Photo: Public Domain
I'm testing a new way to predict the outcome of the NCAA Tournament.
First, I ranked all 68 teams based on their RPI (Rating Percentage Index) with a significant adjustment. I took each team's RPI and then subtracted the percentage of games in which the team lost by seven points or more.
I reason that in most years, every team loses multiple games. Seven-point losses, while they may have been competitive games for the most part, weren't close games. The better teams keep it close to the very end, and I'm using that as a guide to predict how teams will fare in the Tournament.
I then used my ranking to fill out my bracket. I began with the #1 team and filled in all the spots to Champion. Then, I took the next team down and filled all the spots until it ran into an already filled spot. I did this until the bracket was complete, except for one spot.
The one open spot is Illinois (56th) vs. the winner of Xavier (37th) vs Texas (64th) in a play-in game. If Xavier beats Texas, I'll have them beating Illinois as well; if Texas beats Xavier, I'll have Illinois beating Texas. I'll fill that in after Xavier-Texas is decided.
The ranking below shows each team's seeding and abbreviated region name, followed by their adjusted RPI. Sweet Sixteen teams are in bold, and the Final Four teams are in bold and italics. After the top five spots, several teams are seeded much lower than where my ranking has them. They're much better than the Selection Committee thinks.
Overall, however, the bracket is pretty generic:
Only eight (or nine, depending on Xavier-Texas) lower seeds will win in the 32 "Round of 64" games.
Six teams seeded fifth or lower will make the Sweet Sixteen
One team seeded lower than third will make the Elite Eight
One team seeded lower than first will make the Final Four and Championship game.
We know that upsets will happen, but how can we predict them? I'm hypothesizing that a team's performance in losses might be a predictor of Tournament success. This hypothesis is about to be tested.
Please let me know in the comments if you will use this ranking as a guide to filling out (one of) your bracket(s). I'm curious how it would fare in various-sized pools and would like to hear from you if you borrow this:
Duke 1-E .657
Houston 1-MW .652
St. John's 2-W .639
Florida 1-W. .633
Auburn 1-S .629
Maryland 4-W .601
Drake 11-W .572
Saint Mary's 7-E .567
Tennessee 2-MW .567
Michigan St 2-S .559
Clemson 5-MW .558
Liberty 12-E .549
High Point 13-MW .547
VCU 11-E .540
Texas Tech 3-W .533
Memphis 5-W .522
Alabama 2-E .520
Yale 13-S .520
Gonzaga 8-MW .513
New Mexico 10-S .509
Michigan 5-S .507
Wisconsin 3-E .493
UConn 8-W .482
Utah St 10-MW .477
Kentucky 3-MW .476
Akron 13-E .592 .474
McNeese 12-MW .457
Texas A&M 4-S .456
UNC Wilmington 14-W .450
Iowa St 3-S .447
Louisville 8-S .442
Oregon 5-E .437
UCSD 12-S .436
Ole Miss 6-S. .435
Troy 14-MW .422
Marquette 7-S .417
Xavier* 11-MW .412
Bryant 15-S .412
Robert Morris 15-E .404
BYU 6-E .401
Grand Canyon 13-W .396
North Carolina*11-S .394
Lipscomb 14-S .389
Miss St 8-E .388
Colorado St 12-W .381
Purdue 4-MW .377
Missouri 6-W .375
Arizona 4-E. .372
Creighton 9-S .370
UCLA 7-MW .366
Kansas 7-W .361
Baylor 9-E .361
Oklahoma 9-W .353
Montana 14-E .353
San Diego State*11-S .352
Illinois 6-MW. .347
Arkansas 10-W .320
Norfolk St 16-W .316
Georgia 9-MW .315
SIU-Edwardsville 16-MW . .313
Wofford 15-MW .289
Mt St Mary's* 16-E .262
Omaha 15-W. .259
Texas* 11-MW .251
American 16-E*. .246
Vanderbilt 10-E .224
Alabama St* 16-S .143
Saint Francis* 16-S .138
*Teams with an asterisk have a play-in game; aside from Xavier-Texas, the bracket isn't impacted by the play-ins.
James Leroy Wilson writes The MVP Chase (subscribe). Thank you for your subscription and support! James is available for writing, editing, research, and other work: jamesleroywilson-at-gmail.com.
Posts
No comments:
Post a Comment