I realize I'm not the first dope to use statistics to try and predict who will win the NCAA tournament. That said, I have taken the liberty of at least starting with the field of 16 to hopefully give myself a better chance of getting things right. For this exercise, I use a model that estimates a team's probability of winning based on team FT%, FG%, steals per game, rebounds per game and assist-to-turnover ratio (ATR) (I did a similar study estimating exactly how important FT% was when determining team winning percentage).
I'll break this down by games and see how well I can model (predict) the winners of Thursday's and Friday's games. So here goes (a note about the model -- this is just a quick and dirty attempt to see how well I can predict the winners. I realize there are many variables I left out, but I figure at the very least, it's a start):
*The winner of each game is in bold
Vanderbilt vs. UConnSo there it is, the final 8 -- UConn, OSU, Syracuse, St. Joe's, UAB, Ga. Tech, Illinois and Texas. The Thursday games are all won by the higher seed, but the Friday games have two upsets: UAB beats Kansas and Illinois beats Duke. Let's hope I do better here than I did in my NCAA pool (I think I have 5 teams left). I'll post the final four and champion tomorrow.
(Thursday, March 25)
Vanderbilt's chance of winning: 75.8%
UConn's chance of winning: 83.8%
Pittsburgh vs. Oklahoma St.
(Thursday, March 25)
Pittsburgh's chance of winning: 70.9%
Oklahoma St's chance of winning: 79.1%
Alabama vs. Syracuse
(Thursday, March 25)
Alabama's chance of winning: 60.8%
Syracuse's chance of winning: 77.9%
Wake Forest vs. St. Joseph's
(Thursday, March 25)
Wake Forest's chance of winning: 75.9%
St. Joseph's chance of winning: 82.7%
UAB vs. Kansas
(Friday, March 26)
UAB's chance of winning: 76.6%
Kansas's chance of winning: 74.5%
Nevada vs. Ga. Tech
(Friday, March 26)
Nevada's chance of winning: 69.8%
Ga. Tech's chance of winning: 73.4%
Illinois vs. Duke
(Friday, March 26)
Illinois's chance of winning: 74.7%
Duke's chance of winning: 74.2%
Xavier vs. Texas
(Friday, March 26)
Xavier's chance of winning: 62.5%
Texas's chance of winning: 74.5%
*If anyone is interested (and I doubt they are) in the model I used to get these results, let me know.
