Zoltar is my NFL football prediction computer program. It uses reinforcement learning and a neural network. Here are Zoltar’s predictions for week #5 of the 2021 season. These predictions are semi-tentative, in the sense that it usually takes Zoltar about four weeks to hit his stride.

Zoltar:    seahawks  by    2  dog =        rams    Vegas:        rams  by    1
Zoltar:     falcons  by    4  dog =        jets    Vegas:     falcons  by  4.5
Zoltar:    panthers  by    6  dog =      eagles    Vegas:    panthers  by  3.5
Zoltar:     packers  by    0  dog =     bengals    Vegas:     packers  by  3.5
Zoltar:    patriots  by    0  dog =      texans    Vegas:    patriots  by  6.5
Zoltar:      titans  by    4  dog =     jaguars    Vegas:      titans  by  7.5
Zoltar:     vikings  by    6  dog =       lions    Vegas:     vikings  by  8.5
Zoltar:    steelers  by    5  dog =     broncos    Vegas:     broncos  by  1.5
Zoltar:  buccaneers  by    6  dog =    dolphins    Vegas:  buccaneers  by   10
Zoltar:      saints  by    0  dog =    redskins    Vegas:      saints  by    1
Zoltar:     raiders  by    6  dog =       bears    Vegas:     raiders  by  7.5
Zoltar:      browns  by    0  dog =    chargers    Vegas:    chargers  by  1.5
Zoltar:   cardinals  by    6  dog = fortyniners    Vegas:   cardinals  by  2.5
Zoltar:     cowboys  by    6  dog =      giants    Vegas:     cowboys  by  7.5
Zoltar:      chiefs  by    1  dog =       bills    Vegas:      chiefs  by  3.5
Zoltar:      ravens  by    6  dog =       colts    Vegas:      ravens  by  6.5

Zoltar theoretically suggests betting when the Vegas line is “significantly” different from Zoltar’s prediction. In mid-season I use 3.0 points difference but for the first few weeks of the season I go 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 much too strongly biased towards Vegas underdogs. I need to fix this.

1. Zoltar likes Vegas underdog Bengals against Packers
2. Zoltar likes Vegas underdog Texans against Patriots
3. Zoltar likes Vegas underdog Jaguars against Titans
4. Zoltar likes Vegas underdog Steelers against Broncos
5. Zoltar likes Vegas underdog Dolphins against Buccaneers
6. Zoltar likes Vegas favorite Cardinals over 49ers

For example, a bet on the underdog Bengals against the Packers will pay off if the Bengals win by any score, or if the favored Packers win but by less than the point spread of 3.5 points (in other words, win by 3 points or less).

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 #4, against the Vegas point spread, Zoltar went 3-3 (using 3.0 points as the advice threshold).

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 #4, just predicting the winning team, Zoltar went 11-5 which is mediocre.

In week #4, just predicting the winning team, Vegas — “the wisdom of the crowd” — went 10-6.

Zoltar sometimes predicts a 0-point margin of victory. There are four such games in week #5. 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 system is named after the fortune teller machine you can find in arcades. Two nice classical-style paintings of fortune tellers. Center: “The Egyptian Fortune Teller” by Rudoplh Ernst (1854-1932). Right: “The Fortune Teller” by Edouard F. W. Richter (1844-1913).