| Bracketmaster updated with 2008 results! |
Sat,
Jun
14,
2008
11:50 AM, By Pete Tiernan
|
As past members know, I typically don't update the Bracketmaster stats until we open for the new season in December. However, we've been testing out a new Results database link and needed to see if it worked. So the bottom line is that this year's members now have an advance view of the Bracketmaster statistics that will be unveiled for next year's members. Not only that, but the PASE values are also recalculated to reflect the 2008 tourney. Perhaps the biggest historical news from last year is that Duke is no longer the top performing team of the 64-team tourney era. The Blue Devils underperformance dropped them behind three other schools with at least 10 tourney appearances--Florida, Louisville and UNLV. That's pretty much the extent of the new analysis you'll get from me until the fall, but you can do as much of your own analysis with the updated Bracketmaster as your heart desires. And, at least for now, it's gratis. Enjoy!
|
|
| Pomeroy Pythag. game play and Final Four/champ rules fare best in the ESPN Tourney Challenge |
Thu,
Apr
10,
2008
9:05 PM, By Pete Tiernan
|
As was the case from the first round of the 2008 tourney to the last weekend, the two bracket models related to actual game-play statistics fared the best in the ESPN Tourney Challenge. The model based on Ken Pomeroy's combined offensive/defensive efficiency numbers beat out the model giving weight to key tourney-play shooting, rebounding and ballhandling PASE indicators on the tourney's last night. Pomeroy Pythag tabbed Kansas as the champ, while Game Play had Memphis. With the stunning Jayhawk comeback/Tiger collapse, Pomeroy Pythag added 160 points to its tally...and jumped into the 99th percentile. Here's how the models finished:
|
|
| Kansas win gives Don Rega the Great Eight Challenge win |
Thu,
Apr
10,
2008
8:06 PM, By Pete Tiernan
|
When Kansas beat North Carolina last Saturday, Don Rega in Mt. Pleasant, PA no doubt smiled. That's because the Jayhawk's dismantling of the Tar Heels broke a three-way stalemate at the top of the Great Eight Challenge leaderboard and gave him the win--and the iPod touch. "Boston" from--you guessed it--Boston was probably disappointed, since a Kansas loss would've handed him the Great Eight crown.
|
|
| Great Eight Tourney Challenge all boils down to Kansas |
Tue,
Apr
1,
2008
8:45 PM, By Pete Tiernan
|
Heading into the Final Four weekend, there's a three-way tie for the lead in the Great Eight Tourney Challenge. Members AJBRANDNER0, BOSTON and DONREGA9 all have 51 points. However, only DONREGA9 has a dog left in the hunt...and that would be Kansas. If Kansas beats North Carolina, DONREGA9 wins. Period.
|
|
| Does a chalky Final Four make the Selection Committee happy? |
Mon,
Mar
31,
2008
11:11 PM, By Pete Tiernan
|
Just two years after we witnessed our only Final Four of the 24-year, 64-team era without a single top seed, we now have our only Final Four with all top seeds participating. It got me to wondering: if you’re a member of the Selection Committee, does this sort of outcome make you happy?
|
|
| Elite Eight teams continue trend of higher scoring, more experience |
Sat,
Mar
29,
2008
2:53 PM, By Pete Tiernan
|
Given how dramatically scoring has decreased throughout the 23 years of the modern tourney era, it's a little surprising that this year's Elite Eight teams continue to uphold the trend of offensive firepower signalling a deep run. With the tourney field dropping 10 points off their points-per-game average from 1989 to 2008, you'd think that more defensive-oriented squads would find their way into the Elite Eight. Not so. This year's eight squads average 78.2 points a game, 4.3 points more than the overall 2008 tourney field. The Elite Eight is also tougher defensively, yielding just 63.2 points a game compared to the 64-team average of 65.4. But that drop (2.2 points) isn't as big as the rise in offense (4.3 points).
|
|
| Pomeroy Pythag, game play and historical PASE models lead the way |
Sat,
Mar
29,
2008
2:18 PM, By Pete Tiernan
|
The bracket models that are doing the best in the ESPN Tourney Challenge continue to be the ones focusing on game-play statistics and the historical PASE performance of coaches, teams and conferences. After the Sweet 16 round, here's how the models stack up (actual score in parentheses):
|
|
| The Great Eight Challenge is still up for grabs! |
Sat,
Mar
29,
2008
1:00 PM, By Pete Tiernan
|
The top of the leaderboard for the Great Eight Challenge is crowded with players that picked Davidson. But when comes to deciding who the ultimate winner will be, it will be the higher-seeded teams that have the biggest impact.
|
|
| Madometer won't swing too wildly anymore |
Sat,
Mar
29,
2008
11:23 AM, By Pete Tiernan
|
Any chance that the 2008 tourney would break the record as the most unpredictable dance of the 64-team era disappeared in the relative chalkiness of the Sweet 16. Yes, there's one double-digit seed left: as I expected it was Davidson. (Interestingly, I mentioned way back on March 13 that McKillop's bunch had all the champion attributes...except for committing the sin of being Mid-Majors. So high-scoring and high-margin served Davidson well. Eerie.)
|
|
| It's got to get a lot crazier for this to be the most unpredictable tourney |
Tue,
Mar
25,
2008
9:47 PM, By Pete Tiernan
|
If you think the 2008 tourney has been crazy enough already, it's got a heck of a long way to go before it ranks as the maddest tourney of the modern era. Currently, the Madometer sits at 16.1%, just 2.7% off the record unpredictability of 1986. That might seem like an easy gap to make up, but the math says that the odds are low.
|
|
| Sweet 16 teams tourney tested, tough on defense |
Mon,
Mar
24,
2008
9:43 PM, By Pete Tiernan
|
The teams that have reached the Sweet 16 have been to the tourney more consecutive years than the overall field and are lead by more experienced coaches. Whereas the 64-team field had been to the dance 3.37 straight years, the Sweet 16 has been to the tourney 5.06 straight times. Only West Virginia and Western Kentucky did not go to last year's dance. As for coaching, the 64-team field was led by a coach with 5.05 years of tourney experience, while the Sweet 16 is led by coaches with an average of 8.25 dance trips. The only coach making his first tourney appearance is Western Kentucky's Darrin Horn.
|
|
| Pomeroy's a genius; Pythag model has top score |
Mon,
Mar
24,
2008
8:25 PM, By Pete Tiernan
|
After two rounds of the 2008 tourney, the two best performing models in the ESPN Tourney Challenge refer to actual game-play statistics. The model based on Ken Pomeroy's Pythag ratings has done the best, comfirming the value of the statistic in evaluating team quality. Coming in second is the model that merges offensive and defensive efficiency stats with the combined PASE values of 13 game-play statistics.
|
|
| Coaching PASE rankings due for big shakeup after 2008 |
Sat,
Mar
22,
2008
10:41 AM, By Pete Tiernan
|
The spate of first-round upsets is set to wreak havoc on the PASE rankings of the 68 coaches with at least four trips to the dance. While the exact numbers can't be determined until the tourney is over (and the impact on average wins per seed is finalized), three coaches will certainly take hard falls in the rankings, and one coach will find a measure of redemption.
|
|
| Remaining 32 teams more experienced and led by coaching veterans |
Sat,
Mar
22,
2008
9:55 AM, By Pete Tiernan
|
The remaining 32 teams are different from the overall tourney field in a number of respects. The qualities that stand out most relate to experience. Whereas the 2008 tourney field had coaches with an average of 5.05 dance trips and 0.94 Elite Eight advances, the coaches of the 32 survivors have been to the tourney 6.44 times and reached the quarterfinals 1.47 times.
|
|
| The value of historical PASE shines through |
Sat,
Mar
22,
2008
1:47 AM, By Pete Tiernan
|
The opening round of the 2008 tourney was a good one for historical overachievers, as my team/coach/conference PASE model landed at the top with a 98.7 percentile showing in the ESPN Tourney Challenge. This is that model that averaged out the coaching and team PASE values for each squad, then averaged in the conference PASE. In toss-up games, it taked the higher PASE. In upset games, it only taked the lower seed if their combined PASE is 0.500 better.
|
|
| First round features five upsets, Madometer hits 16.0% |
Sat,
Mar
22,
2008
12:49 AM, By Pete Tiernan
|
That's more like it! After the chalky opening round, I was beginning to wonder whether the Selection Committee had gotten so good at seeding that they'd made the tourney boring.
|
|
| 2008 tourney field guard reliant, short on coaching experience |
Thu,
Mar
20,
2008
3:13 PM, By Pete Tiernan
|
The 2008 tourney field is unique in a number of respects. For one thing, they rely on guards for a higher percentage of points than any other field in the 24 years of the 64-team era. The field's 56.4% backcourt scoring load is even higher than last year's 55.6%--and continues the upward swing that's been happening since 1992, when the average tourney team entered the dance relying on guards for just 42.0% of their points.
|
|
| I'll be in Fan Chat for two hours on Wednesday night! |
Tue,
Mar
18,
2008
10:07 PM, By Pete Tiernan
|
|
|
| 2008 team stats Excel sheet updated with RPI, SOS an All-American info |
Tue,
Mar
18,
2008
8:41 AM, By Pete Tiernan
|
As I mentioned in a previous blog, my plan was to push out the team stats as quickly as possible on Sunday night, then double back once I got the RPI, SOS and All-American information on Monday. My source for the All-Americans is the United States Basketball Writers Association, and they named their two teams yesterday. Here's who they picked:
|
|
| ALERT on stats model #3: Two wrong picks |
Mon,
Mar
17,
2008
4:08 PM, By Pete Tiernan
|
For anyone that might've downloaded stats model #3, I double-checked the data, and there were two incorrect picks. South Alabama DID NOT have the statistical criteria to knock off Butler in round one. Meanwhile, Notre Dame DID possess the qualities to beat Washington State in round two. I hope those mistakes didn't throw anyone for a loop.
|
|
| The Great Tourney Test is over and the results are in! |
Sat,
Mar
15,
2008
12:57 AM, By Pete Tiernan
|
I knew The Great Tourney Test would be hard; I just didn’t know it would be this hard. The average score for test takers was 32.4—almost exactly half of the 65 questions. Most players wound up rated as six through 11 seeds, fitting results for a quiz on March Madness. There were, however, a select few that performed up to the quality of one or two seeds. Here are the top ten performers in The Great Tourney Test, by member name:
|
|
| Here’s what’s going to happen on Selection Sunday |
Fri,
Mar
14,
2008
8:52 PM, By Pete Tiernan
|
After the field of 65 gets announced at 6:00pm ET, we’ll be swinging into action at bracketscience.com. Just to set everyone’s expectations about what will happen, here’s the plan:
|
|
| The last check of champ potential |
Thu,
Mar
13,
2008
12:05 AM, By Pete Tiernan
|
There are just four days until Selection Sunday. This is the last time I'll evaluate the AP Top 25 against the six attributes that 15 of the last 17 champs have had. Remember the credentials for tourney titleists:
|
|
| ACC owns most Final Four trips, Mid-Majors have more than Big East |
Sun,
Mar
9,
2008
8:32 PM, By Pete Tiernan
|
Quick...which conference has reached the Final Four the most in the 64/65-team era? Okay, that's an easy one: the ACC has nailed down 21 of the 92 semi-final positions since 1985--nearly one-quarter of the slots. Here's a harder question: which conference has the second-most Final Four trips? If you said the Big East or the SEC, you would be wrong. It's the Big Ten, with 16 appearances, a little better than one out of six available positions.
|
|
| Round-by-round point spreads for the top three seeds |
Fri,
Mar
7,
2008
1:26 PM, By Pete Tiernan
|
Way back on January 10, I examined the point spreads for every seed match-up in the first three rounds of the tourney. I never got around to completing that analysis for the final three rounds, from the Elite Eight to the finals. Until today. I've restricted my analysis this time to the top three seeds, the ones that really matter as the tourney advances.
|
|
| Third- through sixth-seeded Final Four contenders and pretenders |
Wed,
Mar
5,
2008
9:42 PM, By Pete Tiernan
|
Over the last couple weeks, I've examined the statistical conditions that separate top- and second-seeded teams that reach the Final Four from those that don't. This blog analyzes the factors that separate third- through sixth seeded semi-final contenders from pretenders.
|
|
| Texas loses champ credentials, 12 teams miss by one attribute |
Mon,
Mar
3,
2008
8:08 PM, By Pete Tiernan
|
With less than two weeks until Selection Sunday, only five teams still possess the attributes that 15 of the last 17 champs have had. Remember the credentials for tourney titleists:
|
|
| Two seed Final Four contenders and pretenders |
Thu,
Feb
28,
2008
8:24 PM, By Pete Tiernan
|
Last week, I looked at the statistical conditions that separated top seeds that made the Final Four from those that fell short. As part of the research into my "Final Four/champ" article for ESPN.com, I moved on to the two seeds. Here's an advance view of what I've discovered.
|
|
| Can anyone master the Great Tourney Test? |
Thu,
Feb
28,
2008
7:57 PM, By Pete Tiernan
|
Okay, I said it would be hard...and the early returns confirm it. So far, the Great Tourney Test is being won by a seven seed. Two-year member cderrick77584 leads the pack with a score of 39. That's just one ahead of crw78. The average score of contestants who answered every question is just 26.7. That's like knowing 14 of the answers for sure and guessing on the remaining 51. How hard is the Great Tourney Test? Put it this way: I compiled the multiple choice questions two weeks (and thus knew every answer), took the test yesterday...and only scored 57.
|
|
| Which top seeds reach the Final Four—and which fall short? |
Wed,
Feb
20,
2008
9:51 PM, By Pete Tiernan
|
I’m starting to do my research into devising a statistically-driven system for identifying your Final Four and overall champion in the 2008 tourney. Like last year, I offer several methods for filling out your bracket, ranging from low to high-risk models. The model that gets the most publicity on ESPN.com is what I call my “Final Four/champion rules” model. It works backwards from the Final Four, using key performance indicators for identifying semi-final and championship candidates.
|
|
| Notre Dame gains champ credentials, Marquette gets close |
Wed,
Feb
20,
2008
7:57 PM, By Pete Tiernan
|
Believe it or not, there’s less than a month until Selection Sunday. It’s time for the fourth installment of our analysis into the teams with the statistical chops to be the 2008 champ. Remember the six criteria that 15 of the last 17 champs have met:
|
|
| When margin does and doesn’t matter |
Sat,
Feb
16,
2008
11:34 AM, By Pete Tiernan
|
I’ve long trumpeted the value of victory margin as a key indicator of tourney overperformers. In fact, the 107 teams that have entered the tourney beating opponents by more than 15 points per game have a powerful PASE of +.398. That’s the best PASE of the 224 attributes I evaluated. (For the full list, check out “Top Tourney Factors”). And if you consult the Bracketmaster, you’ll see that those 107 teams account for 38 Final Four contenders and 11 champions. Pretty strong stuff.
|
|
| Mythbuster #1: Is dominant guard play really a key to tourney advancement? |
Sun,
Feb
10,
2008
11:49 AM, By Pete Tiernan
|
To hear tourney pundits tell it, if you don’t have a dominant, solid-scoring guard, you’re not going anywhere in the dance. It is true that guard scoring has steadily increased throughout the 64/65-team era. In 1987, the teams came into the tourney getting an average of only 42.2% of their points from the backcourt. In 2007, they depended on guards for an astounding 55.6% of points.
|
|
| Tennessee and Florida join five other schools with champion credentials |
Sun,
Feb
3,
2008
9:53 PM, By Pete Tiernan
|
With just six weeks until Selection Sunday, it’s a good time to take stock which teams have the statistical qualifications to be the 2008 champion. 15 of the last 17 champs have met six criteria:
|
|
| Big East best performing middle seed, SEC top upset seed |
Fri,
Feb
1,
2008
3:11 PM, By Pete Tiernan
|
If you’ve read the analysis of conference performance, the numbers make it pretty clear that the top overachiever among the Big Six power conferences is the ACC. The average ACC squad beats seed expectations by nearly one-fifth of a game (+.185 PASE). The Big East is a solid second at +.154 PASE.
|
|
| How will Memphis perform as a top seed? |
Sun,
Jan
27,
2008
3:33 PM, By Pete Tiernan
|
It's looking more and more like Memphis will bag a top seed in the 2008 tourney. If this comes to pass, the Tigers will be only the ninth Mid-Major to be a No. 1 seed in the 64-team era. (Coincidentally, the last top-seeded Mid-Major was the 2006 Memphis squad.)
|
|
| SEC the top performing conference against tough seeds |
Thu,
Jan
24,
2008
10:52 PM, By Pete Tiernan
|
If you happened to check out the conference rankings under “Top Performers,” you know that the ACC owns the highest PASE among the six Power conferences. Their 118 teams have won about 22 games more than seeding suggests, for a solid +.185 PASE. The Big East is just behind the ACC (+.154 PASE), followed by the SEC (+.120).
|
|
| Quick overview of round-by-round seed performance |
Sat,
Jan
19,
2008
1:29 PM, By Pete Tiernan
|
You don't need to read the entire “Seed Match-ups” article under the Results Analysis menu to know how important seeding is to determining a team's tourney fate. Seeding certainly isn’t a perfect measure of the relative quality of tourney teams. But it does commit teams to a definitive path through the dance.
|
|
| Tempo-free “Pythag” rating could expose tourney over- and underachievers |
Thu,
Jan
17,
2008
7:38 PM, By Pete Tiernan
|
If you haven’t read the article on the value of tempo-free stats, you owe it to your bracket success to give it a look. Click here.
|
|
| Big changes in average wins per seed over last decade |
Wed,
Jan
16,
2008
2:30 PM, By Pete Tiernan
|
Anyone who’s frequented bracketscience.com knows about the concept of “performance against seed expectations,” or PASE. It’s the key statistic for measuring whether a team has overperformed or underperformed in the tournament. PASE is derived by comparing a team’s actual number of wins against the amount it should’ve won based on its seeding.
|
|
| Strength of Schedule now available on Bracketmaster |
Sun,
Jan
13,
2008
8:04 PM, By Pete Tiernan
|
In case you haven’t noticed, there’s a new searchable attribute on the Bracketmaster. At the end of the third row, you’ll find “Strength of Schedule.” This attribute lets you analyze the impact of schedule strength on tourney performance. Strength of Schedule, or SOS, only goes back to 1994, when the statistic was originally devised, but since then it has been a significant performance indicator.
|
|
| Possession-based stats reveal imposter teams |
Fri,
Jan
11,
2008
11:10 AM, By Pete Tiernan
|
If you don’t know about the work of Ken Pomeroy and John Gasaway on what has been alternatively called tempo-free or possession-based stats, you need to get up to speed. Pomeroy and Gasaway aren’t the only tempo-free pioneers—a guy named Dean Oliver was probably the first analyst to use these statistics. But Ken and John are the most prominent advocates for this approach to college hoops analysis. You can read their work at www.basketballprospectus.com.
|
|
| Typical bracket picking performance: 41 for 63 |
Thu,
Jan
10,
2008
6:06 PM, By Pete Tiernan
|
Last year, I pointed out that that most bracket pools are won by someone who gets between 48 and 54 correct picks out of the 63 tourney games. 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.
|
|
| Victory margins in the Sweet 16 |
Thu,
Jan
10,
2008
12:50 PM, By Pete Tiernan
|
Over the last couple weeks, I've tabulated the victory margins for every seed match-up in the first and second round. I figured I might as well finish up that work and deliver the point spreads for the later rounds. So here are the victory margins for every match-up in the Sweet 16 (the number games for each match-up is in parentheses):
|
|
| Seven teams have what it takes to be champ |
Wed,
Jan
2,
2008
10:38 AM, By Pete Tiernan
|
When the 2008 version of Bracketscience went live about a month ago, I filtered the Top 25 teams through six criteria that 15 of the last 17 tourney champs have met:
|
|
| Average victory margins in second round |
Tue,
Jan
1,
2008
4:41 PM, By Pete Tiernan
|
As you might expect, the victory margins for round-two match-ups were a lot tighter than for the opening-round games. Here's how much the favored seed has beaten the underdog seed by in every match-up since 1985:
|
|
| Average victory margins in opening round |
Sun,
Dec
23,
2007
9:52 AM, By Pete Tiernan
|
A lot of people ask me whether my database can spit out the average point spread for various seed match-ups during the modern tourney era. The short answer is no. The long answer is, I probably will go back and feed the game scores into my tourney results--but it's been somewhat of a relief to tell people I don't have those numbers. The fact is, most of the people who are interested in the point spreads need them for gambling, and--believe it or not--I'm not too big into gambling. Sure, I play a lot of tourney pools, but my main motivation for crunching these stats is to test my hoops mettle against other players--not to make a bunch of cash.
|
|
| Tourney history and 2008 outlook available for top 75 teams |
Wed,
Dec
19,
2007
9:19 PM, By Pete Tiernan
|
In case you haven't noticed, there's a new item on the main menu bar. It's called "Team Talk" and it contains historical overviews and the 2008 outlook of the top 75 schools in the 23-year modern tourney era, when the bracket expanded to 64 teams.
|
|
| Are you ready to crown your 2008 champ? |
Tue,
Dec
11,
2007
10:08 PM, By Pete Tiernan
|
Amazingly, all but two of the last 17 tourney champs had six qualities in common:
|
|
| Historical brackets reveal top upset victors and victims |
Thu,
Dec
6,
2007
9:42 PM, By Pete Tiernan
|
One of the new features of bracketscience.com is a complete collection of the 23 bracket results in the modern 64/65-team era. If you haven’t already come across these, click here.
|
|
| Bracket Science is back in session! |
Mon,
Dec
3,
2007
7:47 PM, By Pete Tiernan
|
Welcome to the 2007-08 season of Bracket Science! We said we’d open our digital doors for business on December 1—and we’ve come pretty darn close to that date. There are a few things we’re still trying to iron out (for instance, 2007 results still aren’t completely loaded in the BracketMaster). But there’s a lot of new information for bracketeers to digest.
|
|
Read the daily ramblings of Pete Tiernan as he discusses the strategies behind BracketScience.
