2014 - Round 9 Results: The Smallest of Losses

/

It's disappointing when a round that starts so well, finishes so poorly. Thursday night's victory by the Crows landed our head-to-head and line bets and had us 2.2c closer to profitability, but just a single additional collect throughout the remainder of the extended round saw us hand all of the 2.2c back, with a 0.1c tip toppling us just over the edge from break-even into loss.

All told, the Head-to-Head Fund shed 0.3c on account of its 1 from 2 record, the Margin Fund shed 5c from its 0 and 4 record, while the Line Fund recorded the weekend's only profit, landing 2 from 3 wagers to rise by 2c. Totting up those gains and losses and weighting them appropriately produces a loss for the Recommended Portfolio of 0.1c, leaving it now down by 3.7c on the season.

The Line Fund really is the bright spot at the moment, its 8 and 4 record over the past four rounds giving much hope as we near Round 12, which is when this Fund doubles its bet size. There are some encouraging signs for the Head-to-Head Fund too, it having collected in three of its five most-recent wagers. It will also double its bet size soon, but not until a round later than the Line Fund, in Round 13.

TIPS AND PREDICTIONS

The average Head-to-Head Tipster bagged 3.4 from 6 this week, and most scored 4, despite the fact that only 3 favourites and an equal-favourite were victorious. (BKB scored only 3 because it's assigned CTL's tip when the TAB Bookmaker has equal-favourites and CTL tipped Geelong to beat Fremantle in the relevant game.)

Bookie_9 remains the clear leader of the Tipsters with a score of 52 from 75 (69%), while Combo_7 has broken away to take outright second position, one tip back on 51 from 75 (68%). Bookie_3 has outright third on 50.5 from 75 (67%). Unlike BKB, it's allowed to tip draws, and so picked up a half game in the Freo v Cats contest this week.

The average Margin Predictor recorded a Mean Absolute Prediction Error (MAPE) of 34.5 points per game this week, which was about 2.5 points per game higher than the round's best result of 32.0 points per game, recorded by Bookie_3. That result elevated Bookie_3 into second place on the Leaderboard where it now sits behind Combo_7, who's held the top spot on the Ladder since Round 6.

Bookie_LPSO recorded the round's third-lowest score (ProPred_3 was second-lowest) and in so doing climbed four places into seventh this week, though it's only 1.6 points in total ahead of the next three Predictors, so it's a precipitous fourth at best.

C_Marg produced the worst MAPE of the round at 38.2 points per game, and now finds itself over 10 goals behind the Predictor in second-last position, ProPred_7. C_Marg is still, however, one of the few Predictors with a profitable season-long line betting record.

And speaking of profitable records and the Margin Predictors, only one Predictor managed to select the correct bucket in any of the six contests this week - Bookie_3, who waited until the last game of the round to make its one and only correct selection.

That result wasn't sufficient to drag it into profitability for the season however, a status that's still enjoyed only by Combo_NN_1, and even then this is true only of its home team win or draw predictions.

Returning then to the MatterOfStats Tipster Leaderboard, we'll look next at the Head-to-Head Probability Predictors where we find that Bookie-OE, Bookie-RE and Bookie-LPSO remain in the first three positions, ahead of C_Prob in fourth. Win_Pred, the two H2H Predictors and Pro_Pred still occupy the bottom four places on the Leaderboard, despite being the only Head-to-Head Predictors to generate positive probability scores for the round.

The Line Fund algorithm also recorded a positive probability score this week, its fourth in succession, its fifth in seven weeks, and its seventh of the season. Had we been Kelly-staking on the basis of the algorithm's probability assessments, and had we been allowing ourselves to wager on home and on away teams, our ROI for the season to date would be 39.4%.