Predicting the results of NFL football games is a fascinating problem. I wrote a computer program called Zoltar (named after the arcade fortune telling machine) to do just that. Unlike many prediction programs that use classical statistics techniques, Zoltar uses an information-theoretic approach.

Zoltar’s point spread predictions for the second round of the playoffs (week 19) of the 2015 NFL season are:

Zoltar favorite =    patriots  by    2  underdog =      chiefs
Zoltar favorite =   cardinals  by    6  underdog =     packers
Zoltar favorite =    panthers  by    4  underdog =    seahawks
Zoltar favorite =     broncos  by    5  underdog =    steelers

Here is are the Las Vegas point spreads:

Vegas favorite =    patriots  by    5   underdog =      chiefs
Vegas favorite =   cardinals  by    7   underdog =     packers
Vegas favorite =    panthers  by    3   underdog =    seahawks
Vegas favorite =     broncos  by    6.5 underdog =    steelers

So, Zoltar agrees very closely with the Las Vegas predictions and doesn’t see any good theoretical betting opportunities.

For last week’s predictions, Zoltar was correct in three out of four games when predicting just the winner — Zoltar correctly predicted the Chiefs, Seahawks, and Packers would win but incorrectly predicted the Bengals would beat the Steelers.

Against the Vegas point spread, Zoltar was a perfect 2-0, correctly advising bets on Vegas underdogs Vikings and Bengals, believing that the favorites (Seahawks and Steelers respectively) wouldn’t cover the point spread. The Seahawk won by only 1 point and the Steelers won by only 2 points.

For the 2015 season, through week 18 (first round of playoffs), in head-to-head predictions, meaning just predicting the winner of a game (which isn’t particularly useful except for parlay betting), Zoltar has gone 188-72 for 72% accuracy. I couldn’t find any week 18 predictions for Bing Predictions. For the regular season Bing Predictions was 161-95 for 63% accuracy.

If you just picked the Las Vegas point spread favorite to win each game, for the regular season you’d have gone 153 correct and 91 wrong for 63% accuracy.

Update, Monday, Jan. 18, 2016: In head-to-head predictions (meaning just predicting the winner of a game, not the point spread difference), Zoltar was 4-0 in week 19. The Patriots, Cardinals, Panthers, and Broncos all won (beating the Chiefs, Packers, Seahawks and Steelers respectively). For the year, through week 19, in head to head picks, Zoltar is 192-72 for 73% accuracy. As far as I can tell, Bing Predictions did not make any predictions for the playoffs (weeks 18 and 19) and finished the regular season at 161-95 for 63% accuracy.