Monday, March 14, 2011

96% Success Predicting NCAA Tournament Field

Based on my final NCAA tournament projection, I correctly predicted 65 of the 68 tournament teams (96%), which included the accurate selection of 34 of the 37 at-large teams (92%). Over the past four years, I have correctly forecast 255 of the 263 tournament teams (97%).

The three disagreements I had with the Selection Committee were very minor. The three teams that I projected in that ultimately did not make the field were Virginia Tech, Colorado, and Alabama. I acknowledged all three as bubble teams and even listed Alabama among my last three in. All three were rewarded with 1-seeds in the NIT.

The three teams that I projected out of the field that ultimately made it were VCU, Penn State, and USC. I acknowledged all three as bubble teams and listed VCU and Penn State among my first three out. The other team I acknowledged among my first three out was Boston College, which was awarded a 1-seed in the NIT.

Ultimately, none of these six teams had a realistic shot at running through the tournament to capture the national championship, so the issue is ultimately moot. If I could make one change to the Selection Committee's decisions, it would have been to put Virginia Tech in and take USC out. The resumes for each of the six are highlighted below for your personal review.


#57 VCU, 23-11 (14-7), CAA
  • SOS rank: 85
  • Last 12 games: 6-6
  • Home: 13-4, Road: 8-6, Neutral: 2-1
  • Close Games: 4-3, Best Case: 26-8, Worst Case: 19-15
  • Top 50: 3-5, 51-100: 6-3, 101-200: 2-0, 201+: 12-3
  • Key Wins: Old Dominion, George Mason, James Madison, Drexel (2), UCLA, Wofford, Wichita State
  • Verdict: NCAA = in; Rogers Poll = first three out
#59 Virginia Tech, 21-11 (11-8), ACC
  • SOS rank: 61
  • Last 12 games: 7-5
  • Home: 11-3, Road: 6-7, Neutral: 4-1
  • Close Games: 2-4, Best Case: 25-7, Worst Case: 19-13
  • Top 50: 1-5, 51-100: 4-3, 101-200: 8-3, 201+: 8-0
  • Key Wins: Duke, Oklahoma State, Penn State, Florida State (2)
  • Verdict: NCAA = out; Rogers Poll = in
#67 Alabama, 21-11 (13-5), SEC
  • SOS rank: 98
  • Last 12 games: 8-4
  • Home: 16-0, Road: 5-7, Neutral: 0-4
  • Close Games: 4-1, Best Case: 22-10, Worst Case: 17-15
  • Top 50: 1-4, 51-100: 4-2, 101-200: 5-4, 201+: 11-1
  • Key Wins: Kentucky, Georgia (2), Ole Miss, Tennessee
  • Verdict: NCAA = out; Rogers Poll = last three in
#76 Penn State, 19-14 (12-10), Big Ten
  • SOS rank: 4
  • Last 12 games: 7-5
  • Home: 13-5, Road: 3-8, Neutral: 3-1
  • Close Games: 4-3, Best Case: 22-11, Worst Case: 15-18
  • Top 50: 3-6, 51-100: 6-6, 101-200: 5-2, 201+: 5-0
  • Key Wins: Wisconsin (2), Illinois, Fairfield
  • Verdict: NCAA = in; Rogers Poll = first three out
#88 Colorado, 21-13 (10-9), Big XII
  • SOS rank: 66
  • Last 12 games: 7-5
  • Home: 15-2, Road: 3-10, Neutral: 3-1
  • Close Games: 5-2, Best Case: 23-11, Worst Case: 16-18
  • Top 50: 5-6, 51-100: 3-3, 101-200: 3-4, 201+: 9-0
  • Key Wins: Texas, Nebraska, Kansas State (3), Missouri, Oklahoma State, Colorado State
  • Verdict: NCAA = out; Rogers Poll = in
#99 USC, 19-14 (11-9), Pac-10
  • SOS rank: 37
  • Last 12 games: 7-5
  • Home: 12-5, Road: 5-8, Neutral: 2-1
  • Close Games: 4-5, Best Case: 24-9, Worst Case: 15-18
  • Top 50: 3-4, 51-100: 3-4, 101-200: 8-3, 201+: 5-3
  • Key Wins: Arizona, Texas, Washington State, UCLA, Washington, Tennessee
  • Verdict: NCAA = in; Rogers Poll = out


All 31 teams listed as auto-bids, 4 of which were still to be determined at the time of the projection, indeed received automatic berths to the NCAA tournament. In addition, all 24 of the teams I listed as "locks" were awarded at-large berths.

Moreover, 10 of the 13 "bubble" teams that I expected to receive an at-large bid were granted one. The three exceptions (Virginia Tech, Colorado, and Alabama) instead received a #1 seed in the NIT. VCU, Penn State, and USC received at-large berths in their place, each teams that I had listed on the "bubble" but not expected to earn an at-large (though VCU and Penn State were among my first three out).

In the case of the other 11 "bubble" teams that I had listed and accurately expected to be left out of the NCAA tournament, 6 received berths in the NIT, 2 received berths in the CIT, and 2 declined postseason opportunities (Drexel and Tulsa).

30 teams were profiled in my "miracle watch", a category reserved for key teams available for the NIT, CBI, and CIT selection pools. 16 were awarded berths in the NIT, 2 were awarded berths in the CBI, and 3 were awarded berths in the CIT.

The 9 programs profiled in the "miracle watch" that did not have a postseason destination were Southern Miss, American, Maryland, Minnesota, Baylor, Arkansas, Mississippi State, Iowa State, and Virginia. All nine abstained from the CBI and CIT tournaments, for various reasons, after being passed over for the NIT.