Following up on Bill Simmons’ complaint about the NFC West, it went 3-1 this past weekend, moving it to 5-7 overall. That’s the same as the NFC East, and one game behind every other division except for the AFC’s North and South. So maybe we should cut them a little slack for now and see how they look in the power rankings (coming in the morning!).
On to the prediction results, starting with the moneyline. Mario only picked five games to bet on, choosing Atlanta, St. Louis, Denver, Seattle, and Chicago. Denver was the only loss and the rest were underdogs, so if you bet $100 on each game you would have come out $550 ahead. Luigi had the same picks plus Carolina, which was a loss, so it would have made $450. Since I forgot to get the away team moneyline last week, we’ll say this is the first week of real picks, and so Mario is 4-1 overall and Luigi is 4-2 overall.
For the over/under, the two models had a fair amount of disagreement. Mario went 4-12 while Luigi went 7-9; they went 1-6 on the games they agreed on. There were four games that went off at -105 for the over; Mario went 0-4 in those game, losing $400, and Luigi went 4-0 winning $364.40. So that means Mario went 4-8 in standard -110 games for an overall very bad performance of losing $836.36. Including week 2, Mario is at 12-20 for -$900. Luigi went 3-9 in standard games for an overall performance of -$262.87. On the season, Luigi is 16-16 for -$140.
Mario went 7-8 against the spread while Luigi went 7-7. The numbers don’t add up to 16 because one game, Arizona-Oakland, had a line very close to the prediction. If the two are less than a point apart, I think it’s best to skip that game (Luigi would also skip the Miami-Jets game). Again, the exact performance would depend on what pay odds you got for the games and how much you might adjust your bet size, but neither model did quite well enough to make money. On the season, Mario is 15-15-1 and Luigi is 15-14-1.
Unlike what you would do in real life, Bill Simmons ‘bets’ every game so we will as well instead of ignoring games where the spread is close to the prediction. Against his lines Bill went 10-6, Mario went 7-9, and Luigi went 6-10. On the season overall Bill is 25-20-3, Mario is 15-16-1, and Luigi is 14-17-1. The models have 16 fewer games because they didn’t pick week one, but a different model I tried for fun went 8-6-2 and I like Mario for the spread, so I would consider myself 23-22-3 and two games behind Bill. Bill doesn’t have many 10-correct weeks; I’ll catch up soon.