2001 College Football Forecast (NCAA Div IA & IAA)

3 Jan 2002: Season end model rankings (NCAA Div I & IAA) (rank at final week of the season)

3 Jan 2002: Average model rankings (NCAA Div I & IAA) (average rank for entire season)

2001 Model Performance:

All Div I games               forecast probabilities
                 50-59%  60-69%  70-79%  80-89%  90-99%    100%     all
total games:        119     279     329     348     179       0    1254
number right:        61     174     247     298     167       0     947
number expected:   66.5   180.4   250.5   296.9   163.8     0.0   958.1
right/expected:    0.92    0.96    0.99    1.00    1.02    0.00    0.99

Breakdown by Conference and Division:
Conference games              forecast probabilities
                 50-59%  60-69%  70-79%  80-89%  90-99%    100%     all
total games:         78     167     213     200     118       0     776
number right:        42     100     158     168     107       0     575
number expected:   43.9   108.1   162.5   170.4   108.0     0.0   592.9
right/expected:    0.96    0.93    0.97    0.99    0.99    0.00    0.97

Non-Conference games          forecast probabilities
                 50-59%  60-69%  70-79%  80-89%  90-99%    100%     all
total games:         36     104     103     118      55       0     416
number right:        15      68      79     103      55       0     320
number expected:   19.7    67.2    78.1   100.8    50.3     0.0   316.1
right/expected:    0.76    1.01    1.01    1.02    1.09    0.00    1.01

Non-Division games            forecast probabilities
                 50-59%  60-69%  70-79%  80-89%  90-99%    100%     all
total games:          5       8      13      30       6       0      62
number right:         4       6      10      27       5       0      52
number expected:    2.9     5.2     9.9    25.7     5.5     0.0    49.1
right/expected:    1.39    1.16    1.01    1.05    0.91    0.00    1.06

actual/expected = 1 -- model right on
actual/expected > 1 -- model probabilities too low (more wins than expected)
actual/expected < 1 -- model probabilities too high (less wins than expected)


Please email comments or questions to bfm@BassettFootball.net