Tuesday, March 16, 2010

NCAA Bracket Blog

OK - So my first foray into bracketology was only moderately successful - but I had a lot of fun and learned a few things too. I will not even go into how the committee contradicts itself and will not answer any questions directly on any of the selection shows. That is par for the course. Instead, in case any of you were looking for some in-depth analysis on who to and and who not to pick - here goes. (Note: I have not won an office pool since 1994 when UConn beat Duke. Also, in all of the scenarios I am giving below, there are a few outliers, so nothing is fool proof).

I am going to focus my bracket analysis on Ken Pomeroy's efficiency ratings, which have become a fairly accurate predictor of the tournament results. His data goes back to 2004, so I had 6 years to look at. The first stat is the one to pay the most attention to. Ready??

Number 1
22 out of the 24 Final Four Teams was ranked in the overall Top 10 in efficiency. The two outliers were Villanova in 2009 and George Mason in 2006. 21 out of 24 were ranked in the Top 25 in both offensive and defensive efficiency. The three outliers were UCLA, LSU, and George Mason, all in 2006, so Villanova last year met this qualification. Therefore, other than George Mason in 2006, 23 of the 24 teams was either Top 10 in overall efficiency or Top 25 in both offensive and defensive efficiency. And George Mason had to beat a UConn team in OT that would have made this 24 for 24.

So - for 2010 - Your top 10 in overall efficiency are 1) Duke, 2) Kansas, 3) Wisconsin, 4) Ohio State, 5)Syracuse, 6)Kentucky, 7) BYU, 8)West Virginia, 9) Kansas State, and 10)Maryland. The first 9 teams also ranked in the Top 25 in both offensive and defensive efficiency. No other teams qualified this year in this category. This would indicate to me to pick your final four from this group. In case you didn't pick up on it, that moves Duke straight to the final four line.

Number 2
Out of the twelve Final Four games, 11 were won by the team with the higher overall efficiency. The outlier was Michigan State over Connecticut in 2009.

Number 3
The champion in each of the last 6 years was the 1 or 2 team in overall efficiency and ranked no lower than 4th in offensive efficiency. Duke and Kansas both meet this criteria. Also, 5 of the 6 years, the championship game was won by the team ranked higher in overall efficiency. UConn over Duke in 2004 (2 over 1) was the exception.

Number 4
18 out of 24 Elite 8 games were won by teams with the higher overall efficiency, and 4 of the 6 that were not were won by a team who met the 25/25 criteria. The two outliers were of course George Mason over UConn in 2006 and LSU over Texas in 2006 (LSU did have a higher defensive efficiency than Texas and was #10 overall. So what does this say - at least three of the four elite eight games will be won by the team with the higher efficiency number (3 times it has been all 4, three times it has been 2 of 4 for what that is worth.

Number 5
Except for #10 overall LSU beating #5 Duke in the Sweet Sixteen in 2006 and #4 Pitt beating #5 Wisconsin in the 2nd Round in 2005, the only time teams in the top 6 in overall efficiency did not make the Elite 8, they were defeated by a team with a better offensive efficiency (Pitt and LSU were top 10 teams with a better defensive efficiency). So, in the top 6, Duke is #1 in offensive efficiency, Kansas is #2, Wisconsin is #13, Ohio State is #7, Syracuse is #9 and Kentucky is #18. Duke, Kansas, Ohio State, Syracuse and Wisconsin can not play a team with a higher offensive efficiency before the elite 8. Kentucky at #18 would play Wisconsin at #13 with Wisconsin beating Kentucky via this stat. So let's consider the exception with defensive efficiency - Wisconsin is #7 and Kentucky is #10. Therefore Kentucky does not win the exception either. (As a Bleeding Red Badger fan, I know this will jinx Wisconsin, but it is reason for optimism). None of the other top 6 can play a top 10 team superior in defensive efficiency before the Elite 8 either.

Number 6
OK- How about some first round stats - these have more exceptions than what I gave you above - Are you liking this? Anyway, Number 6...

158 out of 192 First Round games were won by the team with the higher overall efficiency (82.3%) or 26.3 games a year. In 15 of the 34 games won by the lower efficiency team, the higher efficiency team was over-seeded by at least 4 (Over-Seeded - Efficiency rating - (Seed*4)). So now we are up to 173 out of 192 games, or 90.1% of First Round games are won by the team with the higher efficiency or lost by a team over-seeded by at least 4. Please note here, there have been many first round games won by an over-seeded team (see more on this below), but if you are picking against a higher efficiency team, it would be wise if they were over-seeded.

Number 7
More on over-seeded teams (Efficiency rating - (Seed*4)). There have been 33 teams over-seeded by at least 7 on the 1 thru 6 lines. 22 of those 33 won first round games, 18 of those over fellow over-seeded teams. But only 8 won second round games and all 8 beat over-seeded teams in the second round. None won a Sweet 16 game. So who is over-seeded by 7 on the 1 thru 6 lines? New Mexico is over-seeded by 25, Vanderbilt by 20, Pittsburgh by 18, Notre Dame by 14, Tennessee by 11 and Villanova by 7. New Mexico has a ceiling of Sweet 16 if they play Montana/Marquette, but if Washington beats Marquette, Washington is under-seeded. Vanderbilt would be Sweet 16 if they play Murray State/Butler, but would lose to under-seeded UTEP. Pittsburgh will not play an over-seeded team in the 2nd round in either circumstance. No 2nd weekend for them. If Notre Dame beats ODU (more on this later), they would face a properly seeded Baylor squad. Same for Tennessee against Georgetown. Villanova would get an under-seeded Richmond or St. Mary's before bowing out. This does not mean New Mexico would automatically beat Marquette, Vandy beat Butler, or Villanova their 2nd round opponent. Teams over seeded by 7 have lost to a handful of fellow over-seeded teams. This item only provides a maximum of the Sweet 16 for 3 of these teams and 2nd round for the other 3.

Number 8
6 out of the 7 teams under-seeded by at least 17 won their first round game, 4 out of 7 won 2 games. Utah State is under-seeded by 25 this year. 8 out of 10 teams under-seeded by at least 15 won first round games, 5 out of 10 won 2 games (and one of the two who lost was beaten by a higher seeded team who was also under-seeded). Missouri and Minnesota are both under-seeded by 16, but Missouri plays Clemson who is a higher ranked under-seeded team and Minnesota plays Xavier who is a higher ranked properly-seeded team. Proceed with caution. (Ok - Missouri plays Clemson and Oliver Purnell is 0-6 in Tourney games, so either play the streak or law of averages).

Number 9
Under-seeded 11's are 2-0 (see Old Dominion and San Diego State this year). Under-seeded 12's are just 4-4, but have won 3 in a row (hello UTEP and Utah State this year). Under-seeded 10's are also just 4-4 (Georgia Tech and Missouri).

Number 10
The team with the higher efficiency has won 35 out of 48 8/9 and 7/10 games. So there will likely be two instances of the higher efficiency team winning this year. Missouri over Clemson would seem to be 1. Wake over Texas? Louisville over Cal? Richmond over St. Mary's? Oklahoma State over Georgia Tech? Makes these games hard as always. 7 seeds rated 42 or lower are actually 4-2, with the losses coming to #'s20 and 26, but with a win over a 17. ???

I will post my bracket either tomorrow or Thursday morning. Happy bracketing!!


No comments:

Post a Comment

Let me know your feedback...