Following on from previous weeks, here are our mysteriously-generated NFL predictions for week 8. As before, confidence figures (probabilities basically) are given, which are also mirrored by the highlighting (>80% = white, >70% = light grey).
New this week: you will notice that I have highlighted two games in thick grey boxes. These are cases where our system actually favours the underdog to win. This is also reflected in the very high confidences (81% and 79%) that these teams are going to beat the spread. If we were in the betting mood then these picks would seem like particularly good bets, given the odds against these teams.
There are some other particularly bet-worthy picks, but I’d like to draw your attention to what is perhaps the most interesting of the bunch this week…
The red box around SF is there to indicate that there is a very strong incentive to bet against San Francisco in the match @ Jacksonville!
Why, when we think that Jacksonville is going to lose big style, with winning odds of less than 1/6?
Because the available odds against Jacksonville – as set by the bookmakers and the betting public – are only half that, at around 1/12! In other words the world is unreasonably certain that Jacksonville is going to lose, and has already put its money where its mouth is. In terms of the apparent risk/reward, Jacksonville is the most attractive bet by far this week!
It may help to think about it like this: if we bet on Jacksonville in a thousand (presumably weakly-deterministic) parallel universes, the money we’d make from those few universes in which they won would far outweigh the money we lost in all the other universes (assuming our predictions are good)!
Unfortunately we only have practical access to the one universe, and games like this don’t come along every day. So betting on this game entails unavoidable risk.