Zoltar is my NFL football prediction computer program. It uses reinforcement learning and a neural network. Here are Zoltar’s predictions for week #7 of the 2022 season.

Zoltar:   cardinals  by    6  dog =      saints    Vegas:   cardinals  by  1.5
Zoltar:  buccaneers  by    3  dog =    panthers    Vegas:  buccaneers  by 10.5
Zoltar:     bengals  by    6  dog =     falcons    Vegas:     bengals  by    6
Zoltar:     cowboys  by   10  dog =       lions    Vegas:     cowboys  by    7
Zoltar:     jaguars  by    1  dog =      giants    Vegas:     jaguars  by    3
Zoltar:      titans  by    6  dog =       colts    Vegas:      titans  by  2.5
Zoltar:      ravens  by    4  dog =      browns    Vegas:      ravens  by    7
Zoltar:     packers  by    0  dog =  commanders    Vegas:     packers  by  5.5
Zoltar:     broncos  by    6  dog =        jets    Vegas:     broncos  by    3
Zoltar:     raiders  by    7  dog =      texans    Vegas:     raiders  by    7
Zoltar:    chargers  by    6  dog =    seahawks    Vegas:    chargers  by    7
Zoltar:      chiefs  by    0  dog = fortyniners    Vegas:      chiefs  by    3
Zoltar:    dolphins  by    2  dog =    steelers    Vegas:    dolphins  by    7
Zoltar:    patriots  by    6  dog =       bears    Vegas:    patriots  by  7.5

Zoltar theoretically suggests betting when the Vegas line is “significantly” different from Zoltar’s prediction. In mid-season I sometimes use 3.0 points difference but for the first few weeks of the season I am a bit more conservative and use 4.0 points difference as the advice threshold criterion.

At the beginning of the season, because of Zoltar’s initialization (all teams regress to an average power rating) and other algorithms, Zoltar is very strongly biased towards Vegas underdogs. I probably need to fix this. For week #7 Zoltar likes three Vegas underdogs and one favorite:

1. Zoltar likes Vegas favorite Cardinals over the Saints.
2. Zoltar likes Vegas underdog Panthers against the Buccaneers.
3. Zoltar likes Vegas underdog Commanders against the Packers.
4. Zoltar likes Vegas underdog Steelers against the Dolphins.

For example, a bet on the underdog Panthers against the Buccaneers will pay off if the Panthers win by any score, or if the favored Buccaneers win but by less than 10.5 points (in other words, by 10 points or less). If a favored team wins by exactly the point spread, the wager is a push. This is why point spreads often have a 0.5 added — called “the hook” — to prevent pushes.

This is an interesting part of the season for Zoltar. Zoltar has enough data to start getting more accurate, but humans have a lot of information too, including injuries, team chemistry, and so on.

Theoretically, if you must bet $110 to win $100 (typical in Vegas) then you’ll make money if you predict at 53% accuracy or better. But realistically, you need to predict at 60% accuracy or better.

In week #6, against the Vegas point spread, Zoltar went a so-so 4-3 (using 4.0 points as the advice threshold).

For the season, against the spread, Zoltar is 22-12 (~64% accuracy).

Just for fun, I track how well Zoltar does when just trying to predict just which team will win a game. This isn’t useful except for parlay betting. In week #6, just predicting the winning team, Zoltar went only 8-6 which is not very good at all — barely better than a coin flip. Vegas was 7-7 at just predicting the winning team.

Zoltar sometimes predicts a 0-point margin of victory. There are two such games in week #7. In those situations, to pick a winner (only so I can track raw number of correct predictions) in the first few weeks of the season, Zoltar picks the home team to win. After that, Zoltar uses his algorithms to pick a winner.

My Zoltar system evaluates football teams based on statistics, not uniforms. All NFL teams have “throwback” uniforms that pay homage to the early days of the league. Left: Pittsburgh Steelers (originally a steel mill company team). Center: Los Angeles Rams (first owner had a connection to the Fordham University Rams). Right: Green Bay Packers (originally a meat packing company team).