Zoltar is my NFL football prediction computer program. It uses reinforcement learning and a neural network. Here are Zoltar’s predictions for week #10 of the 2022 season.
Zoltar: falcons by 0 dog = panthers Vegas: falcons by 3 Zoltar: buccaneers by 3 dog = seahawks Vegas: buccaneers by 2.5 Zoltar: bills by 6 dog = vikings Vegas: bills by 6 Zoltar: bears by 6 dog = lions Vegas: bears by 3 Zoltar: chiefs by 10 dog = jaguars Vegas: chiefs by 9.5 Zoltar: dolphins by 4 dog = browns Vegas: dolphins by 4 Zoltar: giants by 2 dog = texans Vegas: giants by 6.5 Zoltar: titans by 6 dog = broncos Vegas: titans by 3 Zoltar: steelers by 4 dog = saints Vegas: saints by 2.5 Zoltar: raiders by 5 dog = colts Vegas: raiders by 6 Zoltar: packers by 4 dog = cowboys Vegas: cowboys by 5.5 Zoltar: rams by 5 dog = cardinals Vegas: rams by 3 Zoltar: fortyniners by 5 dog = chargers Vegas: fortyniners by 7 Zoltar: eagles by 5 dog = commanders Vegas: eagles by 10.5
Zoltar theoretically suggests betting when the Vegas line is “significantly” different from Zoltar’s prediction. For this season I’ve been using a threshold of 4 points difference but in some previous seasons I used 3 points.
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 #10 Zoltar likes four Vegas underdogs:
1. Zoltar likes Vegas underdog Texans against the Giants.
2. Zoltar likes Vegas underdog Steelers against the Saints.
3. Zoltar likes Vegas underdog Packers against the Cowboys.
4. Zoltar likes Vegas underdog Commanders against the Eagles.
For example, a bet on the underdog Texans against the Giants will pay off if the Texans win by any score, or if the favored Giants win but by less than 6.5 points (i.e., 6 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 eliminate pushes.
This is the part of the season where injuries start having a big effect on the point spread. I’m surprised that Zoltar thinks the Green Bay Packers are better than the Dallas Cowboys. To the human eye, the Packers have looked terrible and the Cowboys look like champion contenders. We’ll see.
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 #9, against the Vegas point spread, Zoltar went 5-1 (using 4.0 points as the advice threshold). Quite lucky because several big Vegas favorites won easily but didn’t quite cover the spread.
For the season, against the spread, Zoltar is 33-16 (~67% 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 #9, just predicting the winning team, Zoltar went only 6-7 which is terrible. Vegas was much better, going 9-4 at just predicting the winning team.
Zoltar sometimes predicts a 0-point margin of victory. There is one such game in week #10 (Falcons vs. Panthers). 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 prediction system is math-based in the sense that it computes a numeric rating for each team and then uses ratings to compute the predicted margin of victory of the better team. An entirely different approach for predicting NFL football scores is to use a sophisticated simulation program and then simulate a game many thousands of times.
Left: There are many versions of simple football simulation games where a player rolls two ordinary dice and the result is based on the 21 (order doesn’t matter) or 36 (order matters) possible outcomes. This one is called simply “Football Board Game” by The Blue Crab company of Sunrise, Florida.
Center: “1st and Goal” by R and R Games is a fairly sophisticated game that uses several kinds of specialized dice and cards.
Right: “Half-Time Football” by Lakeside was produced in the late 1970s and early 80s. It’s strictly a dice game and is medium complexity.