Dolphin Ratings - College Football

Explanation: Each team is listed with its record, overall score differential, and six ratings. Brief explanations of the ratings follow.

Schedule The number in this column is the effective opponent strength of a team. In other words, they would be expected to have the same record had they played all games against an opponent of this predictive rating at a neutral site. Because this calculation depends on the strength of the team in question, it is not possible to rank schedules using these values.
Standard Ranks teams in an order such that a team is "probably better" than all teams ranked lower than it. This calculation uses margin of victory only for computing a team's opponents strengths; the team's rating itself is computed using only wins, losses, and ties relative to its schedule
Median Likelihood Determines the likely ratings for each team, based on its wins, losses, and ties relative to its schedule. This generally produces the same or similar ratings as the standard ranking.
Predictive Both schedule strength and rating vs. schedule strength are determined considering both the wins and losses and the score differentials. This rating contains none of the biases in the standard rating, but does weight recent games slightly more than past games since those are a better indication of the team's current strength. This rating is the best of the first three for seeing how good teams are, and thus is the best for predicting future results.
Improved RPI Rating The improved RPI formula is similar to the standard RPI, except that the schedule strength is carried out to infinite depth instead of ending with opponents' opponents, thus allows for a better comparison of isolated groups of teams than is given by the standard RPI calculation. It is similar to the simple rating, except that all games are given equal weight.
RPI Rating. Included only because of common real-life usage. The RPI rating has many statistical problems. The football RPI rating is based on the BCS formula, and approximates the schedule, loss, and quality win components.
Pseudo-Poll. A blended ranking based on win-loss and predictive ratings, calibrated to match real-life polls as closely as possible. This gives a rough but unbiased estimate of how typical voters value record vs. impressive wins.
Predictive-Scoring. This value indicates how many points a team would be expected to score if it played an identical team.
Predictive-Offense. This combines the predictive and scoring ratings to measure how many points a team scores. The number is the predictive rating of an opponent against whom the team would be expected to score the league average number of points. This does not necessarily rate a team's offensive abilities, as a fast pace in basketball or big-play defense in football can make a team score more points.
Predictive-Defense. This combines the predictive and scoring ratings to measure how many points a team allows. The number is the predictive rating of an opponent against whom the team would be expected to allow the league average number of points. The same caveat in the predictive-offense rating applies here.

Because these ratings contain no prejudices regarding team or conference strengths, they tend to be quite inaccurate early in the seasons. College football appears to take at least 5 games per team before even remotely reasonable ratings are produced; acceptable ratings are produced with 8 games per team; excellent ratings unfortunately require 14 games per team.

Rankings by division: Overall I-A I-AA II III NAIA NESCAC NSIC

Full Ranking


Unconnected teams:
                                               STANDARD   MED LIKELY  PREDICTIVE  IMPRVD RPI     RPI         POLL       OFFENSE     DEFENSE 
TEAM                     W  L  PF  PA  SCHED  RNK RATING  RNK RATING  RNK RATING  RNK RATING  RNK RATING  RNK RATING  RNK RATING  RNK RATING
Amherst                  7  1 169  94  0.193       1.129       0.997       0.572      0.0000      0.0000      11.279      -0.171       1.315
Wesleyan                 7  1 243 104  0.158       1.057       0.911       0.723      0.0000      0.0000      11.362       0.412       1.033
Middlebury               7  1 238 130  0.041       1.021       0.875       0.591      0.0000      0.0000      11.032       0.504       0.678
Trinity CT               6  2 259 125  0.235       0.760       0.623       0.834      0.0000      0.0000      10.684       0.828       0.839
Bates                    4  4 133 169 -0.149       0.083      -0.021      -0.317      0.0000      0.0000       9.135      -0.830       0.196
Colby                    4  4 168 143 -0.061       0.010      -0.108       0.105      0.0000      0.0000       9.175      -0.330       0.539
Bowdoin                  3  5 133 198 -0.104      -0.132      -0.264      -0.380      0.0000      0.0000       8.558      -0.676      -0.085
Williams                 2  6 139 157  0.050      -0.547      -0.635      -0.057      0.0000      0.0000       8.029      -0.554       0.440
Tufts                    0  8  94 283 -0.009      -1.055      -1.191      -1.063      0.0000      0.0000       6.122      -1.228      -0.898
Hamilton                 0  8  99 272 -0.014      -1.060      -1.198      -1.006      0.0000      0.0000       6.130      -1.155      -0.858


Conference Strengths
CONFERENCE                                                W   L   PCT  RNK RATING

Home field advantage amounts to:
     0.180 points in main ratings
     0.024 points in improved RPI
Average of 6.20 points per score

Predict score:






Team Strength vs. Time Plot:





posted: Mon Dec 23 16:38:06 2019

Back to Dolphin rankings main page.