2016 Round 1 Results : Wholly Unembarrassing

It's a low bar, I acknowledge, but when you completely revamp your major predictive algorithms during the off-season, throwing away the predictive life-force granted you by access to bookmaker prices, I think it's reasonable to commence the new season merely hoping for an absence of profound regret.

That's certainly what we got this week from both a wagering and a prediction basis, with two of the three Funds returning a profit and with the Head-to-Head Tipsters, Margin Predictors and Head-to-Head Probability Predictors all starting the year extremely well.

WAGERING PERFORMANCE

Both the Head-to-Head and the Line Funds went 3 from 3 this weekend, the Head-to-Head Fund climbing by 7.4c as a result, and the Line Fund climbing by 5c.

The Overs/Unders Fund was the weekend's lone loser, it on the end of both a narrow victory and a narrow defeat across its four wagers, one successful by just half a point and another unsuccessful by the same margin. In total, it landed two of its four bets, which, when wagering at $1.90 prices, was enough to see it shed 0.6c on the weekend.

Combined, with the Head-to-Head and Line Funds carrying between them four times the weighting of the Overs/Unders Fund, their success had the overwhelming influence on the Overall Portfolio, it finishing up 4.8c on the round.

TIPSTER AND PREDICTOR PERFORMANCE

For the first time in as long as I can remember, all nine games were won by the Home team, leaving Home Sweet Home atop the MoS Head-to-Head Tipster Ladder I suspect for the first time ever.

C_Marg and MoSSBODS_Marg, both Tipsters that ignore bookmaker input, were next best amongst the Head-to-Head Tipsters, both tipping eight of the nine results, missing only the last game of the round. All of the remaining Tipsters except Consult The Ladder managed six from nine correct tips.

View fullsize

Amongst the Margin Predictors, C_Marg and MoSSBODS_Marg fared less well, though their Mean Absolute Errors (MAEs) of 27.8 and 29.0 points per game were still very acceptable. Best was Bookie_LPSO with an MAE of 26.4 points per game, just slightly ahead of the ENS_Linear (26.5) and ENS_Greedy (26.6) Predictors. Worst was MoSSBODS_Marg and the two ensemble predictors based on it. Even the worst Predictor, however, was only about 26 points in total more distant from the correct margins across the nine games.

On Probability Scoring, the three bookmaker-based Predictors scored best, though MoSSBODS_Prob finished essentially equal with Bookie-OE and currently lies in 4th.