The Bracket was revealed yesterday. Quick thoughts and long ramblings below:
I have two rankings systems: my Bayesian predictive power ratings, which tell how good the teams are, and my DSMRPI ratings, which tell how much they have accomplished. I would put teams in and seed them based on DSMRPI, which looks purely at win-loss data for the team, but for SoS looks at the Ken Pomeroy adjusted efficiency differentials to get a better idea of how good the opponents actually were.
Four teams were left out that I would have put in based on DSMRPI:
- Virginia Tech (as a 9 seed or so)
- Boston College (as an 11 seed or so)
- Minnesota (as a 12 seed or so)
- Cleveland St. (as a 12 seed or so)
The four teams to be displaced?
- VCU (I have them as 16th out!)
- USC (5th out)
- Georgia (4th out)
- Marquette (3rd out)
Of course, by my Bayesian ratings, Marquette is actually quite a good team and a worthy participant (they just perhaps didn’t *deserve* to get in based on record and SoS). In fact, they’re 28th overall in the Bayesian ratings. Perhaps the committee looks a *little bit* at how good the team actually is and mostly at their resume.
Why not run a regression and find out? I used KenPom rank, Bayesian rank, and DSMRPI rank to predict the seed chosen by the committee. KenPom rank was not significant and negative, indicating that recent-play-weighted efficiency is used more by the committee than a season-wide efficiency. So after dropping KenPom out, here’s the result:
| Coefficients | Standard Error | t Stat | P-value | |
|---|---|---|---|---|
| Intercept | 1.863845332 | 0.52705 | 3.5364 | 0.000924 |
| Bayesian | 0.040652063 | 0.02565 | 1.5848 | 0.119712 |
| DSMRPI | 0.135683071 | 0.02740 | 4.9513 | 0.000010 |
So basically, the committee looks at 3.3 parts resume (win-loss + efficiency SoS) and 1 part adjusted efficiency differential, weighted toward more recent results. The R^2 was over 70%, which was good given that I was regressing onto seed-line, which is in bunches of 4 (adding noise).
And I can’t complain with that method of choice. Combine mostly resume with a little Bayesian ratings, and that’s a good way to do the seeding. I’ll call this new rating ExpSd, since it’s the seed/rating we’d expect the committee to follow.
Of course, even by that evaluation method a few teams are a long way from where they perhaps should be. All of the data is below. Utah State is the big discrepancy in the seed line. VT still looks like it should be in at the expense of VCU.
For the latest version of the Bayesian Rankings and this data, see Google Docs.
| Team | Conf | Bayesian | Rk | DSMRPI | Rk | Conf | AtL | Seed | ExpSd | ExpSd |
|---|---|---|---|---|---|---|---|---|---|---|
| Ohio St. | B10 | 33.4 | 1 | 25.8 | 1 | 1 | 0 | 1 | 1 | 1 |
| Kansas | B12 | 31.3 | 3 | 24.2 | 3 | 1 | 0 | 1 | 2 | 1 |
| Duke | ACC | 32.0 | 2 | 23.6 | 4 | 1 | 0 | 1 | 3 | 1 |
| Pittsburgh | BE | 27.3 | 4 | 20.8 | 10 | 0 | 1 | 1 | 9 | 3 |
| San Diego St. | MWC | 24.2 | 11 | 25.6 | 2 | 1 | 0 | 2 | 4 | 1 |
| North Carolina | ACC | 22.7 | 15 | 22.1 | 5 | 0 | 1 | 2 | 6 | 2 |
| Notre Dame | BE | 25.4 | 10 | 21.3 | 8 | 0 | 1 | 2 | 8 | 2 |
| Florida | SEC | 21.3 | 18 | 18.8 | 15 | 0 | 1 | 2 | 16 | 4 |
| Purdue | B10 | 25.6 | 8 | 21.6 | 7 | 0 | 1 | 3 | 5 | 2 |
| Brigham Young | MWC | 23.5 | 13 | 21.9 | 6 | 0 | 1 | 3 | 7 | 2 |
| Syracuse | BE | 25.5 | 9 | 19.9 | 12 | 0 | 1 | 3 | 10 | 3 |
| Connecticut | BE | 21.7 | 17 | 20.7 | 11 | 1 | 0 | 3 | 14 | 4 |
| Wisconsin | B10 | 26.7 | 7 | 19.8 | 13 | 0 | 1 | 4 | 11 | 3 |
| Texas | B12 | 27.3 | 5 | 19.7 | 14 | 0 | 1 | 4 | 12 | 3 |
| Kentucky | SEC | 26.9 | 6 | 18.8 | 17 | 1 | 0 | 4 | 15 | 4 |
| Louisville | BE | 24.2 | 12 | 17.0 | 21 | 0 | 1 | 4 | 18 | 5 |
| West Virginia | BE | 21.8 | 16 | 18.6 | 18 | 0 | 1 | 5 | 17 | 5 |
| Arizona | P10 | 18.5 | 31 | 18.8 | 16 | 0 | 1 | 5 | 19 | 5 |
| Kansas St. | B12 | 19.3 | 24 | 15.4 | 27 | 0 | 1 | 5 | 27 | 7 |
| Vanderbilt | SEC | 18.4 | 32 | 14.3 | 34 | 0 | 1 | 5 | 32 | 8 |
| Georgetown | BE | 19.6 | 22 | 20.8 | 9 | 0 | 1 | 6 | 13 | 4 |
| St. John's | BE | 16.6 | 39 | 17.4 | 19 | 0 | 1 | 6 | 22 | 6 |
| Cincinnati | BE | 18.5 | 30 | 15.3 | 28 | 0 | 1 | 6 | 28 | 7 |
| Xavier | A10 | 18.8 | 26 | 15.2 | 30 | 0 | 1 | 6 | 30 | 8 |
| Washington | P10 | 22.9 | 14 | 15.3 | 29 | 1 | 0 | 7 | 25 | 7 |
| Temple | A10 | 17.3 | 36 | 15.0 | 31 | 0 | 1 | 7 | 31 | 8 |
| Texas A&M | B12 | 17.2 | 37 | 14.9 | 33 | 0 | 1 | 7 | 33 | 9 |
| UCLA | P10 | 14.7 | 51 | 13.5 | 39 | 0 | 1 | 7 | 41 | 11 |
| Nevada Las Vegas | MWC | 19.5 | 23 | 16.4 | 22 | 0 | 1 | 8 | 21 | 6 |
| Michigan | B10 | 16.2 | 42 | 15.0 | 32 | 0 | 1 | 8 | 34 | 9 |
| George Mason | CAA | 15.9 | 44 | 13.2 | 40 | 0 | 1 | 8 | 40 | 10 |
| Butler | Horz | 16.3 | 41 | 12.4 | 43 | 1 | 0 | 8 | 43 | 11 |
| Illinois | B10 | 20.1 | 20 | 15.7 | 25 | 0 | 1 | 9 | 23 | 6 |
| Villanova | BE | 18.9 | 25 | 15.5 | 26 | 0 | 1 | 9 | 26 | 7 |
| Old Dominion | CAA | 15.3 | 49 | 14.0 | 35 | 1 | 0 | 9 | 37 | 10 |
| Tennessee | SEC | 15.4 | 48 | 12.2 | 46 | 0 | 1 | 9 | 46 | 12 |
| Michigan St. | B10 | 17.0 | 38 | 17.2 | 20 | 0 | 1 | 10 | 24 | 6 |
| Penn St. | B10 | 15.5 | 46 | 16.1 | 24 | 0 | 1 | 10 | 29 | 8 |
| Florida St. | ACC | 16.4 | 40 | 13.6 | 37 | 0 | 1 | 10 | 36 | 9 |
| Georgia | SEC | 12.9 | 61 | 11.7 | 54 | 0 | 1 | 10 | 55 | |
| Missouri | B12 | 18.1 | 34 | 12.8 | 41 | 0 | 1 | 11 | 38 | 10 |
| Marquette | BE | 18.7 | 28 | 11.8 | 51 | 0 | 1 | 11 | 44 | 11 |
| Gonzaga | WCC | 18.7 | 27 | 11.0 | 60 | 1 | 0 | 11 | 49 | 13 |
| Southern California | P10 | 15.4 | 47 | 11.7 | 55 | 0 | 1 | 11 | 51 | |
| Virginia Commonwealth | CAA | 9.7 | 75 | 9.4 | 69 | 0 | 1 | 11 | 71 | |
| Utah St. | WAC | 20.5 | 19 | 16.3 | 23 | 1 | 0 | 12 | 20 | 5 |
| Richmond | A10 | 15.2 | 50 | 13.6 | 38 | 1 | 0 | 12 | 39 | 10 |
| Clemson | ACC | 19.6 | 21 | 12.0 | 48 | 0 | 1 | 12 | 42 | 11 |
| UAB | CUSA | 13.5 | 60 | 12.2 | 45 | 0 | 1 | 12 | 48 | 12 |
| Memphis | CUSA | 10.6 | 72 | 10.8 | 64 | 1 | 0 | 12 | 67 | 14 |
| Belmont | ASun | 18.3 | 33 | 10.8 | 63 | 1 | 0 | 13 | 56 | 13 |
| Oakland | Sum | 10.7 | 70 | 11.8 | 52 | 1 | 0 | 13 | 57 | 13 |
| Princeton | Ivy | 6.5 | 97 | 11.7 | 53 | 1 | 0 | 13 | 64 | 13 |
| Morehead St. | OVC | 5.7 | 101 | 5.8 | 99 | 1 | 0 | 13 | 98 | 14 |
| Bucknell | Pat | 4.5 | 114 | 6.8 | 88 | 1 | 0 | 14 | 91 | 14 |
| Indiana St. | MVC | 4.9 | 111 | 6.7 | 90 | 1 | 0 | 14 | 92 | 14 |
| St. Peter's | MAAC | 2.9 | 130 | 5.8 | 100 | 1 | 0 | 14 | 104 | 15 |
| Wofford | SC | 7.9 | 89 | 4.1 | 121 | 1 | 0 | 14 | 112 | 15 |
| Long Island | NEC | 1.4 | 148 | 6.9 | 87 | 1 | 0 | 15 | 102 | 15 |
| Northern Colorado | BSky | 3.0 | 129 | 4.2 | 119 | 1 | 0 | 15 | 123 | 15 |
| UC Santa Barbara | BW | 2.3 | 134 | 3.8 | 125 | 1 | 0 | 15 | 129 | 16 |
| Akron | MAC | 4.6 | 113 | 3.1 | 132 | 1 | 0 | 15 | 131 | 16 |
| NC Asheville | BSth | -0.5 | 165 | 0.1 | 165 | 1 | 0 | 16 | 162 | 16 |
| Boston University | AE | -0.1 | 162 | -0.1 | 166 | 1 | 0 | 16 | 163 | 16 |
| Hampton | MEAC | -5.8 | 226 | -1.7 | 190 | 1 | 0 | 16 | 196 | 17 |
| Arkansas Little Rock | SB | -5.4 | 221 | -2.3 | 194 | 1 | 0 | 16 | 197 | 17 |
| Texas San Antonio | Slnd | -6.4 | 231 | -2.0 | 193 | 1 | 0 | 16 | 198 | 17 |
| Alabama St. | SWAC | -13.9 | 298 | -10.9 | 300 | 1 | 0 | 16 | 303 | 17 |
| Virginia Tech | ACC | 18.1 | 35 | 13.8 | 36 | 0 | 0 | 100 | 35 | 9 |
| Minnesota | B10 | 13.8 | 54 | 12.3 | 44 | 0 | 0 | 100 | 45 | 12 |
| Boston College | ACC | 12.8 | 62 | 12.6 | 42 | 0 | 0 | 100 | 47 | 12 |
| Northwestern | B10 | 12.7 | 63 | 11.9 | 50 | 0 | 0 | 100 | 50 | 100 |
| Cleveland St. | Horz | 8.7 | 80 | 12.0 | 47 | 0 | 0 | 100 | 52 | 100 |
| St. Mary's | WCC | 16.1 | 43 | 11.1 | 59 | 0 | 0 | 100 | 53 | 100 |
| Colorado St. | MWC | 9.3 | 77 | 12.0 | 49 | 0 | 0 | 100 | 54 | 100 |
| California | P10 | 13.6 | 57 | 11.2 | 57 | 0 | 0 | 100 | 58 | 100 |
| Washington St. | P10 | 14.0 | 53 | 10.9 | 62 | 0 | 0 | 100 | 59 | 100 |
| Oklahoma St. | B12 | 11.8 | 67 | 11.2 | 58 | 0 | 0 | 100 | 60 | 100 |
| New Mexico | MWC | 15.7 | 45 | 10.6 | 65 | 0 | 0 | 100 | 61 | 100 |
| Colorado | B12 | 13.5 | 59 | 10.9 | 61 | 0 | 0 | 100 | 62 | 100 |
| Miami FL | ACC | 14.2 | 52 | 10.2 | 66 | 0 | 0 | 100 | 63 | 100 |
| Harvard | Ivy | 7.6 | 91 | 11.4 | 56 | 0 | 0 | 100 | 65 | 100 |
| Maryland | ACC | 18.6 | 29 | 8.3 | 76 | 0 | 0 | 100 | 66 | 100 |
| Alabama | SEC | 13.8 | 56 | 8.9 | 72 | 0 | 0 | 100 | 68 | 100 |
| Southern Mississippi | CUSA | 9.9 | 73 | 9.6 | 67 | 0 | 0 | 100 | 69 | 100 |
| Missouri St. | MVC | 9.7 | 74 | 9.6 | 68 | 0 | 0 | 100 | 70 | 100 |
| Texas El Paso | CUSA | 11.9 | 65 | 8.7 | 74 | 0 | 0 | 100 | 72 | 100 |
| Marshall | CUSA | 9.0 | 79 | 9.0 | 71 | 0 | 0 | 100 | 73 | 100 |
| Seton Hall | BE | 13.5 | 58 | 7.9 | 79 | 0 | 0 | 100 | 74 | 100 |
| Wichita St. | MVC | 12.3 | 64 | 7.5 | 82 | 0 | 0 | 100 | 75 | 100 |
| Dayton | A10 | 8.3 | 83 | 7.9 | 77 | 0 | 0 | 100 | 76 | 100 |
| Mississippi | SEC | 11.9 | 66 | 7.2 | 84 | 0 | 0 | 100 | 77 | 100 |
| Valparaiso | Horz | 5.3 | 104 | 8.7 | 73 | 0 | 0 | 100 | 78 | 100 |
| College of Charleston | SC | 8.2 | 86 | 7.6 | 81 | 0 | 0 | 100 | 80 | 100 |
| Drexel | CAA | 5.2 | 107 | 8.4 | 75 | 0 | 0 | 100 | 81 | 100 |
| Nebraska | B12 | 13.8 | 55 | 6.5 | 91 | 0 | 0 | 100 | 82 | 100 |
| Rutgers | BE | 8.3 | 82 | 7.2 | 85 | 0 | 0 | 100 | 83 | 100 |
| Tulsa | CUSA | 8.3 | 84 | 7.0 | 86 | 0 | 0 | 100 | 85 | 100 |
| Boise St. | WAC | 8.5 | 81 | 6.3 | 92 | 0 | 0 | 100 | 86 | 100 |
| Baylor | B12 | 11.6 | 68 | 5.9 | 96 | 0 | 0 | 100 | 87 | 100 |
| Duquesne | A10 | 10.6 | 71 | 5.5 | 105 | 0 | 0 | 100 | 94 | 100 |
| Iona | MAAC | 11.4 | 69 | 5.5 | 106 | 0 | 0 | 100 | 97 | 100 |
Interesting post. Pomeroy’s numbers are adjusted for recency, however.
Since yours are adjusted more accurately, adding his numbers to the equation adds nothing of statistical significance.
I didn’t think his were adjusted for recency?