2013 : Round 20 - Results

Joy at the start of the round, joy at the end; no joy in between. That was the story of Round 20 wagering for Investors.

On Friday night, though the Saints never seriously looked like troubling the competition leaders for the four points, they did stay close enough on the scoreboard to secure our line bet. Then, late on Sunday, the Dockers put their collective feet down in the 2nd half of their game against the Giants after a troublingly lacklustre 2nd term, eventually running out 113-point victors, comfortably covering the 74.5 point spread we were offering.

Between those two line betting successes - the first of which was accompanied by a losing Head-to-Head wager in the same game, and the second of which came stapled to two losing SuperMargin wagers - were three other line betting failures and a dozen further SuperMargin losses. In total, the losses outweighed the gains, and the Recommended Portfolio fell by about 4c to leave it now up by about 9c on the season.

Still, despite losses by all three Funds this week, each of them remains in profit for the season.


In what was another good round for underdogs, four of them took the competition points, simultaneously cruelling the finals chances of their opponents and the tipping performances of many of the MAFL Head-to-Head Tipsters. The best score amongst these Tipsters this week was six from nine, which was the score attained by the three WinPred-based Tipsters and by Easily Impressed II. Worst were the scores of four from nine recorded by the two Short-Term Memory Tipsters, Shadow and Ride Your Luck.

Combo_NN2's return of five from nine was enough to see it retain outright leadership, but the pursuing pack, just one tip adrift, now numbers 10.

Best amongst the Margin Predictors this week was Combo_7, whose mean absolute prediction error (MAPE) of 28.85 was over one point per game better than the next best, Bookie_9, with an MAPE of 30.1. Combo_NN2 had, comfortably the worst MAPE of all Predictors this week. In fact, its MAPE of 45.5 was the worst single-round MAPE performance of any Predictor so far this season. That goes some way, surely, to explaining the Margin Fund's 0 from 14 record for the weekend.

Combo_7's outstanding performance saw it jump two places to 3rd on the MAFL Leaderboard, the first change we've seen amongst the leading pack for some time, while ProPred_3 and ProPred_7 both suffered significant declines. Combo_NN2 held on to 7th spot despite its poor performance.

The demise of the ProPreds extended to the Head-to-Head Probability Predictor portion of the MAFL Leaderboard, where the ProPred contestant surrendered 4th place to WinPred. Bookie_LPSO continues to lead all-comers in this field.

Meantime the Line Fund algorithm turned in a mildly negative probability score for the round, its second in three weeks.


Two correct predictions from nine was the best that any of the Margin Predictors could muster this week, this result being achieved by WinPred_3, WinPred_7 and Bookie_9. Those successes moved WinPred_7 into outright 2nd in terms of correct bucket predictions so far this season. It's now on 24, just one behind ProPred_3 on 25, and one ahead of ProPred_7 and Combo_NN2 on 23.

Seven Margin Predictors have a positive ROI across the entire season based on all their predictions, while just five Margin Predictors have a positive ROI based solely on the games for which they've selected a home team win (or a draw).