Well that was painful to watch. In three games this weekend the team registering more scoring shots failed to secure the competition points. Investors had money on the loser in all three of them.

On Friday night, the Blues kicked - and I use that verb loosely - 2.9 in the final term only to fall one-and-a-half points short of securing our line bet, in the process recording three more scoring shots than did their opponents, the Cats. Next up, on Saturday afternoon, the Tigers kicked 9.12 for the match (and 4.10 for the last three terms) to Fremantle's 12.6 to lose by 12 points, never looking as though they'd cover the 30.5 points they were offering in the line market. 

Then, lastly, on Saturday evening the Dons racked up a farcical 2.15 in the first three terms of their game against the Swans, only to experience an epiphany minutes into the final term as the distinction between the large and the small posts suddenly became clear, resulting in a 9.1 quarter that left them 5 agonising points short of the greatest final term come-from-behind win in the history of the sport. Far more importantly, it also left Investors 5 points point short of landing their SuperMargin wager and 14 points short of landing their Line bet.

On Sunday, Port Adelaide merely provided the requisite sodium chloride icing. 

All told, Overall Portfolios dropped 13.5c on the round to leave them down by 9.2c on the season.

The six winners this week were Away teams, which was shocking news for Home Sweet Home, but only two of them were underdogs (the Dockers and the Swans) so the average Head-to-Head Tipster still managed to bag 3.9 out of 6. Four tipsters managed as many as 5, while Home Sweet Home registered its first duck egg of the season and ending a four-round streak during which it averaged almost 70%. None of this changed anything much at the head of the MAFL Leaderboard however, which still has 11 joint leaders, all now on 73/96 (76%).

Margin prediction proved relatively easy too, the all-Predictor average for the round coming in at 28.13 points per game, with the best result being turned in by HU_3 (25.90 points per game) and the worst by ProPred_7 (36.51 points per game). Bookie_9 still has the best season-long record, which now stands at 28.59 points per game, ahead of Bookie_3 on 28.98, and Combo_7 on 29.00.

Collectively, in a mass display of regression to the mean, the Margin Predictors turned in sub-chance results on line betting for the second successive week, averaging just 2.7 correct predictions from 6. They're still all above 50% for the season though. 

My mood about our wagering fortunes wasn't at all improved when I reviewed the SuperMargin performance of all Margin Predictors for the weekend only to discover that Combo_NN2, the Predictor whose predictions drive the Margin Fund, nailed the results for the Blues v Cats and Melbourne v Collingwood games - both matches in which the prediction was for an Away team victory and so not acted upon by the Margin Fund.

Had I exhibited more faith in CNN_2's abilities and backed them in every game this season, not just in those where it tipped a Home team win or draw, the ROI for the Margin Fund would now stand at -1% rather than -16%.

Probability scoring was poor for all Probability Predictors this weekend except the TAB Bookmaker, who registered the only positive score for the round and so extended his lead amongst the Head-to-Head Probability Predictors.

The Line Fund algorithm also turned in a negative probability score for the round.