Zoltar is my NFL prediction computer program. It uses a deep neural network and reinforcement learning. Here is Zoltar’s prediction for Super Bowl LIII for the 2018 NFL season, to be played on Sunday, February 3, 2019:

Zoltar:        rams  by    1  dog =    patriots    Vegas:    patriots  by    2

Zoltar theoretically suggests betting when the Vegas line is more than 3.0 points different from Zoltar’s prediction. So for the Super Bowl, Zoltar almost, but not quite, recommends placing a bet on the Vegas underdog Rams. Such a bet would pay off if the Rams win by any score, or if the Patriots win but by less than 2.0 points (in other words, exactly 1 point). If the Patriots win by exactly 2 points, then all bets are a push.

Update: I learned that the point spread opened at “Rams -1” (the Rams favored by 1 point) but in the first few minutes of betting, bettors placed ten of millions of dollars on the then-underdog Patriots. The point spread was adjusted after just one hour to “Patriots -2”. The three-point change is very large. I’m not sure exactly what this means. Anyway, I’ll be happy regardless of who wins Super Bowl LIII because I grew up watching the L.A. Rams in the 1960s, and I’m a huge admirer of the New England Patriots’ model of teamwork.

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.

Just for fun, I track how well Zoltar does when just trying to predict just which team will win a game (not by how many points). This isn’t useful except for parlay betting.

Zoltar sometimes predicts a 0-point margin of victory. When there are such games, in the first four weeks of the season, Zoltar picks the home team to win. After week number four, Zoltar uses historical data for the current season (which usually, but not always, ends up in a prediction that the home team will win).

==

Last week, against the Vegas point spread, which is what Zoltar is designed to do, Zoltar went 1-1 by correctly liking (barely) the Vegas underdog Patriots against the Chiefs. Zoltar missed by liking the Vegas favorite Saints against the Rams because the Rams won outright.

For the regular season, against the Vegas spread, Zoltar went 49-28 which is about 63% accuracy. Including the playoff weekends to date, Zoltar went 52-29 (about 64% accuracy).

Just predicting winners, Zoltar was 0-2. Vegas was also 0-2 as all both favorites, the Saints and the Chiefs lost. For the regular season, just predicting which team will win, Zoltar was 173-81 (about 68% accuracy) and Vegas was 167-85 (about 66% accuracy).

My machine learning system is named after the fortune teller machine you can find in arcades. That Zoltar is named after the machine that appeared in the 1988 movie “Big” starring Tom Hanks. And that Zoltar is named after “Zoltan”, a machine from the 1960s. Far right: fortune teller machines have been around for many decades.