2004 Week 1 (9-13 Sep) Results for the National Football League

prob rnk winning_team        actual pred    rnk losing_team         actual pred

 78%  10 St. Louis               17 34.2,    32 Arizona                 10 13.8
 73%   3 Philadelphia            31 27.7,    29 New York Giants         17 12.8
 63%  17 New York Jets           31 24.1,    27 Cincinnati              24 16.8
 61%  12 Minnesota               35 22.6,    18 Dallas                  17 17.9
 59%   4 New England             27 23.9,     5 Indianapolis            24 20.8
 58%   2 Green Bay               24 22.1,    14 Carolina                14 19.4
 54%  21 Pittsburgh              24 19.9,    19 Oakland                 21 18.7
 52%  28 San Diego               27 21.6,    31 Houston                 10 20.8

 47%   8 Denver                  34 24.7,     1 Kansas City             24 25.5
 46%   6 Tennessee               17 19.6,     9 Miami                    7 20.9
 44%  15 Seattle                 21 22.7,    16 New Orleans              7 24.5
 41%  24 Cleveland               20 15.7,    11 Baltimore                3 19.4
 40%  25 Jacksonville            13 15.0,    20 Buffalo                 10 19.5
 39%  30 Detroit                 20 15.6,    22 Chicago                 16 20.8
 39%  23 Washington              16 14.8,     7 Tampa Bay               10 20.0
 34%  26 Atlanta                 21 18.8,    13 San Francisco           19 27.9

  results by prob (number of games, actual/expected):
               50%-59%  60%-69%  70%-79%  80%-89%  90%-99%     100%
                8 0.90   6 0.54   2 1.32   0 0.00   0 0.00   0 0.00

  total results (num_games,num_right,expected,num_right/expected):  16   8   9.7 0.83

  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)

NOTE: exact predicted scores are not released until after game time.

Back to: Bassett Football Model Home Page

Please email comments or questions to bfm@BassettFootball.net