Animation of shifting efficiency numbers exposes potential weakness of 2- 5 seeds:

Last Tuesday, I examined the correlation between the Pythag ratings of the top teams and tourney unpredictability. Check out the blog just after this. What I discovered was that dance gets more upset- laden the lower the average efficiency of the top ten squads.

A week ago, the 2012 elite teams were the fourth worst top ten in nine years. Today, with their average Pythag falling to .9486, they’re the third worst—right behind the 2010 and 2011 top ten. Oh by the way, those two years also saw high upsets and crazy Madometer readings. So we’re in some dangerous territory.

I also compared the curve of this year’s 20 most efficient teams to their historical counterparts. What I found is that, while one and two seeds might be comparable performers to the last eight tourneys of high seeds, the 3- 5 seeds are much weaker.

That trend hasn’t changed in a week. If anything, the two and three seeds now appear weaker, while the four seeds are a ticker stronger and the fives are markedly softer. Check out this animation showing how the Pythag values have changed in three four- day periods:



What I’m going to be paying special attention to in this final month leading up to Selection Sunday is whether the average Top 10 Pythag falls below last year’s average. If that happens, we could be looking at another crazy dance. Pick your sleeper now.
... more » « less

Animation of shifting efficiency numbers exposes potential weakness of 2- 5 seeds: Last Tuesday, I examined the correlation between the Pythag ratings of the top teams and tourney unpredictability. Check out the blog just after this. What I discovered was that dance gets more upset- laden the lower the average efficiency of the top ten squads. A week ago, the 2012 elite teams were the fourth worst top ten in nine years. Today, with their average Pythag falling to .9486, they’re the third worst—right behind the 2010 and 2011 top ten. Oh by the way, those two years also saw high upsets and crazy Madometer readings. So we’re in some dangerous territory. I also compared the curve of this year’s 20 most efficient teams to their historical counterparts. What I found is that, while one and two seeds might be comparable performers to the last eight tourneys of high seeds, the 3- 5 seeds are much weaker. That trend hasn’t changed in a week. If anything, the two and three seeds now appear weaker, while the four seeds are a ticker stronger and the fives are markedly softer. Check out this animation showing how the Pythag values have changed in three four- day periods: <IMG SRC=/ blog/ Top20_ pythcomp_ 020612.gif> What I’m going to be paying special attention to in this final month leading up to Selection Sunday is whether the average Top 10 Pythag falls below last year’s average. If that happens, we could be looking at another crazy dance. Pick your sleeper now.