NFL 2015 Week 13 Predictions – Zoltar Makes Five Recommendations

I wrote a computer program called Zoltar that predicts the results of NFL football games. For the season, through week 12, in head-to-head predictions, meaning just predicting the winner of a game (which isn’t really useful), Zoltar has gone 138-39 for 78% accuracy. Bing Predictions is 112-64 for 64% accuracy.


Zoltar’s point spread predictions for week 13 of the 2015 NFL season are:

Zoltar favorite =     packers  by    0  underdog =       lions
Zoltar favorite =      texans  by    0  underdog =       bills
Zoltar favorite =       bears  by    6  underdog = fortyniners
Zoltar favorite =     bengals  by   12  underdog =      browns
Zoltar favorite =      ravens  by    0  underdog =    dolphins
Zoltar favorite =    seahawks  by    0  underdog =     vikings
Zoltar favorite =    panthers  by    5  underdog =      saints
Zoltar favorite =      giants  by    5  underdog =        jets
Zoltar favorite =     jaguars  by    0  underdog =      titans
Zoltar favorite =   cardinals  by    5  underdog =        rams
Zoltar favorite =  buccaneers  by    4  underdog =     falcons
Zoltar favorite =      chiefs  by    4  underdog =     raiders
Zoltar favorite =     broncos  by    5  underdog =    chargers
Zoltar favorite =    patriots  by   11  underdog =      eagles
Zoltar favorite =    steelers  by    4  underdog =       colts
Zoltar favorite =     cowboys  by    0  underdog =    redskins

Notice that six of the predictions are a tie game. When this happens, just in order to calculate a head-to-head winning percentage, Zoltar uses its internal ratings to predict the winner. I used to make Zoltar predict the home team as winner when the predicted margin of victory is 0, but using the internal ratings gives more accurate predictions.

Zoltar’s predictions are close to the Las Vegas point spreads, but there are 5 games where Zoltar sees a theoretical betting opportunity:

* Zoltar says the Lions and Packers are evenly matched but Vegas says the Packers are favored by 3.5 so Zoltar recommends betting on the Vegas underdog Lions.

* Zoltar says the Bengals will win by 12 over the Browns but Vegas says the Bengals will win only by 7, so Zoltar recommends betting on the Vegas favorite Bengals.

* Zoltar says the Dolphins and Ravens are evenly matched but Vegas says the Dolphins will win by 4.5, so Zoltar recommends betting on the Vegas underdog Ravens.

* Zoltar says the Giants will win by 5 over the Jets but Vegas says the Jets will win by 2, so Zoltar recommends betting on the Vegas underdog Giants.

* Zoltar says the Redskins and Cowboys are evenly matched but Vegas says the Redskins are favored by 4 so Zoltar recommends betting on the Vegas underdog Cowboys.

Update – Monday, Dec. 7, 2015: Pending the results of tonight’s game (Cowboys at Redskins), Zoltar went 2-2 against the Las Vegas point spread (correct calls on the Bengals and Dolphins games, wrong on the Lions and Giants games). In head-to-head only (meaning just predicting the winner, which isn’t very useful), Zoltar is a mediocre 9-6 so far. Bing is better at 10-5 so far. Interestingly, Zoltar picks the Cowboys to (just barely) win but Bing picks the Redskins.

Update – Tuesday, Dec. 8, 2015:The Cowboys beat the Redskins last night, 19-16 so Zoltar’s head-to-head predictions were an OK, but not great 10-6. Bing also went 10-6 for the week. Zoltar’s final accuracy against the Vegas point spread was 3-2.

Note: I reported incorrect results for week 12. Zoltar was actually 12-4 in head-to-head predictions. (I had a bug in my program’s accuracy calculation). And I also had an incorrect result for week 11 due to a bad data file.

This entry was posted in Machine Learning. Bookmark the permalink.