NFL 2015 Week 15 Predictions – Zoltar Looks for Redemption

I wrote a computer program called Zoltar that predicts the results of NFL football games. Unlike many prediction programs that use classical statistics techniques, Zoltar uses an information-theoretic approach.

Last week Zoltar had his worst week of the season. Zoltar is annoyed and seeks redemption. Zoltar’s point spread predictions for week 15 of the 2015 NFL season are:

Zoltar favorite =        rams  by    2  underdog =  buccaneers
Zoltar favorite =     cowboys  by    4  underdog =        jets
Zoltar favorite =       colts  by    4  underdog =      texans
Zoltar favorite =     falcons  by    0  underdog =     jaguars
Zoltar favorite =     vikings  by    7  underdog =       bears
Zoltar favorite =    patriots  by   18  underdog =      titans
Zoltar favorite =    panthers  by    5  underdog =      giants
Zoltar favorite =      chiefs  by    3  underdog =      ravens
Zoltar favorite =       bills  by    0  underdog =    redskins
Zoltar favorite =     packers  by    5  underdog =     raiders
Zoltar favorite =    seahawks  by   16  underdog =      browns
Zoltar favorite =     broncos  by    0  underdog =    steelers
Zoltar favorite =    chargers  by    1  underdog =    dolphins
Zoltar favorite =     bengals  by    5  underdog = fortyniners
Zoltar favorite =   cardinals  by    4  underdog =      eagles
Zoltar favorite =       lions  by    0  underdog =      saints

Notice that Zoltar predicts a tie for some games. When this happens, just in order to calculate a head-to-head winning percentage, Zoltar predicts the team with the higher internal rating (Zoltar computes ratings for each team) will win.

Zoltar’s predictions are close to the Las Vegas point spreads, but there are 6 games where Zoltar sees a theoretical betting opportunity:

* Zoltar says the Cowboys will beat the Jets by 4 points but Vegas says the Jets will win by 3 points so Zoltar would recommend betting on the Vegas underdog Cowboys.

* Zoltar says he Colts will beat the Texans by 4 points but Vegas says this is a “pick” game so Zoltar would recommend betting on the Colts.

* Zoltar says the Falcons and Jaguars are evenly matched but Vegas says the Jaguars will win by 3.5 points so Zoltar would recommend betting on the Vegas underdog Falcons.

* Zoltar says the Patriots will beat the Titans by a whopping 18 points but Vegas says the Patriots will win “only” by 14 points so Zoltar would recommend betting on the Vegas favorite Patriots to cover the spread.

* Zoltar says the Chiefs will beat the Ravens by only 3 points but Vegas says the Chefs will win by 7.5 points so Zoltar would recommend betting on the Vegas underdog Ravens.

* Zoltar says the Broncos and Steelers are evenly matched but Vegas says the Steelers will win by 6 so Zoltar would recommend betting on the Vegas underdog Broncos.

ZoltarVsBingThruWeek14

For the season, through week 14, in head-to-head predictions, meaning just predicting the winner of a game (which isn’t particularly useful), Zoltar has gone 152-56 for 73% accuracy. Bing Predictions is 131-77 for 63% accuracy. By the way, if you simply picked the home team to win, you’d only get about 55% accuracy.

Update: Tuesday morning, 12/22/2015: In week 15, Zoltar went a decent 12-4 in head-to-head predictions (meaning just predicting the winner, which isn’t hugely useful except in parlay bets). Bing Predictions also did pretty well at 11-5.

Against the Vegas point spread, Zoltar went a so-so 3-2 (note that the Jets vs. Cowboys game was a push because the Jets won by exactly 3 points. Also, I had an incorrect Vegas line on the Colts vs. Texans game — Vegas actually had the Texans favored by 7 points).

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