Biggest Seed Mistakes:
- -2 – UNLV, Florida St, Vanderbilt, Florida, Marquette, Clemson, California
- +2 – San Diego St, Gonzaga, Temple, Washington, St Marys
- +3 – Northern Iowa
- +4 – UTEP
Overall Seed Accuracy:
- 30 exact
- 21 1 seed off
- 14 more than 2
- Oklahoma St/Georgia Tech
- Pittsburgh/Oakland (although my seeding was 4/13)
- West Virginia/Morgan St
- Looking back at the biggest seeding misses, my computer seeding projections overvalued most of the mid-major schools and under-valued some of the mediocre big-conference schools. And the one big-conference school that I overvalued was from the Pac-10, which played like a mid-major this year. I’ll spend the next few days reviewing my calculations to determine how best to stop overvaluing these teams.
- I’m happy to have gotten all 34 at-large teams correct. I didn’t have much time today to review the bracket, and what little time I had I devoted to that last spot. After concluding that Minnesota was in no matter what, and Virginia Tech was not going to make it, I spent some time reviewing the profiles of Mississippi State, Florida, and Illinois. As is often the case with these decisions, it’s a matter of teams not making it because of some disqualifying characteristic rather than a team making it for positive reasons. A I stated in the Final Bracket post, Mississippi State’s SOS and lack of top 25 RPI wins, and Illinois’ bad RPI and record were the disqualifying factors for those teams.
- In the run-up to today, I was mystified by seeing Kentucky being sent to Milwaukee in so many mock brackets. The reason generally given was that Milwaukee was the closest first-weekend site to Lexington. However, in the NCAA bracketing guidelines, no mention is made of exact distance in determining a team’s playing site. The only statement in the guidelines refers to placing a team in their general region of interest. To me, this meant that Kentucky would either play in Jacksonville or New Orleans. Also, since I had 3 Big Ten teams as top 4 seeds, I figured that 2 of them would be playing in Milwaukee. Given the lack of other strong SEC or ACC teams besides Duke, I decided to place Kentucky in Jacksonville. I was wrong, but Kentucky did get sent to New Orleans.
- In the bracketing guidelines, there is mention made of placing teams from the same conference in the same pod (so as to play in the Sweet 16 if both teams get that far), but only if other bracketing possibilities are exhausted. I believe this is the first year this was done, and it was done twice for the Big East. I especially don’t like seeing it done in the lower half of the East regional, as I think there’s a good chance that Marquette and West Virginia will play each other in the Sweet 16. I’d like to see the committee go away from this practice. I don’t think it’s too hard to separate conference teams, even if there are 8 from the same conference.
- Before today, I often heard from the TV experts that the field was wide-open this year, with any number of teams who could win it all. I was surprised to hear this, because to me Kansas has far and away the best team this year, with veteran leadership, strong inside play, and a versatile offense that can beat you in a number of ways. Now it appears that many of the TV experts are picking Kansas to win it all. I think they’ll end up beating Kentucky and would not be surprised to see something similar to the 1992 title game, where a veteran Duke team beat a young Michigan team quite handily in the championship game.
Possible Future Changes:
- Factor in # Conference Losses for Mid-major teams
- Less Weight on Conference Tournament victories (Of the 7 teams that were overvalued, 5 won their conference tournaments, and the other 2 made the finals)
- Factor in a Regular-Season Big Conference Champion Bonus