2012 Round 18 Results : Small Loss, Small Comfort

A Roos goal at the siren was, ultimately, all that stood between profit and loss for Investors this weekend. That goal nudged the Roos' victory margin to 54 points, exceeding the 49-point maximum stapled to our SuperMargin wager. 

Not that Investors weren't without opportunities - on both sides of the ledger - in other contests. Geelong for a long time looked as though they'd win by a margin somewhere in the 10 to 19 point range, which would have landed the SuperMargin wager on them, while the Blues threatened to lose by more than 5 points at various points in their game with the Tigers, which would have scuppered our ultimately successful Line bet on them.

Our 1c loss was probably a poorer result than we deserved, but it's ours to keep nonetheless, and it leaves Overall Portfolios now down by 11.6c on the season. All Funds remain in red ink, notwithstanding the 3c gain by the Line Fund during Round 18 on the back of a 4 and 3 performance.

Head-to-Head Tipping was, once again this week, all too easy. Eight favourites were victorious, so a slew of Tipsters landed eight for the week dragging the all-Tipster average up to 7.7. Worst Tipster honours for the week went to Home Sweet Home, which predicted only five results correctly. The lack of differentiation in the predictions of the Tipsters at the top of the Leaderboard translated, as a matter of logic, into a lack of differentiation in their performances, leaving us with H2H_U10 as our continuing leader, now on 120 from 153 (78%), still one tip ahead of a baying pack of well-credentialled if unadventurous followers on 119.

If you wanted to answer the question "Who won?" this week, the phrase "favourites did" would have sufficed for all but one of the contests, and if this initial inquiry was followed up with the question "By how much?", the equally succinct phrase "a lot" would have been good enough for six of those games. 

Big margins, generally, are not conducive to stellar results for the Margins Predictors, which is why this week saw an all-Predictor MAPE of 40.85 points per game. The round's best result was down to Bookie_3, which recorded an MAPE of just 29.24 points per game, a result so strong that it propelled it into top spot on the MAFL Leaderboard, relegating Combo_7 to 2nd. The round's worst MAPE was attributable to ProPred_7, whose predictions produced a horror MAPE of 48.78 points per game.

These poor MAPEs translated into generally poor line betting performances too, the all-Predictor average this week coming in at just 2.5 from 9. Bookie_3 though, in keeping with its strong MAPE result, correctly picked seven of the nine line market results.

The large and mostly larger-than-expected victory margins meant that the Margin Predictors also fared poorly on SuperMargin wagering, with only Bookie_3 (of course) and Combo_NN1 selecting the correct bucket in any of the nine matches. Regardless, the H2H-based algorithms still show a season-long profit on SuperMargin wagering.

All Probability Predictors, including the Line Fund algorithm, registered positive probability scores this week. Best amongst the Head-to-Head Probability Predictors was the TAB Bookmaker, but he was only a fraction better than H2H. The Line Fund algorithm's result for the round was its 2nd-best all season.