So, after 162 games and one of the most extraordinary seasons of all time for reasons unrelated to football, our three winners are:

Head-to-Head Tipster: MoSHPlay_Marg (116 correct tips, which is an accuracy of 72%)

Margin Predictor: Bookie_9 (MAE of 22.81 points per game)

Head-to-Head Probability Predictor: Bookie_LPSO (Average Log Probability Score of 0.1500 bits per game)

If ever you wanted confirmation of my contention that there’s no such thing as an all-around good model, take note of the fact that MoSHPlay finished 1st as a head-to-head tipster, 2nd as a probability estimator, but 10th as a margin predictor.

Overall, although the MoS trio fell towards the bottom of the Margin Predictor Leaderboard this week, given the relatively small gap between their season-long performances and those of the Predictors nearer the top, I’m happy with their performance. I will do some tweaking in the offseason of the methodology they use to estimate Venue Performance Values, and also look at how MoSHPlay combines team and player information, but I don’t feel as though any major overhaul is required.

In the end, only about one-third of a point per game separated first from last on the Margin Predictor Leaderboard.

WAGERS

One bet, one loss again for Investors this weekend.

So, the Combined Portfolio finishes the season with a price of $1.145, this result coming from a 13.8% ROT on a 1.054 turn of the Portfolio Funds.

Though it always feels a lot better to end the year with some collects, if you’d offered me that final result back in March, there’s not doubt that I would have taken it.