóteams that have what it takes to knock the favorites off their perch. One is Wisconsin, which only has one warning statóa low scoring average. The other is Maryland. The Terps are a big advancer since the stretch-run review, when they weren't even ranked in the AP Top 25 and I tabbed them as a sleeper. Who fell out of this group? Purdue dropped after losing Robbie Hummel. Pity, because I probably would've had them as a favorite otherwise. The other is Texas, which has fallen all the way out of the Top 25óand is now, oddly, listed as a sleeper. CONTENDERS Only one team that I tabbed as a contender last time has remained there. That would be Kansas State. They could've climbed to the rank of a challenger, but the Wildcats have stumbled a little of late. The big movers in this group are BYU, Tennessee and Baylor. I had BYU and the Volunteers as pretenders three weeks ago. But Tennessee's win over Kentucky and BYU's consistently strong possession-based stats have made me a believer. SCRAPPERS This is a new group. I describe scrappers as teams that I don't really think have a chance to reach the Elite Eight, but won't disappoint against seed expectations and could reasonably reach the Sweet 16. For Georgetown, this designation is a step down from where they were in the stretch-run review. Xavier, Temple and Texas A&M were all tabbed as "omissions" last time. And UTEP wasn't even ranked, because they were neither in the Top 25 nor among the most efficient teams. SLEEPERS Only one team that was considered a sleeper at the last pulse check is still in this group. That would be Missouri. The Tigers only have three disqualifying statsóand one of them is not being ranked. It might be surprising to see the Longhorns in this list, but considering their numbers and how much the experts have been running them down, it wouldn't surprise me to see Texas overachieve in the dance. The other three teamsóMarquette, Florida and Californiaówere considered omissions in the last review. California is particularly interesting. The Bears only have three statistical flaws (not ranked, low defensive efficiency and scoring margin), and that's a small amount for a team that no one is talking about. Heck, check out how often Michigan State is on the wrong side of the key stats and I've still got the Spartans as a challenger (all hail the Power of Izzo). PRETENDERS New Mexico, Gonzaga and Butler still make my list of pretenders. And they're joined now by Pittsburgh and Vanderbilt. I want to like both the Panthers and Commodores, but the numbers are rugged. Pittsburgh is a low-scoring and relatively inefficient offensive squad. Vanderbilt's poor defensive efficiency and overall Pythag strength give me pause. If I get time, I'll do this review after Selection Sunday. I'll definitely post the chart, I just may not have time for an extensive review, however.
3 Comments...


Mon, Mar 8, 2010, 8:12 PM	By ptiernan	2 Comments...
On the last stats champ check before Selection Sunday, five teams still have the six basic credentials that 17 of the last 19 champions have possessed: 1) They came from a Power conference (ACC, Big East, Big Ten, Big 12, SEC or Pac 10) 2) They went to the previous year's tourney 3) Their coach had at least five tourney appearances 4) They scored at least 77 points a game 5) Their average margin of victory was 10 points or more 6) They were seeded one through four The only teams that meet all these criteria are Kansas, Syracuse, Duke, Villanova and Maryland. Check out today's results (I'm assuming no team outside the AP Top 25 has a shot at a four seed): All the teams with "Y's" highlighted in dark blue under "Old Rule" have the six attributes cited above. These same five teams also fulfill the extra requirement I've been tracking this yearóthat their coach has made an Elite Eight run in the past. That's what the "New Rule" column signifies. Could any of this Fab Five fall off the list come tourney time? Absolutely. While Kansas, Syracuse and Duke are locks to keep their potential champ credentials, Maryland and Villanova could easily drop out. Right now, Maryland is probably on the outside looking in for a four seed. Lunardi has them at a fiveóand that's historically been a terrible position from which to make a deep run (only a handful of five seeds have gone beyond the Sweet 16óand there have been no champions). Villanova looks like it will safely secure a three seed. But the Wildcats' problem is scoring margin. Right now, their margin stands at just 10.2 points per game. They're just five points away from falling below a ten-point per game average. In the Big East tourney, close games are the order of the day. I've also included the possession-based Pythagorean winning percentage of the top teams, as a way of comparing whether they deserve their lofty rankings. Pythag has helped provide early indications of teams that should make the AP Top 25. This week, there are seven teams in the AP Top 25 whose Pythag rank suggests they shouldn't there. The most glaring "over-rated" team is New Mexico. They're number eight this week, but rank 40th in Pythag winning percentage. That's a 32-position difference. Other undeserving Top 25 teams, based on tempo-free data, include (highlighted in red above): Gonzaga (30-position gap), Butler (18), UTEP (14), Tennessee (12), Vanderbilt (12) and Pittsburgh (10). Which teams would replace them if efficiency data ruled the rankings? Missouri (#13 in Pythag), Texas (#14), California (#16), Clemson (#17), Florida State (#18), Marquette (#22), and Utah State (#24) all have better Pythag winning percentages than the efficiency pretenders cited above.
2 Comments...


Sun, Mar 7, 2010, 1:35 PM	By ptiernan	0 Comments...
Here's part three of my examination of the value of scoring margin. As promised, I broke down the scoring margins for the second-round matchups. Here's a look: If you combine the two matchups of the top four seeds, you'll find that one seeds own an overall round-two scoring margin of 11.9ppg, two seeds have a 5.9ppg margin, three seeds are at 4.6ppg and four seeds squeak by at 1.7ppg. To put those numbers in perspective, the margin of one seeds is comparable to that of three seeds in round one. Fittingly, their winning percentages are similar: in 3v14 matchups, three seeds own a 11.4ppg scoring marginóand an 85% success rate. In their 1v8 and 1v9 round two matchups, one seeds own a slightly higher margin (11.9ppg) and a slightly better success rate (88%). To continue with the comparison, two seeds play the second round (5.9ppg margin) a little better than five seeds taking on 12 seeds in round one (4.8ppg margin). And, not surprisingly, the winning percentages are similarly ordered. Two seeds own a slightly higher winning rate (.667) in round two than five seeds in round one (.660). I also wanted to see if a seed difference in round two matchups was comparable to that of round one. As it turns out, the numbers are very close. Remember: in round one, a seed difference of one in a matchup was worth 1.17 points to the higher seed. If you do the math in round two, a single seed difference is worth 1.15 points. For one seeds, the difference is bigger (1.58) and for four seeds, it's lower (0.52). That's more or less consistent with what happens in round one. So, if we know that a single seed difference is worth 1.15 points or so, we can project what the point spread might be in any matchup, irrespective of the teams. In a 2v7 matchup, with a seed difference of five, we might project that the two seed will beat the seven by 5.75 points. Turns out, the actual margin is 5.7ppg. Of course, this doesn't work as well for all match-ups, but it's a nice starting point in assessing whether actual point spreads are reasonable or not.
0 Comments...


Sun, Mar 7, 2010, 11:29 AM	By ptiernan	0 Comments...
I've been saying for the last couple years that, aside from seeding, the one stat that most foretells a deep tourney run is scoring margin. While it's true that the raw number itself doesn't necessarily signify a team's dominanceómany prefer to look at scoring margin adjusted for strength of scheduleóit turns out that it's a pretty handy stat for assessing overachievement. Take a look at the chart below. With each additional scoring margin point by which you filter the tourney teams, their performance against seed expectations increases. Teams that beat opponents by four or more points per game barely exceed projected win totals (+.005 PASE). But teams that own a scoring margin of 10 or more are better overachievers (+.083 PASE). And PASE values accelerate all the way up to a 15 point-plus scoring margin, where the 118 teams possessing this criterion own a healthy PASE of +.399. Furthermore, 43 of them have reached the Final Four and 13 have won the championship. So, which 2010 teams have scoring margins of 15 points or more a gameóand a better chance to overachieve? Looking at data from the last AP Top 25, only four teams meet this key credential: Syracuse, Kansas, Duke and BYU. If will be interesting to see if they're able to defy the expectations like their past counterparts.
0 Comments...


Sat, Mar 6, 2010, 8:18 PM	By ptiernan	0 Comments...
If you haven't poured over the 2010 Seed Guide yet, you might not have noticed that I've included the scoring margins for every seed matchup since 1985. These margins help explain why certain matchups are lopsided and others are tight. It's not surprising, for example, that one seeds own a 100-0 record against 16 seeds in round one; their scoring margin is 25.4 points a gameónearly 10 points more than their pre-tourney average. It's one thing to read the numbers; it's another to see them compared to one another, side by side, in a chart. Here are the scoring margins of the eight first-round matchups. Take a look at the dark orange bars. It's pretty clear why I've been saying that you really shouldn't pick a seed higher than four to get upset. Two seeds own a 16.8 scoring margin over 15 seedsóand have a 96-4 record against them. Three seeds have a scoring margin that's more than five points loweróand their record is .110 worse, at 85-15. Four seeds have a scoring margin of 9.3 points over 13 seeds and a record that's just .060 lower than three seeds (79-21). For one and two seeds, the scoring margins over their round-one tourney competitors are much bigger than their regular season margins. (The average one seed has a 15.6ppg margin heading into the dance; two seeds have a 12.6ppg margin). For three and four seeds, their first round games are comparable to the typical games they play in the regular season (Three seeds have a regular season scoring margin of 11.3ppg while fours have an average margin of 10.1ppg). When you look at the scoring margins of 5v12 and 6v11 matchups, it's no wonder those are the key games to spot Cinderellas. Five seeds only beat 12 seeds by 4.8 points on averageóand own just a .660 winning record against them (66-34). Six seeds have a slightly tighter average margin (4.3ppg) but a better record (69-31). And the 7v10 and 8v9 are fittingly termed toss-up games, considering their tight margins. I looked at the issue of margin from another angle, and this is what those ghosted orange bars are all about. They measure the difference between the higher seeds' and lower seeds' regular-season scoring margin. For instance, top seeds own a 15.6ppg pre-tourney margin; 16 seeds have just a 2.4ppg margin. The difference is 13.2 pointsóand that's by far the widest gulf between a higher seed and a lower seed. It helps explain why one seeds are a lock against 16 seeds. And take a look at the last four matchups. None of them have a margin difference bigger than half a point. No wonder the matchups are so hotly contested. I thought about this another way today (while I was sitting in a movie theater watching a movie I didn't want to seeólong story). What is the margin value of a single seed gap? Think of it this way: one seeds beat 16 seeds by an average of 25.4 points a game and they're 15 seed positions better. That works out to 1.69 margin points per seed gap. Here's how all the round-one matchups break down: 1v16 - 25.4ppg, 15 seed difference = 1.69 points per gap 2v15 - 16.8ppg, 13 seed difference = 1.29 points per gap 3v14 - 11.4ppg, 11 seed difference = 1.04 points per gap 4v13 - 9.3ppg, 9 seed difference = 1.03 points per gap 5v12 - 4.8ppg, 7 seed difference = 0.69 points per gap 6v11 - 4.3ppg, 5 seed difference = 0.86 points per gap 7v10 - 2.7ppg, 3 seed difference = 0.90 points per gap 8v9 - 0.1ppg, 1 seed difference = 0.10 points per gap If you average all of these matchups, you'll find that the value of a single seed difference is 1.17 margin points. That might come in handy if you're trying to predict how, say, a 3v6 matchup might play out. Without really knowing the teams, I'd say the point spread might be about 3.5 points. Now, I should probably apply this same analysis to the second round and find out if that 1.17 point per gap value holds. By golly, I think I will, but later this week. I've got to finish my night watching a good movie. It is, after all, Oscar weekend.
0 Comments...


Mon, Mar 1, 2010, 10:28 PM	By ptiernan	4 Comments...
Last Friday, I wrote a blog pointing out that this year's top ten teams in terms of efficiency numbers had considerably lower Pythag values than their counterparts from the past six tourneys. One user, my man Spiggz, wondered whether the relative weakness of this year's squads would extend beyond the top ten, down to the top 50óor whether we'd see lesser 2010 teams with better Pythag values than 2004-09 tourney teams. I didn't look at the top 50 for the simple reason that all those teams won't make the tourney. In most years, the top 30 Pythag squads reliably get to the dance. Lower than 30, and it gets iffy. So if I were to compare this season's top 50 to the top 50 ranked Pythag teams for each of the last six tourneys, the lower ranked 2010 teams would have an unfair advantage. So I decided to compare today's top 32 ranked teams in terms of efficiency statistics to their counterparts over the last six tourneys. And here, in a nutshell, is what I found out: 1. This year's top two teamsóDuke and Kansasóare comparable to top-two teams of the past six tourneys. 2. From the third- to the 28th-ranked team, the 2010 squads are less efficient than their counterparts from years' past. 3. The trend shifts from the 29th to the 32nd best Pythag team, with these lesser ranked squads rating out better than similar squads from 2004-09. I didn't add the values to the chartómainly because it would've made it a complicated mess. But the difference in the lines tell the story. Check it out: What does this imply for the upcoming dance? Well, Spiggz had a good suggestion. Once the teams are announced, I'll do this same comparison from the top-ranked to the 64th ranked tourney team. And if we see what I think we mightóthat teams seeded two through seven are weaker than usual while those seeded eight through 13 are strongeróthen it could be one more indication that we're in for an upset-laden March.
4 Comments...


Mon, Mar 1, 2010, 8:49 PM	By ptiernan	4 Comments...
Based on today's AP Top 25, just five teams have the six basic credentials that 17 of the last 19 champions have possessed: 1) They came from a Power conference (ACC, Big East, Big Ten, Big 12, SEC or Pac 10) 2) They went to the previous year's tourney 3) Their coach had at least five tourney appearances 4) They scored at least 77 points a game 5) Their average margin of victory was 10 points or more 6) They were seeded one through four As I warned could happen last week, Texas dropped off the list after failing to crack the AP Top 25. I didn't see Maryland, though. I knew they were a sleeperóI pointed that out in the February 20 blogóbut I didn't realize that once they made the top 25, they would have the numbers to make the potential champ list. Check out today's results: All the teams with "Y's" highlighted in dark blue under "Old Rule" have the six attributes cited above. That includes: Syracuse, Kansas, Duke, Villanova and Maryland. These same five teams also fulfill the extra requirement I've been tracking this yearóthat their coach has made an Elite Eight run in the past. That's what the "New Rule" column signifies. For the last month or so, I've also included the possession-based Pythagorean winning percentage of the top teams, as a way of comparing whether they deserve their lofty rankings. Pythag has helped provide early indications of teams that should make the AP Top 25. Last week, I noted that both Maryland and Xavier had the possession-based numbers to be among the AP's elite. Lo and behold, this week, they've made the list. This week, there are seven teams in the AP Top 25 whose Pythag value suggests they shouldn't be there. The most glaring "over-rated" team is New Mexico. They're number eight this week, but rank 37th in Pythag winning percentage. Other undeserving Top 25 teams, based on efficiency data, include (highlighted in red above): Butler, Tennessee, Pittsburgh, Gonzaga, Temple and UTEP. Which teams would replace them in a Perfect Pomeroy Pythag world? Missouri (#12 in Pythag), Texas (#13), Clemson (#15), Florida State (#16), California (#18), Marquette (#22), and Georgia Tech (#23) all have better Pythag values than the efficiency pretenders cited above.
4 Comments...


Fri, Feb 26, 2010, 10:14 PM	By ptiernan	0 Comments...
Nothing drives home a point like a good picture. In several articles, I've tried to stress the dominance of one seeds in the later rounds of the dance. Here's a pie chart animation of advancers by seed from the Sweet 16 to the championship. Look at how the red top seed slice grows...and how seeds four and lower shrink: Three blog entries in one day. Okay...I'm procrastinating the article analyzing the value of tempo-free data. That's for tomorrow. Gotta finish watching the Canada-Slovakia hockey game. Hey...college hoops isn't all there is to the sporting world.
0 Comments...


Fri, Feb 26, 2010, 7:06 PM	By ptiernan	3 Comments...
A couple weeks back, I looked at the average Pythag of the top 10 most efficient teams in college ball today and noted that it was among the lowest of the previous six tourneys. The average has since climbed and dipped, then climbed again. Today, the average stands at .9643 for the top 10 rated teams. The average for the top 10 rated teams going into the 2004 and 2009 tourneys was .9719. This means, all things staying constant, that the average elite team this year is not as strong in terms of possession-based statistics as elite teams over the past six tourneys. That could indicate a more wide-open tourney. I dug a little deeper this evening and discovered something perhaps even more telling. I noticed when I graphed the 2010 top 10 Pythag teams, that there was a big gap between the top two teams (Duke at .9827 and Kansas at .9797) and the third team (Syracuse at .9645). So I went back over the last six tourneys and came up with the average Pythag for each of the top ten ranked teams. Then I compared those ten averages with the Pythag values of the top ten ranked teams this year. Here's a chart that compares the last six years (in gold) to this year's ten most efficient teams (in blue): Duke is a little below the average top Pythag team...but there's still time for the Blue Devils to improve their number. And Kansas is actually a better second-ranked team than the six-year average of .9768. But that's where the similarities end. For the next eight positions, the 2010 teams are considerably less efficient than their counterparts over the last six years. Syracuse, this year's third most efficient squad with a Pythag of .9645, is a sizeable .0123 behind the average third-seeded team. And the gap between Wisconsin (.9622) and other fourth-most efficient teams is even bigger (.0138). Where does Kentucky fall in all of this? They're the seventh-best Pythag team this year at .9590--but that's .0097 worse than your typical seventh most efficient team. Think of it this way: the average tenth-most efficient team between 2004 and 2009 is as strong as the fourth best team this year. On the face of it, this huge dip between the top two teams and the rest of the elite squads could mean that the dance will be more unpredictable than in recent years. If the two and three seeds are weaker than usual, it could open the door for six and seven seeds. As I look down the rankings of teams based on Pythag, I see schools like Marquette (21), Michigan State (23), Butler (24), Temple (25), and Pittsburgh (28) ranked in positions that six and seven seeds would occupy if the brackets were seeded by Pythag. All of those squads are capable of beating some of those top ten Pythag teams. What I'll be really interested to find out is how Kentucky fares in the dance. The Wildcats already have the fact that they didn't go to the previous dance as an underachievement sign. Now, they're a less-than-average elite team. Can they buck the odds against them? Seems like I've been wondering about that all season long.
3 Comments...


Fri, Feb 26, 2010, 4:31 PM	By ptiernan	0 Comments...
Last year, I did a feature article rating the combinations of two and three attributes that led to the best performance against seed expectations. I'm not going to go through a rigorous analysis of stats power combos this year, because I want to dig into possession-based statistics over this weekend. Then, after Sunday, all Hell pretty much breaks loose for me with Selection Sunday closing in and I won't have time for any more exhaustive research. I had a little time last night, though, and decided that I'd do some quick queries on the Bracketmaster and my more comprehensive Filemaker database to find the combinations of two stats that led to the highest PASE values. For this analysis, I focused on one through six seeds only, since they're the teams most likely to make deep runsóand I made the stipulation that there had to be at least 50 appearances for each combo. (I'll explain the last two entries in the chart below in a minute). I'm not claiming that these are absolutely the best two-stat combos available in the Bracketmaster (maybe someone can find better ones), but all of these are well worth keeping in mind when you fill out your bracket: The most powerful two-stat combo I found involved scoring margin and strength of schedule. The 51 teams that have had a healthy scoring margin of 14.5 points or more while playing a tough but not horrendous schedule (ranked three through 53) have racked up a +.737 PASE. That means they've collectively overachieved at the rate of about three-quarters of a game per tourney. Put it this way: these teams own a better PASE than Billy Donovan. But PASE is just one measure of how well teams with these two stats have performed. Check out the "F4%" and "CH%" columns. More than half of these squads have reached the Final Four and about 18 percent of them have cut down the nets. The top two-stat combo for "Final Four advancement rate" involves star power and scoring margin. The 51 teams with All-Americans and a high scoring margin of at least 16 points a game have gotten to the Final Four nearly 53 percent of the time. And their success rate in winning the tourney is nearly one in five. That's also the best "champion success rate" of any of the top ten two-stat combos. It's not better than the rate of the last two combos in the chart, however. These include Pythag ratings as one of the attributes. The problem is that I only have six years of data for possession-based stats and these two combos didn't make the required 50 appearances. So proceed with caution. That said, the PASE values and advancement rates to the Final Four and championship are too impressive to ignore. Teams ranked in the top ten for Ken Pomeroy's Pythag calculation that also have solid scoring margins of at least 14 points per game own a gigantic PASE of +.827, get to the Final Four half the time and cut down the nets at a better rate than 20 percent. The numbers are pretty similar for top-ten Pythag teams that score at least 77 points a game. Bottom line: if you want to get a quick read on teams that are likely to perform well in the 2010 dance, keep these two-stat combos in mind.
0 Comments...


Wed, Feb 24, 2010, 10:55 PM	By ptiernan	0 Comments...
Based on Monday's AP Top 25, five teams have the six basic credentials that 17 of the last 19 champions have possessed: 1) They came from a Power conference (ACC, Big East, Big Ten, Big 12, SEC or Pac 10) 2) They went to the previous year's tourney 3) Their coach had at least five tourney appearances 4) They scored at least 77 points a game 5) Their average margin of victory was 10 points or more 6) They were seeded one through four The five squads are the same as they were last week: Kansas, Syracuse, Duke, Villanova and Texas. That said, the Longhorns AP ranking of 21 puts them in serious jeopardy of failing to bag a one-through-four seed. So that would leave us with a Formidable Four of the Jayhawks, Orange, Blue Devils and Wildcats. The only one of these teams that's at risk of failing to measure up to the six champ stats by Selection Sunday is Villanova. The Wildcats' average scoring margin is just 11.4óand that number keeps sliding. If they fall below 10 points, they're out. When you compare the AP Top 25 to the rankings based on tempo-free efficiency stats, there are nine ranked teams that don't have the efficiency numbers to measure up: Richmond (#50 according to Pythag), Gonzaga (#46), New Mexico (#42), Northern Iowa (#40), Texas A&M (#34), Temple (#32), Butler (#28), Tennessee (#27) and Vanderbilt (#26). If possession-based data ruled the college ratings, these nine teams would replace the imposters above: Maryland (#12, see the stretch run blog for more on the Terps), Missouri (#13), Florida State (#16), Xavier (#17), Clemson (#18), Georgia Tech (#21), Marquette (#22), California (#23) and Virginia Tech (#25). Who's the best team in the nation according to possession-based efficiency numbers? Surprise, surprise, it's Duke, barely nudging out Kansas. We'll see if that's a harbinger of overachievement for the Dukies in the dance. If so, it will break a string of five underachieving dances.
0 Comments...


Sat, Feb 20, 2010, 10:39 AM	By ptiernan	6 Comments...
With three weeks until Selection Sunday, I took a look at all the AP Top 25 teams and any team off that list that has a top 25 possession-based Pythag ranking. Then I slotted them into six categoriesófive favorites, five challengers, five contenders, five pretenders, five sleepers and the nine remaining teams, which I call omissions. This is a dynamic list. After today's results, I'll no doubt have a different opinion. But at this moment, with 22 days, six hours, 37 minutes and 12 seconds until Selection Sunday, this is where my head's at. I've reviewed 34 teams through the infamous "eye test," as well as 12 key stats that foretell tourney success or failure. In the chart below, a red highlight means that a team fails to meet a basic criteria for overachievement in the dance. The last column, "DQ," tallies up all the disqualifiers. Here's what each column means and the requirement: • 09? - Did they go to last year's dance? • CY - How many tourney trips has the coach made? (Should be three or more) • E8 - How many times has the coach been to the Elite Eight? (At least once) • CO - What's their conference affiliation? (Should be "B" for Big Six) • AP# - What's their AP rating? (The top 15 avoid a red disqualifier) • SOS - What's their strength of schedule (Anything worse than 40 is bad) • PYT - What's their possession-based Pythag ranking (Should be top 15) • OE - Where do they rate for offensive efficiency? (Must be 25 or better) • DE - Where do they rate for defensive efficiency? (Must be 25 or better) • PPG - What's their point scoring average (Should be 75 points per game or more) • MAR - What's their average scoring margin (Needs to be better than 10 points) • MO - How's their current momentum (I gave any team with three losses in the last 10 a minus) • DQ - A tally of all the red highlights for each team And here's the chart, with team write-ups to follow: FAVORITES Kansas - The Jayhawks are one of only two teams that possess all 12 of the key stats. I've been saying all year that I didn't think they could follow through on their status as prohibitive favorite. I'm starting to waver on that assertion. Aldrich and Collins form a great inside/out combo, the Morris twins are playing much more under control of late, the team is deep, and Henry's the X factor. Unless they stumble badly down the stretch, they'll make the Final Four in a lot of my models and be pegged as eventual champ in a fair share. Syracuse - The only chink in the Orange's armor is a relative weak SOS for a team making a deep tourney run. Everything else looks goodósolid coach, Power conference, ten 15 Pythag, strong offensive and defensive efficiency rates, high scoring, big margin and decent momentum. Let's see if the Cuse can bring it home down the stretch. (For some reason, that exhibition loss to LeMoine still sticks in my head, though.) Duke - The Blue Devils are the only other team that meet all 12 criteria for a deep run. Duke's efficiency numbers are betting than its AP rankingóand the Coach K's squad is the most offensively efficient team in the nation. What separates this year's team from the five that have underachieved before them is a strong frontcourt. Duke has the big men to contend in the later rounds of the dance. The question is: can they're backcourt hold up in the early rounds and fend off upset bids? Kentucky - The Wildcat's big bugaboo is that they didn't go to last year's danceóand only two teams in the 25-year, 64-team era have cut down the nets after failing to make the previous tourney (Louisville in 1986 and Syracuse in 2003). I've argued that Kentucky has the intangibles to shed that particular monkey on its back. But the soft SOS is also a concern. That said, this team seems to find a way to pull out close gamesówhether against weaker competition or strong teams. Villanova - All the numbers would seem to align for Novaóexcept one. They are not a very defensively efficient team. A ranking of 68 is extremely low for a prospective Final Four combatant. Just how low? Consider this: the worst defensive semi-finalist of the last six tourneys was North Carolina last year. And their defensive efficiency ranking was 25ófar better than the Wildcats. So why do I still believe in Villanova. Two words: Scottie Reynolds. CHALLENGERS Purdue - I debated putting the Boilermakers or Buckeyes into the "Favorite" category and dropping Villanova, but then I watched the two teams play each other and couldn't pull the trigger. I was impressed with Purdue's ability to bag a big road win, but they struggled down the stretch. Still, I've liked the Boilermakers all year long, going back to my early December 28 review of the nation's top teams. I worry that Matt Painter is a snakebitten coach (will have five tourney appearances with no Elite Eight runs) and the offensive efficiency and firepower are concerns. But Purdue has a strong frontcourt, a solid guard and good, old-fashioned toughness. They're a Final Four challenger for sure. Ohio State - Ohio State is reasonably efficient offensively, but they're not as prolific as a typical Final Four candidate. Plus, the SOS is a little cream-puffy. On the plus side, the Buckeyes boast one of the nation's very best players in Evan Turneróand star power matters. Plus, they have good balance in their efficiency numbers and a hefty scoring margin. I wouldn't be surprised at all to see them in the final weekend. I would be surprised if they cut down the nets. West Virginia - Three things concern me about the Mountaineers. Four really. But we'll get to that. As we've already mentioned, Final Four-caliber teams are more efficient defensively than a 27 ranking suggests. And while West Virginia is the third most offensively efficient team in the nation, they put up fewer than 75 points a game. Plus, the Mountaineers are struggling of late. If they don't right the ship, I can't see them getting to the Final Four. And that brings me to the fourth reason I'm leery about West Virginia: Bob Huggins. His record of underachievement in the dance doesn't exactly inspire confidence. All that said, the talent is definitely there for the Mountaineers to do well. If they go into the dance on a roll, I'll feel differently. And that's why they're listed as challengers. Michigan State - The Spartans have the most flaws of any team I've pegged as a favorite or challenger. Their Pythag rating is outside the top 15, they're surprising weak on defense, they don't score enough, their margin is in the single digits and they've been limping down the stretch. So why didn't I drop MSU down to contender status? The reason is purely, Tom Izzo. This guy just knows how to get the most out of his team come tourney time, and he owns the best PASE of any active coach (+.864). But Izzo can't play center for this team, and they sorely miss a post presence like they had last year in Goran Suton. For the Spartans to reach the final weekend again, Lucas will have to come back 100%, Luscious has to play better, and MSU's wingsóSummers, Roe and Morganówill have to compensate for the hole in the middle. Lots of ifs, Texas - I know, I know. What are the Longhorns doing in the second echelon? I was pretty high on Texas early, and I'm not ready to give up on them, but Barnes' bunch have issues. It starts with getting good offense when they really need it. The Longhorns' offensive efficiency ranking of 26 is far too low for a Final Four contender, their SOS is too soft, and, of course, they are struggling big time. I have them hanging on to the bottom rung of the challenger tier only because I still think they've got the talent, the balance and the depth to make noise in the dance. But they better turn things around soon. CONTENDERS Kansas State - The only statistical weaknesses that the Wildcats have relate to experience. They didn't go to last year's tourney, and their coach, Frank Martin, is greener than your average deep dancer. Aside from that, KSU is strong: highly ranked in the AP, tough schedule, solid efficiency numbers, high scoring, comfortable margins and decent momentum. Consider this: experience aside, only two other teams can boast qualifying numbers for those nine statsóin-state rival Kansas and Duke. That's good company. So why didn't I bump them up to a challenger? That lack of experience is a big thing for me. Martin will have to prove himself before I push his squad too far into my bracket. Wisconsin - There's a lot to like about the Badgers. They own the third best efficiency rating the in the country, despite a ranking of 14 in the AP Top 25. And their strength is based on balance, as both their offensive and defensive numbers are solid. That said, Wisconsin is still a slow-down team that averages just 68 points a game. That usually doesn't bode well for a deep tourney run. Nor does sputtering momentum, and Wisconsin is struggling down the stretch. There's time to rectify that. The low offensive output is another matter. Sweet 16? Sure. Eight Elite? Mmm, maybe. I can't a Final Four run though. Georgetown - There's a lot to like about Georgetownódominant big man, stabilizing point guard, recent history of tourney success. But there are also some nettlesome flaws that have me scratching my head. Like, what's with the losses to Rutgers and South Florida? More importantly, when you didn't go to the previous dance, can only muster 73.8 points a game, don't have an average margin above 10 points, and rank 36th in defensive efficiency, you're odds of a Final Four run are steep. I can see the Hoyas in the Elite Eightóor getting shocked in the opening round. Crazy team. Pittsburgh - User "Spike" noticed that I inexplicably forgot to add the write-up on Pitt. That's what happens when you try to write on a cramped airplane on the way to Montreal (to see Clapton and Beck no less!) So I'm writing this three days--and one big win over Villanova--after the rest of the write-up. I still like where I put Pitt. They aren't quite challengers in my book, but certainly contenders. Dixon shed the stigma of being a snakebitten "no Elite Eight run" coach last year. And the Panthers tough schedule and decent possession-based numbers (better now than when I posted the chart above) bode well for a surprise tourney run. Let's face it: nearly every year, a middle-of-the-pack Big East squad overachieves in the dance. After years of disappointment, this could be Pitt's year. Vanderbilt - It was between Vandy and Tennessee for the last slot in the contenders tier. I chose Vandy, mainly because they're playing better right now. There are some sleepers that I'd pick for deeper runsóbut, as I mentioned in the intro, I've any surprise teams that aren't ranked automatically fall into the sleeper field. Truth be told, Baylor, Missouri, and Maryland are all solid choices to go further than the Commodores. Still, I like this team. I like their size, their balance and their ability to put up points. I think their defensive inefficiency is going to doom them. And Kevin Stallings is another one of those snakebitten coaches. PRETENDERS Tennessee - Speaking of snakebitten coaches, Bruce Pearl falls into that ignominious club as well. He's been to the dance six times without reaching the Elite Eight. That's one reason why I have the Vols on the cusp of the Contender/Pretender categories. But it's not the biggest. I just find this team to be bewilderingly inconsistent. Some nights they look fantastic, and others they look like they aren't even awake. As far as the numbers go, their schedule is pretty weak and their offensive efficiency ranking (#76) is gruesome. Plus, they're starting to sputter down the stretch. Here's one reason for optimism, though: unlike years past, this Vol squad is not a defensive sieve. They're the sixteenth most efficient team in the nation at stopping opponents. If Tennessee is going to make a deep tourney run, it's going to be because of defense. BYU - With a Pythag rating of seven and only four statistical flaws, Brigham Young might be the team least deserving to be tabbed a pretender. But I base my decision on a few key facts: 1) the Cougars haven't won a tourney game in six straight dances, 2) they own one of the worst historical PASE values of any team since 1985, and 3) they have the cushiest schedule of any team I examined here. Sure, they're efficient, they score a ton and their average margin is second only to Kansas. But prove it to me, Cougars. Prove it to me. New Mexico - Okay, so they have a strong record and they're the twelfth ranked team in the AP poll. I'm not soldóand seven disqualifiers tell the story. The Lobos rank just 30th in Pythag, with mediocre offensive (#45) and defensive (#33) efficiency. These are not Elite Eight numbers. Add to that the fact that they didn't go to last year's dance, Steve Alford is on the verge of ësnakebittenness" and their SOS is super gooshy, and you have the makings of a tourney disappointment. Gonzaga - The Bulldogs used to be a fashionable Cinderella; but once they started grabbing higher seeds, they quickly became Cinderellees, victims of upsets and pre-mature departures. I don't see this year as being any different. Mark Few is the poster child for snakebitten coaches, having gone to the dance coming on 11 times without an Elite Eight run. Besides that, the Zags play a soft schedule and their efficiency numbers are middling. I'll give them this though: they can fill it up. Potent offenses are sometimes enough to win a couple rounds. The Sweet 16, that sounds about rightómaybe a little optimistic. Butler - The Bulldogs East were anointed The Great Mid Hope before the season began. They're having a decent year, but they aren't exactly blowing way their milktoast competition (SOS of 116). Only one other team I reviewed this week has more statistical danger signs than Butleróand that's why Richmond fell to the "Omissions" group. Not only that, but the Bulldogs struggle to score and are offensively inefficient. As Wisconsin shows, the two don't necessarily go hand in hand. Unless their road to the Sweet 16 includes other flawed squads, I can't see Butler springing any substantive surprises this year, despite the pre-season hype. SLEEPERS Maryland - Only five teams have zero or one statistical flaws. Four of themóKansas, Duke, Syracuse and Villanovaóare ranked as favorites. The other team is, amazingly, Maryland. The only thing the Terrapins lack is that they aren't ranked in the AP Top 15 (or the top 25 for that matter). Maybe the voters just don't know what they're talking about, because everything else checks out. Successful veteran tourney coach, check. Tourney-tested squad, check. Right conference affiliation, yep. Decent SOS, okay. Solid efficiency numbers (11 Pythag, 20 offensive efficiency, 18 defensive efficiency), oh yeah. Prolific offensive and healthy average margin, check. Ohóand then there's Greivous Vasquez, a legit go-to guy. If they weren't unranked, I'd have them in the contenders category for sure. Baylor - By the time the dance rolls around, Baylor is going to be the chic sleeper pick among the media. Just watch. What's not to like. Okay, so the Bears are ranked just 22nd in the AP. But their Pythag efficiency rating is strong at 14 and they're the seventh most offensively efficient team in the country. Right now, they've hit a bit of a rough patch, their defensive numbers aren't what you'd expect of a deep dancer, and coach Drew is a relative tourney newbie. But I like the make-up of this team. Just watch, Missouri - After the Tigers' Elite Eight run last year, tabbing them as a sleeper isn't much of a shock. But after losing so much in the frontcourt, I didn't think Mizzou would get here at the beginning of the season. It's a testament to coach Mike Anderson and his frenetic style of play. The Tigers have the thirteenth best efficiency numbers in the country and their defense is particularly good (#12). The two chinks in their armor are a so-so schedule and middling offensive efficiency numbers. That said, their up-tempo style of play helps them pile up the points. You don't see many unranked Big Six conference teams averaging nearly 80 points per game with an average margin near 15. Wake Forest - Here's another ACC team to keep an eye on. The Demon Deacons play a tough schedule (eighth hardest in the country) and manage to put up respectable efficiency numbers. They're better defensively (23) than offensively (57), but they still manage to score more than 75 points a game. Dino Gaudio will only be making his second tourney appearance, and Wake Forest is historically an underachiever. But the Deacons seem to struggle most when they have a high seed. That won't be the case this time. They could sneak up on a couple teams. Virginia Tech - I don't know why I picked Virginia Tech as the last sleeper in over Xavier and Marquette. Call it the eye test. Call it "buying in to Seth Greenberg's years of lobbying. I guess what sold me most were three things: 1) Greenberg will be making his fourth tourney trip, 2) Virginia Tech is on a roll, and 3) they have the fifth stingiest defense in the nation. For all those reasons, I'll make Va Tech my last sleeperóat least for now. OMISSIONS - I could be persuaded that a number of these teams could climb into the sleeper category--Xavier, Marquette, Cal, Florida State and Georgia Tech are my favorites here. Virginia Tech is a shaky sleeper for me. If I did this article in another week, I'm sure there would be several changes.
6 Comments...


Thu, Feb 18, 2010, 11:08 AM	By ptiernan	3 Comments...
If you separate the teams at every seed position that make the Final Four from those that don't, you'll find that the winners tend to share attributes that the losers don't possess. Consider the top seeds. Beware of a one seed that has any one of these 11 attributes: 1. They didn't go to the previous dance. (Hel-lo Kentucky...) 2. They've lost two or more games in a row. 3. Their coach has been to the dance five or more times without reaching the Elite Eight. 4. They have a winning percentage lower than .780. 5. They have a "strength of schedule" (SOS) rating weaker than 80. (Syracuse's SOS is currently ranked 94th) 6. They score 72 or fewer points per game. 7. They allow 80 or more points per game. 8. Their scoring margin is 10.5 points per game or less. 9. They get imbalanced scoring, relying on either the backcourt or frontcourt for 73% or more of their points. 10. They get outrebounded. 11. They shoot less than .465 from the floor. Top seeds possessing any of these attributes are just four for 37 in reaching the Final Four; that's a 10.8% success rate. The four exceptions are Oklahoma in 1988 (gave up 81.4 points per game), Minnesota in 1997 (didn't go to the dance in 1996), Texas in 2003 (Barnes had been to the dance 11 times without reaching the Elite Eight) and UConn last year (two-game losing streak). Meanwhile, top seeds that aren't hampered by any of these weaknesses are 40 for 63 in reaching the Final Fouróa 63.5% success rate. That's a 44.3% improvement over the typical success rate of top-seeded semi-finalists (44 for 1000. Just as importantly, these exclusion rules identify 40 of the 44 actual contenders, for an accuracy rate of more than 90%. This is just a taste of the information you'll find in the Final Four/Champ Strategy feature article. Make sure to check it out come bracket-pondering time.
3 Comments...


Mon, Feb 15, 2010, 9:50 PM	By ptiernan	0 Comments...
Based on today's AP Top 25, only five teams now have the six basic credentials that 17 of the last 19 champions have possessed: 1) They came from a Power conference (ACC, Big East, Big Ten, Big 12, SEC or Pac 10) 2) They went to the previous year's tourney 3) Their coach had at least five tourney appearances 4) They scored at least 77 points a game 5) Their average margin of victory was 10 points or more 6) They were seeded one through four Tennessee dropped off the list after a week that saw its scoring average dip below 77 points a game. Check out today's results: All the teams with "Y's" highlighted in orange under "Old Rule" have the six attributes cited above. That includes: Kansas, Villanova, Syracuse, Duke, and Texas. If Tennessee increases their scoring average, they could climb back onto the potential champ list. But I don't see any other squads joining this elite group. Could any of the current five teams slide out of stats champ contention? Villanova and Texas are probably the most at risk. The Wildcats' average scoring margin is just 12.0 points per game after their upset loss at the hands of UConn. And if the Longhorns keep struggling, they're at risk of failing to bag a top four seed. That means the three most solid prospective champions are Kansas, Syracuse and Duke. For the last three weeks, I've been tracking a new stat in the "champ credentials" chart. See that second column called "PY RK"? That's how the AP's top 25 teams rank according to Ken Pomeroy's Pythagorean possession-based efficiency statistics. According to efficiency statistics, the nation's top five Division I basketball teams are Kansas, Duke, Wisconsin, West Virginia and Syracuse. Kentucky is just the eleventh most efficient team in the nation. If Pomeroy's Pythag data determined the rankings, ten teams in today's AP Top 25 would be on the outside looking in. Pittsburgh (#28), Wake Forest (#29), Butler (#30), Vanderbilt (#31), Tennessee (#32), New Mexico (#33), Texas A&M (#36), Temple (#37), Gonzaga (#40) and Richmond (#45) all don't have good enough efficiency numbers to merit a high rating. Who would replace these ten squads in a tempo-free world? Maryland (#13), Missouri (#14), California (#16), Marquette (#18), Xavier (#19), Clemson (#20), Old Dominion (#21), Florida State (#22), Virginia Tech (#24) and Utah State (#25) deserve a second look, especially if these rankings hold until Selection Sunday.
0 Comments...


Sat, Feb 13, 2010, 9:26 AM	By ptiernan	10 Comments...
If Jay Bilas has said it once, he's said it a thousand times: "There are a lot of good teams in college basketball this year; there just aren't any great ones." I would tend to agree with him, based on the eye test, but do the numbers bear out his assertion? The best statistic for evaluating the relative quality of teams from one year to the next is "Pythag," a possession-based calculation of a team's combined offensive and defensive efficiency developed and made popular by Ken Pomeroy, John Gasaway and Dean Oliver. How does the average Pythag of this year's top ten most efficient teams compare to that of the top ten teams heading into each of the last six tourneys? In short, not so good. Take a gander at the left chart below: This year's top ten Pythag teams are tied with 2009 for the weakest year of overall team efficiency. And here's the thing: that number is bound to fall as the season winds to a conclusion and these teams complete their tough conference schedules and tourneys. So it's probable that this year's higher-seeded teams will be less dominant overall than they've been in seven years (that's as far back as my pre-tourney possession-based data goes). So what does that matter? Well, there is somewhat of a correlation between elite team strength and tourney predictability. See the ghosted orange bars overlaying the red Pythag values above? Those are measures of tourney predictability based on the readings I explained in the Madometer feature article. The most predictable tournament in the 25-year history of the 64-team era was 2007. If you measure the deviation from perfect higher-seed advancement (read that Madometer piece if this concept is murky), it was 95.9% predictableóeasily more sane than the next calmest tourney (2009: 90.8% predictable). And, lo and behold, 2007 was also the year when the top ten teams had the strongest collective Pythag rating. And in 2006, the most unpredictable tourney since 1999 (remember George Mason), the top ten squads had one of the weaker average Pythags. Of course, the correlation isn't super strong. As mentioned, last year was the second most predictable tourney of the modern era, and the top teams owned the weakest Pythag of the last seven years. That said, my guess is that this year's top teams will enter the 2010 tourney with a significantly lower average Pythag than in year's past, and the tourney will be more unpredictable and upset-prone because of it. This doesn't exactly resolve the issue I brought up at the beginning of this blog, however. Just because this year's top ten teams have a low collective Pythag, that doesn't mean there aren't one or two great teams out there. In fact, if you chart the individual Pythags of the top teams as of this morning, you see that Kansas and Duke are head and shoulders above the rest of the top ten. Then there's a grouping of three contendersóSyracuse, West Virginia and Wisconsinóthat are a notch below the top two, followed by a bunched-up group of five squads rounding out the top ten. So Kansas and Duke are, at this point, the most efficient teams of 2010. But that doesn't mean they're necessarily "great." If you rank them against the 70 top ten teams since 2004, Kansas is just the seventh best Pythag squad of the era and Duke is seventeenth best. The top Pythag team since 2004, ironically is the 2008 Kansas squad that cut down the nets. They're followed by the 2007 and the 2005 Tar Heels (another tourney champ). In fact, all of the tourney champs since 2004 have been among their year's top ten Pythag squads (file that little nugget away for bracket-brooding time), though they're not necessarily highly ranked among the 70 squads studied. The Gator team that won in 2006 was just the sixtieth best Pythag team of the 70 top ten most efficient teams. Last year's Tar Heels, surprisingly, were just the forty-eighth best Pythag team on this list. For what it's worth, if the current Pythag values held (and they'll almost certainly go down), the rest of this year's top ten would be ranked as such among the 70 top ten Pythag teams since 2004: Syracuse - 44 West Virginia - 55 Wisconsin - 58 Purdue - 64 BYU - 65 Kentucky - 66 Kansas State - 67 Maryland - 68 Them's some pretty low rankingsóthat are bound to go lower. And that's important to bear in mind. Why? Because, not including 2010, the bottom ten lowest Pythag teams on the list of 70 have beaten seed expectations just four times. Meanwhile the top ten Pythag teams overachieved seven times. Here are the five things I take away from all of this: 1. Jay Bilas is right. There aren't any great teams this year. 2. That's likely to make the tourney more unpredictable. 3. This year's tourney champ will likely have a top-ten Pythag rating (That doesn't bode well for Villanova, Michigan State, Tennessee, Georgetown, etc.) 4. The most likely squads to make early exits are the ones whose Pythag ranks in the bottom ten of the 70 top-ten squads since 2004. 5. I'm going to be paying extra-special attention to Pythag when I fill out my bracket this year.
10 Comments...


Mon, Feb 8, 2010, 10:18 PM	By ptiernan	0 Comments...
Based on today's AP Top 25, six teams have the six basic credentials that 17 of the last 19 champions have possessed: 1) They came from a Power conference (ACC, Big East, Big Ten, Big 12, SEC or Pac 10) 2) They went to the previous year's tourney 3) Their coach had at least five tourney appearances 4) They scored at least 77 points a game 5) Their average margin of victory was 10 points or more 6) They were seeded No. 1 through No. 4 The same six teams as last week met all these criteria. Check out today's results: All the teams with "Y's" highlighted in orange under "Old Rule" have the six attributes cited above. That includes: Kansas, Syracuse, Villanova, Duke, Tennessee and Texas. This year, we're tracking a seventh champ criterion. Sixteen of 19 champions have possessed all the qualities listed aboveóand were led by a coach who had been to the Elite Eight in a previous year. Adding this requirement eliminates Tennessee. This leaves just five teams that have the numbers to cut down the nets in 2010: Kansas, Syracuse, Villanova, Duke and Texas. Tennessee is also in danger of failing to meet the six basic stat champ criteria. The Vols average exactly 77 points a gameóright down to the point. Should they fail to average 77 points this week, you won't see them on either the old or new list next week. For the last three weeks, I've been tracking a new stat in the "champ credentials" chart. See that second column called "PY RK"? That's how the AP's top 25 teams rank according to Ken Pomeroy's Pythagorean possession-based efficiency statistics. According to efficiency statistics, the nation's top five Division I basketball teams are Kansas, Duke, Syracuse, Wisconsin (whoa!) and West Virginia. Villanova? They're just the fourteenth best team by possession-based stats. (Of course, the Cats are liable to move up after tonight's thumping of the Mountaineers.) Actually, if Pomeroy's Pythag data ruled the rankings, eight teams in today's AP Top 25 would be on the outside looking in. Butler (#26 in Pythag), Vanderbilt (#27), UNLV (#32), Pittsburgh (#33), Gonzaga (#42), Temple (#47) and Northern Iowa (#48) all don't have good enough efficiency numbers to merit a high rating. Who would replace these eight squads in a tempo-free world? Maryland (the eighth most efficient team in the nation!), Missouri (#13), California (#17), Marquette (#18), Clemson (#22), Florida State (#23), Utah State (#24) and Louisville (#25) are better basketball teams than the pundit think. If these numbers hold up until March, keep your eye on these squads. They're good candidates to spring a tourney surprise. For the first time all season, the top ranked team in the AP is also the best team in terms of playing efficiency. And Syracuse is both the third-ranked squad and the third-best Pythag team. That's pretty much where the similarities end, though. Villanova is the second best team according to the AP, but its efficiency numbers rank a surprisingly low thirteenth. Fourth-ranked Kentucky is the ninth best Pythag squad. And #5 Michigan State has just the nineteenth-best possession-based statistics. Meanwhile, Duke (2), Brigham Young (4) and Wisconsin (5) join Kansas and Syracuse as the top five Pythag teams. Several ranked teams in the AP poll wouldn't even be in the top 25 if the list was based on efficiency numbers. Gonzaga, Mississippi, Northern Iowa, New Mexico, Pittsburgh, Temple and Butler would all tumble from the Top 25. They would be replaced by Maryland (11), Missouri (15), Marquette (17), Xavier (20), California (21), Clemson (22) and Florida State (24).
0 Comments...


Mon, Feb 8, 2010, 9:33 PM	By ptiernan	3 Comments...
At one point or another, every college basketball announcer has annoyed me (I'm easily annoyable). Bilas whines too much about the refs. Knight belabors his point ad nauseum. Musberger lays the homespun schtick on too heavily. And, of course, Vitale is just maddeningly Vitale. But of all the national announcers out there, nobody grates on me more than Bill Raftery. I'm tired of "The Kiss," I rarely find a basketball play "gorgeous," and I don't agree that every center "shuffles the big puppies." It got to the point where I had to turn off the volume to watch a Raftery-called game. But, wonder of wonders, my son actually likes Bill Raftery. So I've grudgingly agreed to put up with him. And last week, I found the perfect coping mechanism. Now, whenever Bill lets loose one of his incomprehensible blatherings, I write it down. It's gotten to the point where I can't wait to hear what sort of idiotic nonsense will come out of his mouth. Here are the top five Rafteridiculous comments of the last week or so: 5. Pretty move! Oohh, the ability to sliver with that broad shoulder. Gorgeous! (This isn't the only quote in the top five that sounds suspiciously...well...you know. Not that there's anything wrong with it.) 4. You gotta go quick and strong and, of course, he may have gotten a little knick-knack there. He might've gotten a piece of the cuticle. 3. Bob Huggins said you may love him some nights. Others, he might not ring the bell. Tonight, Huggie's gonna caress him! (Well...hmmm...) 2. He's got arms! He should take the collection up at church. Goodness! He can reach deep into a pew! And the number one, Rafteridiculous comment: 1. Wow! Is he tough! He just knows and knows and knows when to lay the wood on the guy! I'm at a loss.
3 Comments...


Thu, Feb 4, 2010, 10:27 PM	By ptiernan	4 Comments...
Last week, I gave a quick, basic tutorial on how to use the Bracketmaster. I explained how to answer easy questions like "How has Tom Izzo done in the tourney when his Spartans have been seeded lower than third?" (Answer: 12-8, with an overachieving PASE of +.515.) And I also walked you through more complicated searches like: "How do ACC teams perform that are making more than their third straight trip, score 75 or more points, don't have an All-American and have a strength of schedule tougher than fiftieth? (Answer: 28-16 with a PASE of +.115óbut no championships). I use the Bracketmaster all the time to get quick answers like that. But there's a more valuable way to use the tool. When I go to fill out my bracket this year, I'll be asking myself, "Which kinds of teams are most likely to get to the Final Four?" The Bracketmaster can answer this question too. Let's say that you want to construct a rule that would identify at least 75% of the Final Four teams with maximum reliability. The first factor I would consider is seeding. While 100 of 1600 teams in the tourney fieldójust 6.2%óhave reached the Final Four since 1985, the overwhelming majority of them have been seeded one through six. If you restrict your choices to these high seeds, you increase your odds of picking a Final Four team to 15.8%...and the rule still encompasses 95 of the Final Four teams. You can do even better than that if you eliminate teams that are struggling heading into the dance. Add these two requirements to the Bracketmaster: 1) focus on teams that have won at least six of their last ten pre-tourney games, and 2) make sure they don't have more than a two-game losing streak. All 95 of the Final Four teams found in the previous search are also in this smaller set of teamsóbut we've eliminated 62 more pretenders. So now our conditions include 538 candidates while describing 95 Final Four squads. That's an ARF (Ability to Reach the Final Four) rate of 17.7% We're getting betteróbut we've only just begun. Knowing that teams making deep tourney runs tend to be higher scoring, we can play with the "Team PPG" dial. After a few searches, we discover that 81 of our remaining Final Four teams have scored at least 75 points per game. They also win by a margin of 7.5 points per game or more. With these two factors added to our conditions, we eliminate 199 more pretenders. So now, we have a rule that includes 81 semi-finalists, but is satisfied by only 339 teams. That's an ARF of 23.9%. Let's push that up even higher, while still describing 75% of the Final Four contestants since 1985. Restrict your selection to teams that also have a winning record higher than .715, give up no more than 81.5 points a gameóand are just seeded one through four. (Yeah, we went back to the seeding variable; that's the kind of "what-if" tweaking you have to do in these sorts of analyses.) Only 270 teams satisfy all these conditions in your new ruleóbut 75 of them reached the Final Four. That works out to an ARF rate of 27.8%. I just did this analysis on the fly, so it's likely that there are several combinations of statistics that will yield a better ARF while retaining 75% of the semi-finalists. (In fact, I just came up with one that included 256 teams while retaining 75 Final Four teams, all while watching the new episode of the "The Office.") I didn't use some of the variables like "# of straight bids," "# coach's bids" or "All-Americans"ójust because I got working on some of the other stats. Then there were a few stats I didn't use because I don't have 25 full years of data. It wasn't until 1994 that "Strength of Schedule" started getting tracked. And I only have "# Jr/Sr Starters" and "Guard % of Points" information for 23 years. Bear that in mind as you do your advanced Bracketmaster research. So, here's a question you can test your new Bracktetmaster knowledge on: What set of conditions describe the smallest set of teamsówhile retaining 20 of the 25 past champions?
4 Comments...


Mon, Feb 1, 2010, 8:19 PM	By ptiernan	0 Comments...
Based on today's AP Top 25, only six teams now have the six basic credentials that 17 of the last 19 champions have possessed: 1) They came from a Power conference (ACC, Big East, Big Ten, Big 12, SEC or Pac 10) 2) They went to the previous year's tourney 3) Their coach had at least five tourney appearances 4) They scored at least 77 points a game 5) Their average margin of victory was 10 points or more 6) They were seeded No. 1 through No. 4 Last week, seven teams met the criteria above, but Michigan State fell off the champ list, as predicted, because the Spartan's offense is sputtering in the grind-it-out Big Ten. Check out this week's results: All the teams with "Y's" highlighted in blue under "Old Rule" have the six attributes cited above. That includes: Kansas, Villanova, Syracuse, Texas, Duke and Tennessee. This year, we're tracking a seventh champ criterion. Sixteen of 19 champions have possessed all the qualities listed aboveóand were led by a coach who had been to the Elite Eight in a previous year. Adding this requirement eliminates Tennessee. This leaves us with the just five teams that have the numbers to cut down the nets in 2010: Kansas, Villanova, Syracuse, Texas and Duke. Last week, I debuted a new stat in the "champ credentials" chart. See that second column called "PY RK"? That's how the AP's top 25 teams rank according to Ken Pomeroy's Pythagorean possession-based efficiency statistics. For the first time all season, the top ranked team in the AP is also the most efficient team in the country. And Syracuse is both the third-ranked squad and the third-best Pythag team. That's pretty much where the similarities end, though. Villanova is the second best team according to the AP, but its efficiency numbers rank a surprisingly low thirteenth. Fourth-ranked Kentucky is the ninth best Pythag squad. And #5 Michigan State has just the nineteenth-best possession-based statistics. Meanwhile, Duke (2), Brigham Young (4) and Wisconsin (5) join Kansas and Syracuse as top five Pythag teams. Several ranked teams in the AP poll wouldn't even be in the top 25 if the list was based on efficiency numbers. Gonzaga, Mississippi, Northern Iowa, New Mexico, Pittsburgh, Temple and Butler would all tumble from the Top 25. They would be replaced by Maryland (11), Missouri (15), Marquette (17), Xavier (20), California (21), Clemson (22) and Florida State (24).
0 Comments...


Sun, Jan 31, 2010, 2:51 PM	By ptiernan	0 Comments...
Long-time member Skillet300 emailed me a question this morning that gave me a great blog topic--and has occupied most of my day. He wondered whether a team's road/neutral record was a reliable indicator of performance in the tourney. Put another way, Skillet300 wanted to know whether teams with better road records tended to overachieve in the dance, while those with weaker ones underperformed. I only have five years of home/neutral records in my tourney database (I plan to go back to 2000 in the off-season), but I figured it would be enough to draw a tentative conclusion. Here's what I did: for the top six seed positions, I compared the overall and road/neutral winning percentages of overachieving and underachieving teams. On the surface, it seems clear-cut that overachievers in the dance have better road records than underachievers. One through six seeds that beat expectations in the dance own a .681 road winning percentage; those that fall short of expectations have a road winning percentage of .652. But the answer isn't that easy. That fact is, high-seeded overachievers also have a better overall winning percentage than underachievers--.799 to .788. So you can't just focus on road winning percentage by itself to get to the bottom of whether it's a good indicator of tourney success. What you have to do is examine the gap between overall and road winning percentages for overachievers and underachievers. If road success were truly an indicator of overachievement in the dance, the gap for overachievers between overall and road winning percentage would be smaller than it was for underachievers. Let's re-examine the numbers we already presented. One through six seed overachievers have an overall winning rate of .799 and a road winning rate of .681. That's a .118 gap. Meanwhile, the difference between overall winning percentage (.788) and road rate (.652) for underachievers is .136. So there's a greater discrepency in overall and road rates for underachievers than overachievers. And that's what you'd expect if road success were an indicator of overperformance in the dance. But the gap isn't exactly huge--just .018--so it's hard to draw any definitive conclusions about the value of road success as an indicator of tourney success. Maybe if we examined the gap between overall and road winning rates for over- and underachievers at each seed position. Take a look at this chart: The blue bars represent the overall and road winning rates of overachievers; the orange bars represent the two rates of underachievers. And the small red bars represent, for each of the six seeds, how much bigger (or smaller) the gap is between overall and road winning rates for underachievers and overachievers. A positive gap--like what you see for one (.033), two (.035), three (.012) and four seeds (.093)--means that there's a bigger difference between overall and road success for underachievers than overachievers. And that squares with the theory that road/neutral game performance is an indicator of tourney overachievement. Of course, for five (-.042) and six seeds (-.010), the gap is actually smaller for underachievers. In fact, for every seed below four (with the exception of eight), the difference in overall and road winning percentage is smaller for underachievers than overachievers. And that definitely does not square with the notion that road success correlates with tourney success. Of course, most of the lower seeds only have to win a single game to overachieve--and we don't know whether those overperforming squads are upsetting higher seeds with poor road records or not. (I could spend all week doing that analysis.) But there's no denying that the seeds which truly matter to your bracket welfare are the ones through sixes. And they clearly trend in favor of the teams with better road records. Here's another indication that road success is a solid tourney performance indicator for the top six seeds: I figured out the median road winning percentages for each of the top six seeds, then calculated the PASE value of teams above the median road winning rate for their seed and those below the median. The 58 teams above the median road winning percentage were +.179 PASE overachievers--and beat expectations 55.2% of the time. Meanwhile, the 62 teams below the median road winning rates for their seed were -.168 PASE underachievers, and beat expectations just 35.5% of the time. A +.179 PASE rate is pretty impressive for a single attribute. To put it in perspective, if you looked at the top ten attributes in the "Key Performance Indicators" study under "Feature Articles," you'd find that the PASE of teams with above-median road winning percentages would rank seventh on the list, just below frontcourt scoring and above three-pointer frequency. Sure, I only have five years of data on road records, but I think the point is clear: a team's road/neutral winning rate does indeed matter--and the better high-seeded road warriors are more likely to beat expectations in the tourney.
0 Comments...


Thu, Jan 28, 2010, 10:10 PM	By ptiernan	0 Comments...
The Bracketmaster gives users a small window into the database I've been building for two decades on tourney stats. It enables you to submit queries to the database using any combination of 18 statistics. Then it returns the round-by-round record of teams fulfilling the conditions you specify, along with their cumulative PASE valueóa measure of their degree of over- or underachievement. For some of you, this explanation may be obvious. Others don't quite get the hang of it. Every year, I answer a number of frantic post-Selection Sunday emails from people who wonder where to find the current tourney field in the Bracketmasteróor just can't figure out how to work the darn thing. The fact is, the Bracketmaster only contains the results of tourneys that have already been completed (the 2010 is in there do database purposes only), so it will never have the current year's tourney field in it. The philosophy is that past tourney trends can provide some insight into how you should fill out your bracket for the impending tourney. That's the easy misconception to clear up. Now, for those of you who struggle to make sense of the Bracketmaster, here's a quick primer. There are 18 statistics you can examine in your search queries. While only 12 are visible in the default view, if you click "More Options" at the bottom of the brown rectangle, you'll get six more. Some of the statistical fields have single drop-down menusólike "Round Reached," "School," "Conference," "Coach" and "All-Americans"óand others have two menus, allowing for you to set a range for the statistic. So let's say you wanted to know how ACC teams seeded one through six have done in the dance over the last decade when they had an experienced tourney coach, an All-American and a high-powered offense averaging more than 77 points per game. Here's what you'd do: • Under "Conference," pick "ACC" • In the first menu under "Seed," you could specify "1," but "Any" includes the lowest seed possible, so there's no need • In the second drop-down, pick "6" • For "Year," pick "2000" in the first menu and leave the second menu, since "Any" will take you up to 2009's results • For the coach's bids, you want to put "2" in the first column, since you're interested in coaches that are making at least their second trip to the dance • Under "All-Americans," pick "At Least One" • For points per game, under "Team PPG," type "77" Now that you have all your conditions defined, hit the "Submit" button. You'll see that 13 teams fulfill your conditions. They've gone 45-9 (.833), won four championships and overachieved at a +.177 PASE rate. If you want to see who the 13 teams are that meet these conditions, click "view tourney results." You'll get a pop-up window that lists the teams by year, along with their coach and record. That's a more complicated query than the average Bracketmaster user might feel up to submitting. Of course, you can also answer easy questions like, "How has Minnesota done in the 64-team tourney era?" (Answer: a 10-8 record, one Final Four appearance, and an overachieving PASE of +.104.) Or "How do one through six seeds coming into the dance with a two-game losing streak or worse fare in the tourney?) (Answer: 65-44 with an underachieving PASE of -.070 and zero championshipsósteer clear of them.) This should give you the basics of working with the Bracketmaster. In a later blog, I'll give more advanced advice on how you can use the tool to identify characteristics that correlate with underachievement and overachievement in the tourney. Then, you can use the Excel sheet I provide on Selection Sunday to find out which teams in the current field possess those qualities.
0 Comments...


Mon, Jan 25, 2010, 9:10 PM	By ptiernan	0 Comments...
Based on yesterday's AP Top 25, only seven teams have the six basic credentials that 17 of the last 19 champions have possessed: 1) They came from a Power conference (ACC, Big East, Big Ten, Big 12, SEC or Pac 10) 2) They went to the previous year's tourney 3) Their coach had at least five tourney appearances 4) They scored at least 77 points a game 5) Their average margin of victory was 10 points or more 6) They were seeded No. 1 through No. 4 Last week, eight teams met the criteria above, but Clemson fell off the champ lists, as predicted, because it dropped out of the AP Top 25. Check out this week's results: All the teams with "Y's" highlighted in blue under "Old Rule" have the six attributes cited above. That includes: Kansas, Villanova, Syracuse, Michigan State, Texas, Duke and Tennessee. This year, we're tracking a seventh champ criterion. Sixteen of 19 champions have possessed all the qualities listed aboveóand were led by a coach who had been to the Elite Eight in a previous year. Adding this criterion eliminates Tennessee. This leaves us with the same six teams as last week that have the numbers to cut down the nets in 2010: Kansas, Villanova, Syracuse, Michigan State, Texas and Duke. Last week, I figured that Clemson and Michigan State were in danger of dropping off the champ list. The Spartans managed to keep their offensive output above 77 points a gameóbut just barely with a 77.8 point average. It could be hard to accomplish that this week; MSU has two games against slow-down Michigan and Northwestern squads. This week, following my comparison of the AP Top 25 and the top possession-based teams, I'm debuting a new stat in my "champ credentials" chart. See that second column called "PY RK." That's how the AP's top 25 teams rank according to Ken Pomeroy's Pythagorean efficiency statistics. Some interesting points jump out. For one thing, top-ranked Kentucky only has the ninth-best possession-based numbersóand Villanova is fifteenth according to Pythag. Meanwhile, Duke, BYU and Wisconsin have much better efficiency ratings than their AP ranking would indicate. As a matter of fact, several teams wouldn't even be in the top 25 if the rankings were based on possession-based statistics. UAB, Gonzaga, New Mexico, UConn, Pittsburgh, Temple and Vanderbilt would all slip out of the top 25. They would be replaced by Missouri (the tenth-best Pythag team!), Maryland, Clemson, Marquette, Minnesota, California and Florida State.
0 Comments...


Fri, Jan 22, 2010, 6:52 PM	By ptiernan	0 Comments...
I just concluded negotiations with CBSSports.com to do a series of tourney articles for them starting a week or two before Selection Sunday and going up to tourney tip-off. While I miss all my ESPN.com contacts--Lunardi, Glockner (at SI.com now) and Patrick Stiegman--I'm really excited about the CBS gig. Editor Mark Swanson is great to work with, and his designers do fantastic things with my articles. The tourney teams article will be on CBS, as will the upset piece (yet to be posted), the seed guide and a few special articles. But fear not: all the real meaty stuff will remain on bracketscience.com.
0 Comments...


Wed, Jan 20, 2010, 9:22 PM	By ptiernan	0 Comments...
I have member "gbbound" to thank for this blog entry. He asked whether I was going to be researching the value of possession-based stats in predicting tourney performance. The quick answer is: absolutely. Every year, I do at least one feature article on Ken Pomeroy's possession-based, or tempo-free, statistics. For those who aren't familiar with tempo-free stats, the philosophy is simple. Ken Pomeroy and other tempo-free pioneers Dean Oliver and John Gasaway contend that raw numbers like points scored and allowed per game are only meaningful in the context of the number of a times a team possesses the ball or defends against a possession. In other words, the most accurate way to gauge a team's offensive or defensive ability is to analyze its efficiency in scoring or preventing scores. Consider this hypothetical example: which team is better offensivelyóa Wisconsin team that has 50 possessions in a game and scores 60 points, or a North Carolina team that has 80 possessions and gets 80 points? Sure, the Tar Heels score more points, but they get an average of only one point per possession while the Badgers score an average of 1.20 points. Pomeroy and company have come up with a measure of team quality called "Pythagorean winning percentage" for every team in Division I based on their offensive and defensive efficiency. I could go into great detail on how Pythag is calculated, but it would take a lot of explaining. Besides, Ken has already done it, and much better than I ever could at www.kenpom.com, including adjusting the stats for the quality of opponent and home court advantage. What I want to examine is whether the Pythag possession-based rankings bear any resemblance to the AP Top 25. The short answer is no. Take a look at the bolded "AP RK" and "PYTH" columns in this chart: According to the AP rankings, the top five teams are Texas, Kentucky, Kansas, Villanova and Syracuse. However, Pythag possession-based stats say that Duke is the country's best team, followed by Kansas, Brigham Young (say what?), Syracuse and finally Texas. Just how much exactly do the two ranking methodologies differ? Take a look at the "+/-" column beside "PYTH." This shows how many positions the Pythag ratings differ from the AP rankings. For instance, Texas is ranked first in the AP, but fifth in Pythag. That means the Longhorns' possession-based quality is four positions lower than their AP ranking. Or, if you believe in the efficacy of Pythag, you could say that Texas is overrated compared to their possession-based numbers. Now, look at the bottom of the "+/-" column. That's the average number of positions that the top 25 deviates from Pythag rankings. As it turns out, the Longhorns are fairly typical; the average team is overrated by four positions. (Just for giggles, I also looked at the absolute value of the deviations in the final "ABS" column. It turns out that the average AP Top 25 ranking is 8.4 positions off Pythag ratings, regardless of over- or underratings.) As you look down the "+/-" column, you'll see that the Longhorns aren't alone in being overrated according to Pythag numbersóand they certainly aren't the most egregious example of being ranked much higher than efficiency numbers say they deserve. The top five "overrated" teams are: Gonzaga (22 positions higher than Pythag indicates), North Carolina (also 22 positions), Villanova (15), Pittsburgh (15), and Mississippi (14). On the flip side of these overrated teams, there are 11 schools in the AP Top 25 that are underrated according to possession-based statistics. The most glaring of these is Wisconsin, which owns the sixth best Pythag winning percentageóbut is only ranked eighteenth in the AP Top 25. BYU is the second most underrated (11 positions lower than its Pythag rank), followed by Duke (6), Baylor (6) and Ohio State (5). While possession-based disciples might consider some of these under-ratings to be unjust, there are seven top-25 Pythag teams that didn't even make it into the AP Top 25. The most criminal omission is Missouri, which owns the twelfth-best Pythag winning percentage. Arizona State is the next best Pythag team on the AP 25 outs (ranked fourteenth), followed by Marquette (17), Minnesota (20), California (21), Vanderbilt (22) and Maryland (23). Come Selection Sunday, make sure you consult the Pythag winning percentage rankings (I'll have them in my big ol' Excel stat sheet). If you buy into the value of tempo-free statistics, you might consider teams like Missouri and Arizona State to be sleepers in the dance. And overrated teams like North Carolina, Gonzaga and Mississippi could be headed for a fall. (One more word on the chart above: I also included the offensive and defensive efficiency rankings, deviations and absolute valuesójust in case you wanted to see how those compared to the AP Top 25 rankings. I found it interesting that the defensive efficiency ratings were closer, on average, to the AP rankings than the offensive efficiency numbers.)
0 Comments...


Mon, Jan 18, 2010, 11:03 PM	By ptiernan	2 Comments...
Based on yesterday's AP Top 25, only eight teams have the six basic credentials that 17 of the last 19 champions have possessed: 1) They came from a Power conference (ACC, Big East, Big Ten, Big 12, SEC or Pac 10) 2) They went to the previous year's tourney 3) Their coach had at least five tourney appearances 4) They scored at least 77 points a game 5) Their average margin of victory was 10 points or more 6) They were seeded No. 1 through No. 4 Last week, 10 teams met the criteria above, but North Carolina's struggles pushed its average scoring margin below 10 points per game (by one measly point!) and Purdue's wrestling tour of the Big Ten has slowed their offense down to just 75 points per game. Check out this week's results: All the teams with "Y's" highlighted in blue under "Old Rule" have the six attributes cited above. That includes: Texas, Kansas, Villanova, Syracuse, Michigan State, Duke, Tennessee and Clemson. This year, we're tracking a seventh champ criterion. Sixteen of 19 champions have possessed all the qualities listed aboveóand were led by a coach who had been to the Elite Eight in a previous year. If we re-examine our top 25 teams by adding the requirement that their coach have an Elite Eight appearance on his resume, two more squads drop out: Purdue and Clemson. This leaves us with only six teams that have the numbers to cut down the nets in 2010: Texas, Kansas, Villanova, Syracuse, Michigan State and Duke. This is the shortest list of potential champs so far this year. My guess is that both lists will get shorter in the coming weeks. I don't think Michigan State will keep its scoring average above 77 points through the Big Ten meat grinder. And the same fate is likely to befall Clemsonóthat is, if the Tigers don't drop out of the AP Top 25 altogether.
2 Comments...


Fri, Jan 15, 2010, 8:24 PM	By ptiernan	2 Comments...
Since the tourney expanded to 64 teams in 1985, only two teams have cut down the nets one year after failing to make the tournament. The first team to do the trick was Denny Crum's Louisville Cardinals in 1986, and the second was Jim Boeheim's 2003 Syracuse Orange. This year, the Kentucky Wildcats find themselves in the same position as these two former champs. Do they have what it takes to buck the odds? The three teams actually have some key similarities. Kentucky has a rookie phenom in John Wall, just as Louisville did with "Never Nervous Pervis" Ellison and Syracuse did with Carmelo Anthony. And all three teams had or have veteran coaches who've tasted tourney success. Louisville's Denny Crum was making his twelfth tourney appearance, had five previous Elite Eight runs and one championship (1980) before cutting down the nets in '86. Jim Boeheim was making his twenty-second tourney appearance, had three previous Elite Eight trips, yet was still looking for his first crown. And John Calipari, like Crum, will be making his twelfth appearance in the dance with five Elite Eight runs. Unlike Crum, however, he doesn't have a championship under his belt. (He could've, if his 2008 Tigers shot free throws better, but I digress.) There are other important differences between the three teams. Louisville won in 1986 as a two seed and Syracuse surprised the 2003 field as a three seed. My guess is that Kentucky will cruise into the 2010 tourney as a top seed. Not only that, but the Wildcats have a more potent offense (currently 82.8 points per game) than either Louisville (78.2 ppg) or Syracuse (80.1 ppg) didódespite the fact that scoring is way down compared to those earlier years. And Kentucky's average scoring margin of 17.6 points a game is much higher than either the Cardinals or the Orange, who owned margins of 10.0 and 10.2 points per game, respectively. So let's assume for the sake of argument that Kentucky gets a top seed in the 2010 tourney. That begs a new question: how have top-seeded teams that didn't go to the previous dance performed? Here again, the odds aren't in the Wildcats' favor. There have been only seven such one seeds over the last 25 years (out of 100), and all but one of them has underachieved. Only Minnesota got as far as the Final Four in 1997. Overall, the seven squads fell considerably short of seed expectations, posting a horrid -.867 PASE. That's nearly one-game-per-tourney underperformance. Here's a look at the records and relevant statistics of the nine teams: How does Kentucky compare to these teams? On the whole, I'd say the Wildcats are better than all these squads. Only four of the seven teams had an All-American (OSU had Jim Jackson, Minnesota had Bobby Jackson, Auburn had Chris Porter and Memphis had Rodney Carter). But I would venture to say that John Wall is a better player than all of them. As far as coaching experience goes, all these seven top seeds were led by relatively green coaches. Four of the seven coaches had never been to the Elite Eight before. And only Jud Heathcote had won a tourney. Then, of course, Memphis had a less experienced and successful Calipari. This year, he'll have four more appearances, three more Elite Eight trips and one finals under his belt. As for offensive firepower, only Michigan State scored more points per game than the average top seed (83.3)óor more than Kentucky (82.8). And the only team that had a higher average scoring margin than the typical one seed (15.6 ppg) was Auburn. Kentucky is also above the average. The simple fact is, these seven top seeds have failed to make deep tourney runs largely because they were weaker and less experienced than the typical one seed. It remains to be seen whether this year's Wildcat squad will fall into the same category. They're certainly a young teamóand they could very well wind up with a weaker record than some of these seven squads. But there's no question that they have a more experienced and successful coach than any other top seeds that hadn't gone to the previous dance. And they also have the best go-to player. All that said, I'm still nervous about Kentucky. These are serious odds they're up against. An Elite Eight run? No sweat. An overachieving trip to the Final Four? Maybe. Cutting down the nets? I'm not quite ready to go that far.
2 Comments...


Tue, Jan 12, 2010, 7:24 PM	By ptiernan	0 Comments...
Based on yesterday's AP Top 25, only ten teams have the six basic credentials that 17 of the last 19 champions have possessed: 1) They came from a Power conference (ACC, Big East, Big Ten, Big 12, SEC or Pac 10) 2) They went to the previous year's tourney 3) Their coach had at least five tourney appearances 4) They scored at least 77 points a game 5) Their average margin of victory was 10 points or more 6) They were seeded No. 1 through No. 4 Last week, 10 teams also met the criteria above, but it was a slightly different list. Washington had the chops to be champs then, but the Huskies have fallen out of the Top 25. The Clemson Tigers replace them on the champ listóafter having dropped off themselves last week. Check out this week's results: All the teams with "Y's" highlighted in blue under "Old Rules" have the six attributes cited above. That includes: Texas, Kansas, Villanova, Syracuse, Purdue, Michigan State, Duke, Tennessee, North Carolina, and Clemson. This year, we're tracking a seventh champ criterion. Sixteen of 19 champions have possessed all the qualities listed aboveóand were led by a coach that had been to the Elite Eight in a previous year. If we re-examine our top 25 teams by adding the requirement that their coach have an Elite Eight appearance on his resume, three more squads drop out: Purdue, Tennessee and Clemson. This leaves us with seven teams that have the numbers to cut down the nets in 2010: Texas, Kansas, Villanova, Syracuse, Michigan State, Duke, and North Carolina. Interestingly, despite all the upheaval among AP Top 25 teams last week, there weren't a whole lot of changes to the champ list. I'm betting that's going to change after this week's games. Purdue and Clemson are the two schools at greatest risk of losing their potential champ cred. Purdue is just half a point per game over the 77-points-scored thresholdóand the Boilermakers host two slow-tempo teams, Ohio State and Northwestern. Clemson is 0.9 points per game away from missing out on the scoring criterion. Plus, the Tigers are ranked 24th, so they're teetering on the brink.
0 Comments...


Tue, Jan 12, 2010, 12:24 PM	By ptiernan	3 Comments...
As I went through my research for the contender/pretender feature article, I discovered that there were a common set of attributes across nearly all seeds that underachievers tended to possess. Since the one-through-six seeds are most likely to make deep runs in the dance, I decided to examine which attributes were the biggest indicators that a high-seeded team would fail to meet seed expectations. Here's what I discovered: Not a single one-through-six seed possessing any of these 10 qualities has won the championship. And, as you can see by the ARF (Ability to Reach the Final Four) percentages, very few of these teams get to the semi-finals. Bear in mind that the average one-through-six seed is 15.8% proficient in reaching the Final Four. According to this list, the top high-seeded team you should avoid has a rookie coach leading a team that's made fewer than five consecutive trips to the tourney. These 33 squads have all failed to reach the Final Four and fall short of seed expectations by more than half a game per tourney. That's bad. The next worst type of team has a combined rebounding and turnover edge of less than five. (If the rebounding margin is two and the team commits two fewer turnovers than their opponents, then the combined total is four.) These 31 teams have also failed to reach the Final Four, but have a slightly better PASE of -.462. Beware, also, of any team with a weak strength of schedule rating lower than 80th. Yes, two of these 30 squads have snuck into the Final Four, but their PASE is a disappointing -.321. Other attributes that stand out as being key signs of underachievement include: • The old snakebitten coach with more than five tourney bids and no Elite Eight trips (10.2% ARF and -.261 PASE) • Shooting worse than .470 from the fieldó64 teams have done it, only four have reached the semifinals, and they're -.208 PASE underperformers • A scoring margin of fewer than seven points per gameómore than 100 teams fall into this category, but only two have made the Final Four • Fewer than six wins in the last ten gamesówhile these 57 teams are slight PASE overachievers, they've never gone to the semi-finals.
3 Comments...


Mon, Jan 11, 2010, 9:36 PM	By ptiernan	0 Comments...
Nearly a month ago, I analyzed the differences between top seeds that overachieve and those that exited the dance pre-maturely. At that point, I said that I'd do the same analysis for all the other seeds by Christmas. Well, it took a heck of a lot longer than I thought it would, but the piece is finally done and posted under "Feature Articles." It turned out to be an interesting, illuminating and valuable study. In addition to examining how overachieving contenders differed from underachieving pretenders in 18 statistical categories, I also identified the conditions that improve your odds of picking an overachiever at every seed position. Check it out--and keep this article in mind com bracket picking time.
0 Comments...


Mon, Jan 4, 2010, 9:48 PM	By ptiernan	0 Comments...
Based on today's AP Top 25, only ten teams now have the six basic credentials that 17 of the last 19 champions have possessed: 1) They came from a Power conference (ACC, Big East, Big Ten, Big 12, SEC or Pac 10) 2) They went to the previous year's tourney 3) Their coach had at least five tourney appearances 4) They scored at least 77 points a game 5) Their average margin of victory was 10 points or more 6) They were seeded No. 1 through No. 4 Last week, 13 teams met the criteria above. West Virginia, which just got back onto the champ list last week, fell out again as its scoring average dropped to 75.2 per game. Ohio State and Clemson are also no longer potential champs, since they fell from the ranks of the AP Top 25. Check out this week's results: All the teams with "Y's" highlighted in blue under "Old Rules" have the six attributes cited above. That includes: Kansas, Texas, Purdue, Syracuse, Duke, Villanova, North Carolina, Michigan State, Tennessee, and Washington. This year, we're tracking a seventh champ criterion. Sixteen of 19 champions have possessed all the qualities listed aboveóand were led by a coach that had been to the Elite Eight in a previous year. If we re-examine our top 25 teams by adding the requirement that their coach have an Elite Eight appearance on his resume, three more squads drop out: Purdue, Tennessee and Washington. That leaves us with seven teams that have the numbers to cut down the nets in 2010: Kansas, Texas, Syracuse, Duke, Villanova, North Carolina and Michigan State.
0 Comments...


Sat, Jan 2, 2010, 2:09 PM	By ptiernan	0 Comments...
Have you ever seen more "decade's top ten" lists? From sports teams and movies to celebrity couples and wardrobe malfunctions, every subject is fair game. With that in mind, I figured I'd rank the ten schools that won NCAA championships in the decade. So here's the list, counting down from tenth to first. It wasn't easy: 10. 2002 Maryland It's hard to say any one of the champions of the decade is the tenth-ranked team. But someone has to fill the slot. The 2003 Terps squad is my choice. While they were a top seedóand three teams on this list weren'tóthey didn't exactly plough through their competition in the tourney. Their scoring margin in the dance was just the seventh best of the ten teams on this list, and their regular season margin was ninth best. Finally, time has shown that their starting lineup wasn't comparatively strong. Juan Dixon and Chris Wilcox are still in the NBA, and Lonny Baxter had a cup of coffee. Pretty much every other team on this list had players with better NBA credentials. 9. 2003 Syracuse By all rights, the 2003 Orange squad should be the tenth-ranked team. They were seeded three in the tourney, the worst of any champion on this list. They owned the lowest regular-season scoring margin and the lowest margin against their tourney opponents. And their lineup only boasted two NBA players. One of them, however, was Carmelo Anthony. That's worth something. In addition, this Syracuse team had the hardest road to the championship, by average seeding of opponents, of any of the decade's ten tourney winners. For those two reasons, I ranked them above Maryland. 8. 2006 Florida I was sorely tempted to rank this year's Florida squad above the 2007 team, because I think they were hungrier and playing better come tourney time. But I had to take into account the fact that the 2006 Gator team was seeded third, and owned just the eighth-best win/loss record and average scoring margin of the ten champions. Plus, despite being a three seed, the Gators actually had the second easiest road to their tourney victory, lucking out by getting an 11 seed in round two (Wisconsin-Milwaukee), a seven seed (Georgetown) instead of a two seed in the Sweet 16 and eleventh-seeded George Mason in the Final Four. Those are the reasons I put the 2006 Gators eighth. 7. 2000 Michigan State Tom Izzo's 2000 Spartan squad sent three players to the NBAóMo Peterson, Charlie Bell and Mateen Cleaves. Plus, they owned the fifth-best average scoring margin in the regular season among the ten champions and the fourth-best margin in the tourney. But this Spartan team also had the third easiest road to the crown and the worst win/loss record of any champion in the decade. And let's face it: their three big players haven't torn up the NBA. It was hard for me to rationalize putting this team any higher. 6. 2004 Connecticut Jim Calhoun's 2004 Huskies had a powerful starting lineup, with future-NBAers Ben Gordon, Emeka Okafur, Charlie Villanueva and Josh Boone. I ranked that as the third-best starting players among the decade's ten champions. However, UConn was a two seed, owned the seventh-best regular-season scoring margin and the ninth-best margin in the tourney, despite having the sixth easiest road to their crown. 5. 2005 North Carolina Roy Williams' first championship team boasted four future NBA players in Sean May, Marvin Williams, Rashad McCants and Raymond Felton. They had the fourth-best win/loss record and the fourth-best regular-season scoring margin. However, they also had the fourth easiest road to their championship, but the eighth-highest scoring margin against those fairly easy opponents. Those two factors dropped the 2005 Tar Heels into fifth place on this list. 4. 2007 Florida Billy Donovan's repeat championship squad didn't exactly storm through their regular season. The 2007 Gators owned only the sixth-best win/loss record of the decade's champions. But they also had a target on their backs all season longóand got the job done despite that pressure. Add to this the fact that they had the third hardest road to the tourney championship and put four players into the NBA, and the '07 Gators deserve the fourth spot on this list. 3. 2008 Kansas I was surprised when I did this analysis at how well the 2008 Jayhawks rated out. Not only did this team have the best regular season record of any champion in the decade, but they also ranked highest in terms of possession-based stats and second highest for average scoring margin. That said, the Jayhawks had the easiest road to their championship by cumulative seeding of opponentsóand only the fifth-best tourney scoring margin. Plus, it looks like they'll only have three future NBA players. That's not good enough to overtake either of these top two teams... 2. 2009 North Carolina I went back and forth on which of the two final teams should earn the top spot. I don't think there's any question that the 2001 Duke team and the 2009 Tar Heel squad are a notch above the other eight champions of the decade. A case could certainly be made for the Tar Heels to be rated first. They had the highest scoring margin against tourney opponents despite having the second toughest road to the championship. And their regular-season win/loss record was third best among the decade's ten champs. Plus, the 2009 Heels will likely wind up sending six players to the NBA. That said, it's debatable whether Lawson, Green, Ellington, Hansbrough, Ginyard and Thompson will have the kind of impact in the pros that the 2001 Duke players had. And that brings us to the top team of the decade, 1. 2001 Duke Duke gets the nod over North Carolina because they had the second-best regular season win/loss record, the top scoring margin and the second-best margin against their tourney opponents. In addition, their future pro playersóSean Battier, Carlos Boozer, Chris Duhon and Mike Dunleavyóhave had a significant impact in the NBA. For all these reasons, I rated the 2001 Blue Devil squad as the decade's best college basketball team.
0 Comments...


Fri, Jan 1, 2010, 10:08 AM	By ptiernan	0 Comments...
Last year, sportscaster Tim McCormick made the comment during a Michigan State-Penn State game that the Big Ten had the best coaches of any conference in the country. It was a pretty gutsy thing to say. At first, I ascribed it to McCormick's affiliation with Michigan and the Big Ten. Then I decided to find out if he was rightóat least through the measuring stick of tournament performance. And guess what I discovered? McCormick had a pretty darned good argument. Last year's crop of Big Ten coaches (same group as this year) was easily the top group in terms of performance against seed expectations, posting an impressive PASE of +.349, well ahead of the second and third place coaches in the Big East (+.167) and the ACC (+.136). Granted, ACC coaches notched more Final Four trips and championships than either the Big Ten and Big East, but they also owned a higher average seed position. This analysis was done before the results of the 2009 tourney where North Carolina downed MSU for the championship. So I got to thinking: did last year's dance change the standings among the coaches in the six Power conferences? Just as important, did such noted coaching moves as John Calipari going to the SEC and Sean Miller going to the Pac-10 have any effect on the analysis? The short answer is: not really. Tim still has it right. (Not only that, but he's a heck of a nice guy, too; he called me out of the blue shortly after my blog on conference coaches made the rounds on the internet and we shot the bull for a good 15 minutes. We had a lot to talk about. We both went to Michiganóand my high school (Andover) had the misfortune of being in his Clarkston team's conference, so we took regular poundings from Tim.) The long answer is, of course, an in-depth analysis involving a lot of numbers. I grouped all the coaches from each of the six Power conferences and compared their cumulative success in the tourney based on win/loss records, accomplishments and, most importantly, performance against seed expectations. Here's what I learned: The current 11 Big Ten coaches should've won about 89 games based on their various seed expectations. They actually won nearly 25 games more than that, for a strong PASE of +.372. For perspective, the two individual coaches with similar PASE values are Coach K (+.381) and Gary Williams (+.371). That's pretty good company. What amazes me about the Big Ten coaches is that they actually increased their PASE lead over the second-place Big East coaches., adding +.023 to their PASE over the last year, while the Big East lost .009. Perhaps more astounding is the frequency with which Big Ten coaches have beaten seed expectations. In 39 of 67 appearances, they exceeded projected win totals, for a 58.2% SOAR (seed overachievement rate) that's head and shoulders above the 50.0% SOAR of the second-best SEC coaches. Here's another impressive indication of the quality of Big Ten coaches: eight of 11 of them are historical PASE overachievers. Only Ed DeChellis (-.040), Tom Crean (-.144) and Bill Carmody (-.325) have failed to meet seed expectations. If you're keeping score, the top three Big Ten coaching overperformers are John Beilein (+.744), Tom Izzo (+.864) andóget thisóTodd Lickliter, whose two seed-defying runs with Butler have earned him a gaudy +1.175 PASE. The nearest competing stable of coaches for collective PASE performance is the Big East. The 14 coaches (Fred Hill and Norm Roberts have yet to dance) have made 105 tourney appearances and won 16.6 games more than the 160.4 that seeding projected they should have. That works out to a +.158 PASE, slightly down from last year, as I already pointed out. Eight of 14 Big East coaches have beaten seed expectations, making the conference the only one besides the Big Ten to have more than half its coaches be historical overachievers. The top three coaching PASE performers are Rick Pitino (+.662), Jay Wright (+.436) andósurprise, surpriseóStan Heath (+.393). The group is weighed down by the underachievement records of Bob Huggins (-.241), Mick Cronin (-.345) and Keno Davis (-1.130, ouch). The third best conference of coaches in terms of PASE is the ACC. They've won 8.8 games more than they should have in 90 appearances, for a PASE of +.098. Surprisingly, however, that figure is down .038 from last year's +.136 PASE. How is it possible in a year when North Carolina cut down the nets? Well, considering that Duke, Wake Forest, Florida State and Clemson all fell short of seed expectations in 2009 it more than offset the two-plus game overachievement of the Tar Heels and the fractional improvement of Maryland. Careful, though, about running down ACC coaches too much. Check out the "F4" and "CH" columns in the chart above. The ACC coaches (read: Mike Krzyzewski and Roy Williams) own more Final Four trips (20) and championships (six) than their competing coaching groups. And the last time I checked, cutting down the nets is still the ultimate in tourney success. That said, no conference is more reliant on the performance of two coaches than the ACC. Without the records of Roy and Coach K, the ACC coaches are -.105 PASE underachievers. No other conference suffers such a PASE decline when their two best coaches are taken away. The ACC's top three PASE performers are Coach K (+.381), Gary Williams (+.371) and Paul Hewitt (+.292). The bottom three are Leonard Hamilton (-.565), Oliver Purnell (-.770) and Dino Gaudio (-1.480, yowdio!). Last year, there was a noticeable gap between the coaching ranks of the Big Ten, Big East and ACC and those of the other three Power conferences. But the ACC's PASE decline coupled with the +.010 PASE improvement of the SEC means that these two conferences could easily be switching positions after this year. In 50 appearances, SEC coaches have won 4.3 games more than the seed-projected total of 69.7 wins, for a PASE of +.085, just .010 away from ACC coaches. If it's any consolation to SEC fans, their coaches are more frequent overachievers in the dance than ACC coaches. While ACC coaches have beaten seed expectations in 38 of 90 chances (42.2%), SEC coaches have done the trick in half of their chances (25 of 50). Who are the top and bottom SEC coaches? Darrin Horn (+1.480), Billy Donovan (+.702) and Trent Johnson (+.140) top the list. Bruce Pearl (-.102), Mark Fox (-.187) and Rick Stansbury (-.485) hold the conference back. The stable of coaches that made the biggest PASE improvement in the 2009 tourney is, surprisingly, the Big 12. Last year, Big 12 coaches were -.023 PASE underperformers. After the dance, they were +.025 PASE overachieversóa +.048 PASE improvement. That's what happens when all six of your coaches beat seed expectations in a single tourney. The Big 12's best PASE coaches are Mark Turgeon (+.663), Mike Anderson (+.625) and Frank Martin (+.550). Bill Self is just fourth in the conference, thanks to a couple embarrassing flame-outs in the first round. Greg McDermott (-.430), Scott Drew (-.480) and Doc Sadler (-.480) are the biggest underachievers And that brings us to the Pac-10, the only Power conference with a group of coaches that has collectively underachieved in the danceóto the tune of a -.075 PASE. Pac-10 coaches have beaten seed expectations in just 20 of 48 appearances (41.7%), the lowest frequency of any group of coaches among the Big Six. There is a ray of hope, though. The current crop of Pac-10 coaches can boast a better PASE than the conference's overall -.167 PASE throughout the 25 years of the 64-team era. But that's small consolation. Besides newcomer Sean Miller (+.478), Ben Howland (+.459) and Herb Sendek (+.099), no other Pac-10 coach is an overachiever. Meanwhile, five have failed to live up to seed expectations. (Craig Robinson and Johnny Dawkins have yet to earn dance cards.) The three biggest culprits? Kevin O'Neill (-.127), Lorenzo Romar (-.458) and Mike Montgomery (-.494).
0 Comments...


Mon, Dec 28, 2009, 11:00 PM	By ptiernan	0 Comments...
The new year is just days away and the college season is about to move into conference play. I figured it's a good time to take stock of where the nation's top teams might wind up in March. These are the top 15 teams according to the AP poll. As you're about to read, though, I don't see the tourney shaking out according to these rankings (seed and round reached in parentheses): Kansas (1 seed, Elite Eight) I don't know why I'm so skeptical of the Jayhawks. But nothing I've seen so far changes my mind that this team won't be cutting down the nets in March. Kansas has yet to play a tough road game. Sure, UCLA was supposed to be hard, but the Bruins are just 5-7óand that game wasn't exactly a laugher. The Jayhawks also struggled with Memphis on a not-so-neutral St. Louis court. I'll be very interested to see how Kansas handles two big upcoming road games against Temple and Tennessee. If they get by those opponents in convincing fashion, I may have to readjust my prediction. Texas (1 seed, Champ) Okay, so the same argument about "no impressive road wins" could be made about the Longhorns. That's fair. But unlike Kansas, Texas has played some of the elite teams in the countryóNorth Carolina and Michigan Stateóand handled them pretty well. With all their offensive weapons, the Longhorns are my early pick to be champs. Of course, Rick Barnes' less-than-sterling tourney performance worries me, and there's a long haul through the conference schedule. But right now, at this very moment, this is the best team I've watched. Kentucky (2 seed, Elite Eight) If Texas is the best team I've seen so far, John Wall is the most impressive player. Wow, is he fast. And he's already earned his reputation as a crunch-time player, saving the Wildcats' bacon against Miami of Ohio and coming through down the stretch against the Tar Heels. Still, I get this sneaking feeling that Kentucky's youth and inexperience is going be their undoing. The optimistic view is that the Wildcats are 2003 Syracuse redux; the pessimistic view is that all these near-losses will catch up with them. At this point, I'm still a skeptic. Purdue (2 seed, Final Four) We'll have a better read on Purdue within the next couple weeks, after the results of games against West Virginia, Minnesota, Wisconsin and Ohio State. For now, I like the Boilermakers in the Final Four. Besides having three strong scorers in Moore, Hummel and Johnson, they've also got the defensive toughness, the experience and the depth to get the job done. Of course, they haven't waded through the sluggish, slow-down gauntlet of the Big Ten yet. Let's see how their offense is looking after that. If Purdue is still enforcing their tempo on teams and scoring more than 75 points a game, I'll remain optimistic. Syracuse (2 seed, Elite Eight) Who would've thought before the season started that the Orange would be considered a potential top seed at this point? That's what wins against Cal, North Carolina and Florida will do for you. I'm soldóto a point. I think Syracuse or West Virginia will bag a top seed, but I don't see Syracuse overcoming their youth to pull off a Final Four run. The Elite Eight is as far as it goes for the Orangeóand that should still be considered a surprising performance for Boeheim's squad. West Virginia (1 seed, Final Four) The Mountaineers are coming off their most impressive week of the season, with wins against two solid squads in Mississippi and Seton Hall. They've got a big road game coming up on New Year's Day against Purdue that could be a preview of the Final Four in March. But even if West Virginia takes a loss in that one, I like this team. The frontline of Da-Sean Butler, Kevin Jones and Devin Ebanks is as good as there is in the country, though they're not exactly the beefiest bunch. Despite the tourney underachievement of Bob Huggins, I like the Mountaineers to reach the Final Four. Duke (1 seed, Finals) Given the Blue Devils' recent habit of underachieving, this projection is a bit of a risk. They've already shown they're vulnerable, with a close road loss against better-than-expected Wisconsin. Then again, the Dukies have also racked up impressive victories against Arizona State, UConn and Gonzaga, who they absolutely pulverized before the holidays. What makes this year's Duke squad different from the last six teams that exited early from the dance? Well, they're certainly stronger in the post than they've been in the past. If anything, it's the guard play that will be their Achilles Heel. But the real reason I'm saying Duke will overachieve this year is the law of averages. They can't fall short of seed expectations seven years in a row. Can they? Villanova (3 seed, Sweet 16) I was a big Villanova believer early in the season. So why can't I see them even getting to the Elite Eight? I've watched the Cats a couple times on TV, and despite the strength of their backcourt (Reynolds, Redding, Fisher and Stokes), I'm struck by how undersized they are. Cunningham was an essential cog in this lineupóand given that frontcourt-dominant teams rule in the later rounds of the dance, I just don't see the Cats repeating the magic of last year. I do, however, reserve the right to change my mind yet a third time on Villanova. If they get through the Big East in reasonable shape and prove that they can handle squads like Georgetown and West Virginia with tough big men, they've got the crunch-time players to go deeper, at least into the Elite Eight. North Carolina (3 seed, Sweet 16) The Tar Heels have shown that they can hang with some of the tougher teams in the land, like Michigan State and Ohio State (even when they had Evan Turner). But Roy's boys have also failed their share of tests, getting spanked by Syracuse at home and losing road games to Kentucky and Texas. (Yes, technically, that was a neutral court, but come onóit was Arlington, Texas.) While North Carolina has a lot of talent, it's raw and frontcourt-dominant. There aren't enough proven answers in the backcourt for me to see the Tar Heels getting past the Sweet 16. Connecticut (4 seed, Sweet 16) I could be way wrong on this one, but there's something about this team that makes me leery. On paper, the Huskies look strong, well deserving of their tenth-best AP ranking. Jerome Dyson and Kemba Walker are a solid guard combo. And Robinson, Edwards and Oriakhi form an intimidating frontcourt. But UConn struggles to scoreóand has an average scoring margin of just 11.4. Those are head-scratching numbers that don't figure to improve through the meat grinder of Big East play. Maybe the Huskies will gel over the course of the season. Then again, their struggles could deepen. I don't think they'll bag a top-three seed. And that makes it difficult to reach the Elite Eight. Michigan State (2 seed, Elite Eight) Nobody can accuse Tom Izzo of playing a cupcake schedule. Unlike Kansas and Texas, MSU hasn't avoided tough road games. Sure, the Spartans lost to North Carolina and Texas on their home courts, but both games were competitive until late. And Izzo learned some things about his team that should help him adjust quicker than squads that shy away from demanding schedules. Unfortunately, I think what Izzo is learning about his team is that they don't have a solid big man to compete with the elite teams that have tough frontcourts. That's not to say MSU can't spring a few surprises; Izzo can practically be counted on for that. But this year, I don't see the Spartans getting past the Elite Eight. Kansas State (5 seed, Second Round) Why is it that I'm not convinced the Wildcats are for real? Heck, they've notched a bunch of big winsóagainst Dayton, against Washington State, against Xavier. And UNLV and Alabamaóboth on the road. There aren't many teams in the country that can boast such a resume of wins. Not only that, but you're not going to find many backcourt tandems that are more productive than Jacob Pullen and Denis Clemente. So why do I have these guys seeded fifth when they're ranked twelfth in the AP Top 25? I think they'll struggle against some of the better frontcourt-dominant teams in the country. Plus, Frank Martin has to prove to me he can make a deep tourney run before I pencil the Wildcats in for for Sweet 16. Maybe next year. Georgetown (4 seed, Sweet 16) Here's one I can see myself regretting. But I'm going to go with my instincts. I don't think the Hoyas have the offensive firepower to go any deeper in the dance than the Sweet 16. Yes, Greg Monroe is one of the best big men in college. Yes, Chris Wright, Austin Freeman and Jason Clark form a strong backcourt trio. But Georgetown is a little thin on the frontline and I see them struggling in the Big East. Hence, the four seedóand the predictable inability to upend a top seed. Tennessee (3 seed, Sweet 16) By all rights, I should be projecting a better fate for the Volunteers. They're big, they're deep and they showed they can hang with the likes of Purdue, losing by just a single point on a neutral court. But I'm snakebit on Tennessee. The team always finds a way to disappoint. Usually it's their commitment to defense and lack of poise that get the better of Pearl's squad. Yes, they're playing better D this year. But the Vols really haven't beaten anyone, and they suffered a humbling blow-out loss to USC. Somehow, I feel like I'm being charitable pegging Tennessee for a three seed and a Sweet 16 run. We'll see. Ohio State (5 seed, Second Round) You've got to give the Buckeyes their due. They've played better than anyone ever expected after the loss of Evan Turner. It's a credit to the team's depth and offensive firepower that they're still ranked this highly. Unfortunately, I don't think OSU will hold up through the Big Ten season. Bagging a five seed in the dance is nothing to sneeze at, but it's not the ideal position from which to make a deep run. Darkhorses If you're keeping score, you know that there are still three positions in the top four seeds that weren't covered by today's top 15 ranked teams. There's a three seed still available and two four seeds. At this stage of the season, while it might be easy to spot the truly elite teams, it's difficult to determine those second-echelon squads. We're going to have to get into the rigors of conference play before we discover some of the better teams in the land. So if today's top 15 ranked teams won't scarf up all the three and four seeds, which teams might? Don't be surprised if you see the likes of Wisconsin or Clemson or Minnesota in the three- and four-seed slots of the March bracket. Heck, Missouri, Seton Hall or Florida might also sneak in there. And Mississippi and Mississippi State are also viable candidates. And those are just some of the Power conference hopefuls. It wouldn't shock me to see a Mid-Major sleeper nab a three or four seed. If that happens, I'd put my money on BYU, New Mexico, Dayton or Butler. The bottom line: this season has already answered a lot of questions about who might make a deep run in the dance, but there's still plenty of time for teams to rise in the brackets.
0 Comments...


Mon, Dec 28, 2009, 8:40 PM	By ptiernan	0 Comments...
Based on today's AP Top 25, 13 teams now have the six basic credentials that 17 of the last 19 champions have possessed: 1) They came from a Power conference (ACC, Big East, Big Ten, Big 12, SEC or Pac 10) 2) They went to the previous year's tourney 3) Their coach had at least five tourney appearances 4) They scored at least 77 points a game 5) Their average margin of victory was 10 points or more 6) They were seeded No. 1 through No. 4 Last week, 12 teams met the criteria above. West Virginia returned to the champ list thanks to a 90-point overtime win that pushed its scoring average above 77 points a game. Check it out: All the teams with "Y's" highlighted in blue under "Old Rules" have the six attributes cited above. That includes: Kansas, Texas, Purdue, Syracuse, West Virginia, Duke, Villanova, North Carolina, Michigan State, Tennessee, Ohio State, Washington and Clemson. This year, we're tracking a seventh champ criterion. Sixteen of 19 champions have possessed all the qualities listed aboveóand were led by a coach that had been to the Elite Eight in a previous year. If we re-examine our top 25 teams by adding the requirement that their coach have an Elite Eight appearance on his resume, four squads drop out: Purdue, Tennessee, Washington and Clemson. That leaves us with nine teams that have the numbers to cut down the nets in 2010: Kansas, Texas, Syracuse, West Virginia, Duke, Villanova, North Carolina, Michigan State, and Ohio State.
0 Comments...


Tue, Dec 22, 2009, 11:16 AM	By ptiernan	0 Comments...
Anyone who's followed the NCAA tourney knows that top seeds are by far the most dominant seeds in terms of advancement in the dance. Their success in the tournament is out of proportion to their seed advantage over the other high seeds. It's one thing to say that and back it up with numbers. It's another to see the degree of dominance. I put together this simple animated seed-advancement graph to illustrate how much one seeds tower over their second- through sixth-seeded rivals. Check it out: It's a simple yet striking illustration of the dominance of one seeds in the dance. If a picture doesn't tell the story, how about these numbers? - Top seeds are more likely to reach the Elite Eight (73%) than five or six seeds are to win their opening-round games (66% and 69%, respectively) - Top seeds reach the Final Four at nearly the same rate (44%) that two seeds reach the Elite Eight (46%) - Top seeds are about as likely to reach the finals (24%) as two through six seeds combined (25%) - Top seeds win the championship at a higher rate (15%) than four seeds reach the Elite Eight (14%) or three seeds reach the Final Four (13%)
0 Comments...


Mon, Dec 21, 2009, 8:09 PM	By ptiernan	0 Comments...
Based on today's AP Top 25, 12 teams now have the six basic credentials that 17 of the last 19 champions have possessed: 1) They came from a Power conference (ACC, Big East, Big Ten, Big 12, SEC or Pac 10) 2) They went to the previous year's tourney 3) Their coach had at least five tourney appearances 4) They scored at least 77 points a game 5) Their average margin of victory was 10 points or more 6) They were seeded No. 1 through No. 4 Last week, 11 teams met the criteria above. Clemson snuck back onto the champ list by returning to the AP Top 25 after two weeks out of it. Check out the list: All the teams with "Y's" highlighted in blue under "Old Rules" have the six attributes cited above. That includes: Kansas, Texas, Purdue, Syracuse, Duke, Villanova, Michigan State, North Carolina, Tennessee, Ohio State, Washington and Clemson. This year, we're tracking a seventh champ criterion. Sixteen of 19 champions have possessed all the qualities listed aboveóand were led by a coach that had been to the Elite Eight in a previous year. If we re-examine our top 25 teams by adding the requirement that their coach have an Elite Eight appearance on his resume, four more squads drop out: Purdue, Tennessee, Washington and Clemson. That leaves us with an elite eight group of teams that have the numbers to cut down the nets in 2010: Kansas, Texas, Villanova, Syracuse, Duke, North Carolina, Michigan State, and Ohio State. Who's at risk of falling off the champ list in the near future? Purdue is getting close to the 77-point scoring barrier, with a per-game average of 79.3 points. Amazingly, North Carolina is also teetering on the edge of dropping off the champ list. It's not because of offense. The Tar Heels average 84.9 points a game, fourth best among the Top 25 behind Kansas, Syracuse and Mississippi. It's an old bugabooóporous defenseóthat's narrowed NC's average scoring margin to just 11 points, one away from failing to meet the 10-point barrier. I'm also leery of Ohio State's, Washington's and Clemson's staying power on this list of potential champs. Once these three teams get into conference play, I think they'll struggle to stay in the AP Top 25. Time will tell.
0 Comments...


Sun, Dec 20, 2009, 10:54 AM	By ptiernan	0 Comments...
In the course of researching my last feature article about key performance indicators, it occurred to me that it might be a good idea to figure out what factors separated overachievers from underachievers at each seed position. It's one thing to do this for an entire group of seeds, like I did when I focused on one through six seeds in the KPI piece. But the fact is, each seed is unique and must take a different path through the bracket than other seeds. The best way to identify contenders from pretenders, therefore, is to do 16 separate analyses. (Okay, so we really only need to do 15, since there are no 16-seed overachievers.) Here's a taste of the new feature article I'm writing, called "09-10 Seed Contenders/Pretenders." Let's look at one seeds. As the chart below indicates, top seeds are 344-85 over the last 25 years of the tourney, winning an average of 3.44 games per appearance. Of the 100 top seeds to play since 1985, 44 have beaten seed expectations for a +1.446 PASE; that means they've averaged 4.89 wins per tourney. Meanwhile, 56 top seeds have fallen short of seed expectations, with a PASE of -1.136. That means they've averaged 2.30 wins per dance. Check it out...and look how contenders and pretenders compare in 18 statistical categories: The five factors that have the biggest percentage difference between contenders and pretenders are highlighted above in orange. The number-one factor is pre-tourney momentum. Top seed overachievers come into the dance with an average winning streak of 6.11 games, while underachievers average just 4.54 games. The second biggest factor separating top seed contenders from pretenders is having an All-American. Over 90% of the top seed overachievers have an All-American on their squad; less than 70% of underachievers have a star on their team. The third factor that distinguishes top seed overperformers from underperformers relates to coaching experience in the Elite Eight. One seeds that overachieve were led by coaches with an average of 3.8 Elite Eight trips; those that underachieved had coaches that reached the quarterfinals only 3.1 times. As you've probably noticed if you've read any of the 2010 blogs, having a coach that's gone to the Elite Eight in the past seems to be a significant prerequisite for a deep tourney run. Rounding out the top five factors separating top-seeded contenders from pretenders are Strength of Schedule and scoring margin. One seeds that overachieve play a schedule that's 15% tougher than top seeds that underachieve. And, given its high rankings in the KPI article I posted last week, it shouldn't be any surprise at all that scoring margin is a key indicator of a top-seeded overachiever. The 44 top-seeded contenders had an average scoring margin that was 12% better than the 56 pretenders. I plan to crunch the numbers for every seed except 16 (since there are no overachievers) and compile them in a single article. It should be a nice companion piece to the "Seed Match-up Guide" I write every year. Look for it before the end of 2009.
0 Comments...


Mon, Dec 14, 2009, 9:10 PM	By ptiernan	0 Comments...
After three weeks of Top 25 rankings, only 11 teams now have the six basic credentials that 17 of the last 19 champions have possessed: 1) They came from a Power conference (ACC, Big East, Big Ten, Big 12, SEC or Pac 10) 2) They went to the previous year's tourney 3) Their coach had at least five tourney appearances 4) They scored at least 77 points a game 5) Their average margin of victory was 10 points or more 6) They were seeded one through four Last week, 13 teams in the AP Top 25 met the criteria above. This week two fewer teams have what it takes to be champs. Check out the list: All the teams with "Y's" highlighted in blue under "Old Rules" have the six attributes cited above. That would be: Kansas, Texas, Purdue, Syracuse, Duke, Villanova, Tennessee, North Carolina, Michigan State, Ohio State, and Washington. West Virginia and UConn, a new addition last week, fell off of the stats champ list because they both failed to meet the 77-point-per-game scoring condition. The Mountaineers average 75.9 and the Huskies average 75.0 points a game. This year, we're tracking a seventh champ criterion. Sixteen of 19 champions have possessed all the qualities listed aboveóand were led by a coach who had been to the Elite Eight in a previous year. If we re-examine our top 25 teams by adding the requirement that their coach have an Elite Eight appearance on his resume, three more squads drop out: Purdue, Tennessee and Washington. That leaves us with an elite eight group of teams that have the numbers to cut down the nets in 2010: Kansas, Texas, Villanova, Syracuse, Duke, North Carolina, Michigan State, and Ohio State. The one glaring omission on this list is Kentucky. Why don't the Wildcats make the grade? They didn't go to last year's tournament. Only one team in the 25-year history of the 64-team era won the NCAA tourney without having gone the previous year. That was the 2003 Syracuse squad, which succeeded on the strength of a young team with a dynamic freshman, Carmelo Anthony, who gave veteran coach Jim Boeheim his first NCAA championship. Hmm, young team, dynamic freshman, veteran coach with every accolade but a tourney crown. Could Kentucky buck the odds like Syracuse did seven tourneys ago?
0 Comments...


Sun, Dec 13, 2009, 11:08 AM	By ptiernan	0 Comments...
Yesterday, I looked at the top indicators of tourney success. Today, I offer this quick overview of the factors that correlate with tourney failureómeasured both by performance against seed expectations (PASE) and ability to reach the Final Four (ARF). I evaluated every one through six seed since the tourney expanded to 64 teams back in 1985 and looked at more than two hundred statistical attributes to determine the worst PASE performers and the factors that led to the lowest percentage of Final Four contenders. Here are the five factors that correlate to the biggest underachievement against seed expectations: • RPI (-.253 PASE) - getting underseeded according to your RPI position • Three-point shooting (-.175 PASE) - Taking no more than 32.5 percent of your shots from beyond the three-point line • Rebound and turnover margin (-.160 PASE) - having a combined rebound and turnover margin of no more than six • Rookie tourney teams (-.146 PASE) - not going to the previous year's tourney • Field goal percentage (-.144 PASE) - shooting no better than 46.7 percent from the field The fact that RPI is the biggest indicator of underachievement speaks volumes about its validity as a measure of overall team quality. The class of teams that struggled the most to live up to seed expectations were those that didn't get seeded as high as their RPI position suggested they should. For example, Iowa owned an RPI of seven going into the 2006 tourney. If RPI were a valid measure of their relative strength in college basketball, the Hawkeyes should've gotten a two seed. Instead, they got a three seed. So, for those that believe in the value of RPI, Iowa should've been a team that overachieved, since they got underseeded according to their ability and weren't saddled with the higher expectations of actual two seeds. But what happened? Iowa got upset by 14 seed Northwestern State in the first round. The Hawkeyes are just one among many of these "RPI-underseeded" teams that have failed to meet expectations. Beware of them when you fill out your bracket. The second leading indicator of underachievement in the tourney won't be a surprise to those that read yesterday's blog on key overachievement indicators. Teams that take fewer than the median percentage of three-pointers (32.5 percent), irrespective of how proficient they are at making them, are -.175 PASE underachievers. Go figure. The final three factors rounding out the top five signs of early tourney exits are more intuitively obvious. Teams that struggle to rebound or protect the ball fail to meet seed expectations at a -.160 PASE rate. Teams that didn't go to the previous year's dance and those that shoot below the median for one through six seeds are also more likely to disappoint in the dance. PASE is one way to measure tourney failure. Another is ARF, or the ability to reach the Final Four. As I explained yesterday, ARF measures the percentage of teams in a given set that have made it to the semi-finals. For instance, of the 600 one through six seeds in the last 25 years, 95 have made the Final Four. The ARF for one through six seeds is therefore 95/600 = 15.8 percent. Which statistical conditions have led to the lowest ARF values for one through six seeds? Here are the top three unique factors: • Winning five or fewer of the last ten pre-tourney games (0.0 percent ARF) • Having an average scoring margin of fewer than seven points a game (1.9 percent ARF) • Owning a winning percentage less than .750 (4.9 percent ARF) The 57 high-seeded teams that came into the dance winning five or fewer of their last ten pre-tourney games all failed to reach the Final Four. That works out to a big, fat ARF of nada. The 105 high-seeded teams that don't beat their opponents by an average of seven or more points a game aren't much more proficient at getting to the semifinals. Only two have turned the trickóMississippi State in 1996 and Indiana in 2002. Steer clear of these squads in your bracket. Same goes for high-seeded squads with a winning rate below .750. There have been 204 such teams since 1985 (about eight per dance) and only ten have reached the Final Four. The two most celebrated of these are the Danny Manning-led 1988 Jayhawks and the 1997 Arizona Wildcats, both of which cut down the nets. They're definitely the exception, though. I'd think long and hard before I penciled in these "low winning percentage" teams for a deep tourney run. Some time this week, I'll be posting a feature article on key performance indicators in the tourney. It will go into more depth on the statistical conditions that lead to tourney success and failure. It should be a valuable resource for helping you separate this year's tourney contenders from the pretenders.
0 Comments...


Sat, Dec 12, 2009, 9:42 AM	By ptiernan	2 Comments...
I've nearly wrapped up my research into the key performance indicators for tourney success. This is a preview of what I've discovered. I'll be posting an extensive article some time this week. If you were deciding the fate of a team in your bracket and could only know five things about them besides seeding, what would you focus on? I evaluated every one through six seed since the tourney expanded to 64 teams back in 1985 and looked at more than two hundred statistical conditions to determine the best PASE performers and the factors that led to the highest percentage of Final Four contenders. I found out that the five factors that correlate with the highest PASE values are: • Scoring margin (+.402 PASE) - beating your opponents by more than 15 points a game • Points scored (+.367 PASE) - averaging more than 85 points a game • Coaching experience (+.285 PASE - having a coach who's reached the Elite Eight at least five times • Frontcourt scoring (+.227 PASE) - getting at least 60 percent of your points from forwards and centers • Three-point shooting (+.175 PASE) - Taking more than 32.5 percent of your shots from beyond the three-point arc It should be no surprise that scoring margin is the top indicator of a deep tourney run. This hasn't changed since I've been examining tournament stats. The fact is, scoring margin has an eerie correlation to overachievement in the dance. With each succeeding point in scoring margin, teams have increasingly higher PASE values. I'll be dedicating a later blog to this, but here's a demonstration. Check out the escalating PASE of teams satisfying these conditions: • Greater than 10 points per game - +.093 PASE • Greater than 11 points per game - +.123 PASE • Greater than 12 points per game - +.175 PASE • Greater than 13 points per game - +.219 PASE • Greater than 14 points per game - +.334 PASE • Greater than 15 points per game - +.402 PASE Offensive firepower, the second leading factor above, is a more unreliable indicator of tourney success, simply because there aren't a lot of teams scoring more than 85 points a game anymore (though NC fit the bill last year). Coaching experience, on the other hand, is something I'm vowing to remember come bracket time. If you read the earlier blog about snakebitten coaches, you already know how important it is for deep tourney runs to have a coach who's been to the Elite Eight before. The two factors among the top five that most surprised me are frontcourt scoring and three-point shooting. While tourney teams are getting progressively more guard-orientedólast year's field averaged getting 54 percent of its points from the backcourtóit's still the teams that rely on their frontcourt that most exceed expectations. And what amazes me about the three-point shooting stat is that it's not the accuracy that correlates with overachievement; it's the frequency. The teams that chuck it from the deep seats more often (that 32.5 percent figure is the median of one- through six-seeded teams) beat expectations at an impressive +.175 PASE rateóregardless of their three-point field goal percentage. That should make coaches like John Beilein smile, except you've got to get into the dance firstóand Michigan doesn't look like it's going to get it done this year. One last thing: I'm bringing back a stat I tracked in 2007 called ARFóor ability to reach the Final Four. It's a simple stat that reflects the percentage of teams in a given set that have made it to the semi-finals. For instance, of the 600 one through six seeds in the last 25 years, 95 have made the Final Four. The ARF for one through six seeds is therefore 95/600 = 15.8 percent. I looked at all the same factors I examined for the PASE analysis and crunched the ARF numbers too. Not surprisingly, scoring margin came out as the top performance indicator for ARF (and second and third best too). One- through six-seeded teams that beat opponents by more than 16 points a game have reached the Final Four 44.4 percent of the time. The next best ARF indicator, aside from margin? Winning percentage. Teams with a winning rate of .875 and higher have made the semi-finals 38.5 percent of the time. Keep your eye out for the upcoming feature article on key performance indicators in the tourney. We'll not only examine the top factors that lead to tourney success, but we'll also unveil the factors that foretell failure.
2 Comments...


Mon, Dec 7, 2009, 8:04 PM	By ptiernan	0 Comments...
Last Monday, I kicked off what will be weekly check of the teams that satisfy a basic set of six credentials that 17 of the last 19 champions have possessed. These are the six qualities that all but two champions have had since 1991: 1) They came from a Power conference (ACC, Big East, Big Ten, Big 12, SEC or Pac 10) 2) They went to the previous year's tourney 3) Their coach had at least five tourney appearances 4) They scored at least 77 points a game 5) Their average margin of victory was 10 points or more 6) They were seeded No. 1 through No. 4 The only two teams that haven't satisfied these conditions since 1991 are Michigan State (Izzo only had three bids and his Spartans scored less than 77 points per game) and Syracuse (the Orange are the only champ that hadn't gone to the previous year's tourney). Last week, 14 teams in the AP Top 25 met the criteria above. This week the 13 teams have what it takes to be champsóbut there are two important differences. Check out the list: All the teams with "Y's" highlighted in blue under "Old Rules" have the six attributes cited above. That includes: Kansas, Texas, Villanova, Purdue, West Virginia, Syracuse, Duke, Tennessee, North Carolina, Michigan State, Ohio State, Washington, and the one newcomeróUConn. Louisville and Clemson fell off the stats champ list because they both failed to make the AP Top 25. This year, we're tracking a seventh champ criterion. Sixteen of 19 champions (got this wrong last time) have possessed all the qualities listed aboveóand were led by a coach who had been to the Elite Eight in a previous year. (Only the 1998 Kentucky squad won the championship with a coachóTubby Smithówho hadn't reached the quarterfinals.) So let's re-examine our top 25 teams by adding the requirement that their coach have an Elite Eight appearance on his resume. Three squads drop out: Purdue, Tennessee and Washington. That leaves us with ten teams who have the numbers to cut down the nets in 2010: Kansas, Texas, Villanova, West Virginia, Syracuse, Duke, North Carolina, Michigan State, Ohio State and Connecticut. I'll put my money on this group right now--though Kentucky fans might have an argument. Who's in danger of dropping off either the six- or seven-rule champ list next week? Look for Texas, West Virginia and Connecticut all to struggle to meet the 77-point-per-game barrier. The Huskies, in particular, are at risk. They're already right on the dividing line.
0 Comments...


Sun, Dec 6, 2009, 9:13 PM	By ptiernan	0 Comments...
I'm starting to do my analysis into the key performance indicators of the 64-team tourney era. I'll be doing extensive analysis into the individual and multiple factors that lead to overachievement in the tourney. The first factor I looked at was having an All-American on the roster. It might seem obvious that having a bona fide star on your team would help in the tourney. But the extent of that benefit might surprise you. Take a look at the numbers below: For this analysis, I focused only on one through six seedsóthe seeds most likely to make deep tourney runs. Face it: 95 of the 100 Final Four contenders since 1985, and 24 of 25 champs, have been among the top six seeds. So I restricted my research to the 600 top six seeds of the last quarter century. What did I find out? Let's look at teams without stars first. There were 404 of them with an average seed of 4.03. Only 35 have reached the Final Fourójust 8.4% of these "no-star" squads. (That's what the "F/Ap" column means.) And just five of them, or 1.2%, have cut down the nets. Altogether, these 404 no-star teams have fallen 26.4 games short of the number of wins projected by seeding, for an underachieving PASE of -.065. Contrast this with the teams that have an All-American or two on their squad. There are 196 of them, with an average seed of 2.40. Sixty of them, or 30.6%, have reached the Final Four and 19 (9.7%) have won the championship. Those numbers are considerably better than what no-star teams have accomplishedónearly four times better for the semifinals and eight times better for the championship. Clearly, seeding is part of the reason why such a higher rate of these teams get Final Four berths and cut down the nets. But the fact is that star squads are significant overachievers, with a solid PASE of +.134. What could be better? Having two All-Americans on your team! There have only been 22 such squads, but half of them have reached the Final Four and about a quarter (five) have won the tourney. Ironically, their PASE (+.283) is about twice as good as that of one-star squads. Of course, their average seed is 1.36, so we're talking about teams that are well positioned to make deep tourney runs in the first place. But there's no denying that the Final Four and championship achievement rates, as well as the PASE value, outstrip even the rates of top seeds. When you fill out your bracket in about 100 days, take a close look at high-seeded teams with top stars. (I usually wait for U.S. Basketball Writers Association to tell me who they are right around Selection Sundayóthat's what I've used to define an All-American.) Sixty percent of the last 100 Final Four combatantsóand 76% of the tourney champsóhave had at least one All-American.
0 Comments...


Thu, Dec 3, 2009, 10:23 PM	By ptiernan	2 Comments...
Last year, I went back through all 24 brackets from the 64-team tourney era and discovered that Louisiana State and Villanova were the two schools that had sprung the most upsetsósixósince 1985. Conversely, I found that Duke and Oklahoma were the most victimized by Cinderellas. Each had suffered eight upsets. (I count an upset as any game that a team wins when it's seeded four or more positions lower than its opponent. A six seed beating a two seed is an upset; a six beating a three isn't.) One year lateróand after a quarter century of 64-team brackets, LSU and Nova remain as the top upset victors and Duke and Oklahoma are the top upset victims. But I noticed as I did this year's research that some teams appear high on both the victor and victim lists. Take LSU. Yes, the Tigers have sprung six surprises, but they've also suffered four upsets. So their net upset margin is just two games. Alabama and North Carolina are tied for thirteenth best among upset springersóbut both have been victimized five times. So their net upset marginóheck, let's make an acronym out of it and call it NUMóis negative two. So what if we reckoned the upset victor/victim balance sheet for all the teams over the last 25 years? Here are the lists of teams with the highest plus and lowest minus margins: Villanova is still tied for the top victor marginóbut it's with Auburn now and not LSU, which doesn't even appear in the top 16. The top victim margin now belongs exclusively to Duke. That's because the Blue Devils have never sprung an upset themselves, while getting humbled eight times. Oklahoma, which also owns eight upset losses, at least managed to spring a couple surprises. In fairness to Duke, the school has only played in one game where it had a low enough seed to be considered a Cinderella. In 1987, the Blue Devils lost as a five seed to top-seeded Indiana, which went on to win the tourney. Nonetheless, Duke owns the dubious distinction of having the worst victim margin of any team in the modern era. Oklahoma and Syracuse come in second, with a margin of negative six. Like Duke, Syracuse finds itself in this disappointing position because they haven't sprung an upset. Unlike Duke, however, the Orange have had their chances, playing in four upset-potential games (1985, 1992, 1995 and 1998), but failing in each.
2 Comments...


Tue, Dec 1, 2009, 12:03 PM	By ptiernan	0 Comments...
For the last three seasons, I've used a simple set of credentials that 17 of the last 19 tourney champs have possessed to project which teams might have the right stuff to cut down the nets in the current year. These are the six qualities that all but two champions have had since 1991: 1) They came from a Power conference (ACC, Big East, Big Ten, Big 12, SEC or Pac 10) 2) They went to the previous year's tourney 3) Their coach had at least five tourney appearances 4) They scored at least 77 points a game 5) Their average margin of victory was 10 points or more 6) They were seeded No. 1 through No. 4 The only two teams that haven't satisfied these conditions since 1991 are Michigan State (Izzo only had three bids and his Spartans scored less than 77 points per game) and Syracuse (the Orange are the only champ that hadn't gone to the previous year's tourney). So now that many schools have faced their first challenges of the young season, which currently have the statistical chops to be champs? Believe it or not, a record 14 teams meet the criteria above. (I base the seeds on the current AP Top 25). Here's a chart I'll be using all season for my weekly report on potential tourney champs: All the teams with "Y's" highlighted in blue under "Old Rules" have the six attributes cited above. That includes: Kansas, Texas, Villanova, Purdue, Duke, West Virginia, Syracuse, Michigan State, North Carolina, Tennessee, Washington, Georgetown, Clemson and Louisville. Teams in the Top 25 that don't meet the six key criteria are: - Kentucky, Florida and Georgetown (didn't go to last year's dance) - Gonzaga, Butler, UNLV and Portland (they're Mid-Majors) - Connecticut, Texas A&M, Florida State and Cincinnati (they don't average 77 points a game) Some of these teams fail in other categories as well. Check out the red numbers on the chart. They reflect criteria that teams don't meet. Notice the column titled "E8"óand the orange zeros beside certain coaches? This is a seventh champ criterion I'm going to track this year. Seventeen of 20 champions have possessed all the six qualities listed aboveóand were led by a coach that had been to the Elite Eight in a previous year. Only the 1998 Kentucky squad won the championship with a coach that hadn't reached the quarterfinals. That would be Tubby Smith. So let's re-examine our top 25 teams by adding the requirement that their coach have an Elite Eight appearance on his resume. Which teams drop out? Purdue's Matt Painter, Tennessee's Bruce Pearl, Washington's Lorenzo Romar and Clemson's Oliver Purnell are yet to reach the quarterfinals during March Madness. It's unlikely that they'll cut down the nets this year. That leaves us with ten teams at this early date who have the numbers under the "New Rules" (see the last column in the chart?) to cut down the nets in 2010: Kansas, Texas, Villanova, Duke, West Virginia, Syracuse, Michigan State, North Carolina, Ohio State and Louisville. Sounds like a pretty reasonable group. Of course, as the season wears on and we get into conference play, some of these teams will struggle to meet the scoring and margin requirements. West Virginia (78.8 points a game) and North Carolina (12.8 average scoring margin) are already teetering close to the dividing lines. For now, though, hope springs eternal for ten squads, and if the last few years are any indicator, the champion is among this group.
0 Comments...


Sun, Nov 29, 2009, 2:59 PM	By ptiernan	2 Comments...
Why do I never listen to my own advice? I few weeks before Selection Sunday last year, I stumbled across a pattern with veteran coaches that had seen to the dance more than five times without ever reaching the Final Four. As it turns out, not only are these types of coaches solid underachievers, but they have never won a championship. That's right: no coach who's been to the dance more than five times without reaching the Elite Eight has ever won a 64-team tourney. Here are the numbers: A coach who's been to the dance before, but hasn't reached the fourth round is a -.012 PASE underachiever. (See the second line above?) That's a slightly better performance against expectations than a coach making his first tourney appearance, but not by much. And with each year that a coach gets to the dance without reaching the Elite Eight, his PASE sinks deeper into underperformance. So by the time a coach has earned six tourney bids but failed to make a fourth-round run, he's a -.136 PASE underachiever with virtually no shot at winning a championship until he reaches the Elite Eight barrier. In fact, only two coaches have cut down the nets without ever making it to the quarterfinals in a prior year. The first to do it was rookie coach Steve Fisher in 1989, but he doesn't really count, since he never had a chance to get to the Elite Eight in a previous year. So that leaves Tubby Smith as the only coach to win a championship with first getting to the quarterfinals in one of his previous three appearances. This is what I'm now calling "The Law of the Snakebitten Coach"óthat is, a coach who hasn't reached the Elite Eight in six or more tourney tries will not win the championship. It would have been good for me to have observed this law last year when, in one of my "from the gut" brackets I chose Jamie Dixon. Big mistake. He was a snakebitten coach last year, making his sixth appearance without a quarterfinal under his belt. The good news for Pitt fans is that Dixon was able to reach the Elite Eight last year, so henceforth he is no longer officially snakebitten. The bad news is that Pitt doesn't have the sort of team that will come close to a championship in 2010. I did some further research into those 136 appearances by snakebitten coaches in the final line of the chart above and discovered that they were made by 51 coaches. Of these coaches, 37 never did reach an Elite Eight or are yet to accomplish the feat. Nine made it to the quarterfinals but never cut down the nets. They are Gene Keady, Pete Gillen, Pat Kennedy, Dave Odom, Rick Majerus, Mike Jarvis, Kelvin Sampson, Rick Barnes and Mike Montgomery. Only Barnes and Montgomery are active. That leaves five coaches that were at one point in their tourney career considered snakebitten, but managed to reach the Elite Eight, then went on in a subsequent year to win a championship. Who are they? Nolan Richardson, Jim Harrick, Gary Williams, Jim Calhoun and Jim Boeheim. Now, for the really important listóand one you should definitely consult come bracket picking time. Which active coaches are snakebitten, fulfilling the requirement of at least six tourney appearances without reaching the Elite Eight? I would definitely think twice before picking any of these coaches' teams to make a deep tourney run: 1. Fran Dunphy, Temple - 11 appearances without an Elite Eight run. 2. Mark Few, Gonzaga - 10 3. Al Skinner, Boston College - 9 4. Dana Altman, Creighton - 7 5. Stew Morrill, Utah State - 7 6. Homer Drew, Valparaiso - 7 7. Herb Sendek, Arizona State - 7 8. Mike Brey, Notre Dame - 7 9. Gregg Marshall, Wichita State - 7 10. Ralph Willard, Holy Cross - 6 11. Jeff Jones, American - 6 12. Bruce Pearl, Tennessee - 6 13. Oliver Purnell, Clemson - 6 14. Rick Stansbury, Mississippi State - 6 Some of the these coaches you would never consider for a deep tourney run whether or not they were snakebitten. Dunphy, Altman, Morrill, Drew, Marshall, Willard and Jones coach smaller schools that don't really have a prayer of cutting down the nets. But it's certainly good to know the other seven coaches from prominent schools that you should probably not advance too far in your bracket.
2 Comments...


Sun, Nov 29, 2009, 2:57 PM	By ptiernan	0 Comments...
After fully partaking in college basketball's Feast Week and madly writing team overviews, I can now declare Bracketscience.com officially open for the 2009-10 season. The Bracketmaster is up to date. Team Talk is filled out with histories, outlooks and key questions for 99 teams. The blog is in full swing. And there are seven in-depth feature articles on everything from coaching, team and conference performance to top trends and potential champs. As we go through the season, I'll be adding more articlesóand once Selection Sunday hits we'll have plenty of bracket models and contests for you to test your powers of prognostication. Enjoy!
0 Comments...


Mon, Nov 9, 2009, 11:48 AM	By ptiernan	0 Comments...
Even though Bracketscience.com doesn't officially open until December 1, I've been quietly updating parts of the site and adding new articles. That's a good thing, because many of you are already signing up and I want to reward your enthusiasm. Here's what's new: - The Bracketmaster is updated with last year's tourney results. - The first six articles in "Feature Articles" are new, including analysis on the top coaches, teams and conferences of the 25-year, 64-team era; a more robust look at champ credentials; new Madometer readings and the outlook for unpredictability in 2010; and a countdown of the top 10 tourney outliers. - All the charts for the 75 schools under "Team Talk" have been updated. I'm in the process of writing up the previews for all the schools--and will likely be adding 25 more teams to the list this year. By the time December 1 rolls around, all the graphics on the home page will be updated, more articles will get posted, and the blog will begin in earnest. I'll send out an email note letting everyone know when the site is ready to explore. Thanks to everyone who's signing up early!
0 Comments...


Sun, Oct 18, 2009, 4:47 PM	By ptiernan	0 Comments...
I'm working away to get the site ready for the December 1 opening. For those early birds who've signed up early for the 2009-10 season, I've posted two new articles analyzing the top coaches and teams of the 64-team tourney era. With the completion of the 2009 dance, there have now been 25 tourneys that used the 64-team bracket (including these past few years of play-in games). It's a perfect time to ask the questions: Who's the best coach and team of the last quarter century? You might think you know the answer to both--but you might be surprised. Here's one clue: the best coach didn't coach the best team. 'Nuf said. Check out the new pieces under "Feature Articles."
0 Comments...


Sat, Oct 10, 2009, 11:23 AM	By ptiernan	0 Comments...
I've been getting a bunch of email asking when the new material is coming for the upcoming year. Officially, we don't unveil all the new stuff until December 1 (that's because I'm busy writing it all right now), but there's still plenty of content that's fresh now. Most importantly, all the data in the Bracketmaster is up to date. So you can query my database to your heart's content. As for the feature articles, top performer lists and 2010 resources, they're coming. Over the next few weeks, I'll be quietly posting the coach, team and conference analyses, along with the tourney trends article. Then we'll start digging into the 75 team outlook pieces and more in-depth research pieces. This year, I'll be writing again for CBSSports.com--and maybe a couple of other prominent sites to be named later. I'm still in negotiations on that front. In the meantime, thanks to everyone who's been signing up so early. Sit tight: all the good stuff is coming. In fact, just to keep your appetites whetted, I'll be posting blogs on all my findings on the top coaches, teams and conferences of the last quarter century. That's right: the 64-team tourney is 25 years old now--a perfect time to settle all those heated debates that have been raging over the years. More to come...
0 Comments...