2011 Round 17 Results: Just 2c Worth

Sometimes I catch myself harbouring a sense of entitlement about our football wagering, a belief that the universe somehow "owes" me and all other Investors a result to make up for what was clearly an earlier injustice. It's a nonsense notion, of course, but one I'd be willing to wager is not uncommon amongst regular gamblers.

Anyway, whatever agency you care to blame it on - the universe, the coach, the players, the heat (well it was in Darwin, after all) - Melbourne should surely have covered the spread on Saturday, an outcome that looked all but assured when they turned for home up by 39 points with the spread already covered by 2 clear goals. But the Dees managed just 1.3 in the final term to Port's 4.3, consigning our thankfully small line bet to the "win" column of the TAB Sportsbet bookmaker's ledger.

Sydney then conspired to grant Freo a 6-and-a bit goal lead at the final change before coming home with a flurry that petered out with the Swans still 2 goals short of dragging our head-to-head bet out of that same column. 

At that point Investor Portfolios were down by almost 6c on the weekend with only an out of form Roos team to save them, which the Roos duly did by summoning an 8.5 to 4.3 final term that stretched a 5-point lead at three-quarter time into a 31-point triumph.

So, after all that, Investor Portfolios jump 2c for the week.

The Portfolio price now stands at 92.4c, the highest it's been since the end of Round 9.

(My apologies, by the way, to anyone who was confused by the weekend's Ready Reckoner, which showed last week's teams with this week's zeroes for the first four games of the round. Since we had no bets on these games I didn't check the names of the teams that were showing in the table against them. Fortunately, all the details for the games in which we had stakes were correct, as was the table that appeared at the start of the blog and the pictorial representation of the wagers that appeared straight after the Ready Reckoner in tabular format.)

The Roos' upset victory was one of three for the weekend but was by no means the most surprising of them. That description definitely belongs to the Suns' win over the Tigers, which seemed all the more unlikely at the end of a first term that saw the Tigers lead 44-8, and still looked long odds at the final change by which time the Tigers lead - 8.16 to 7.9 - was less imposing but was still, surely, enough.

Actually, based on the bookmaker's prices, the Roos' victory was also less surprising than the Swans' loss to Fremantle. It was certainly more enjoyable.

With three favourites losing it's to be expected that the Head-to-Head Tipsters will struggle a bit, which they did, averaging just 4.9 from 8 for the round, only their 12th highest performance of the season. The worst performance, 3 from 8, belonged alone to Home Sweet Home, while a number of Tipsters recorded the round's best score of 6 from 8.

None of the top six Head-to-Head Tipsters scored other than 5, however, which left the top of the leaderboard unchanged. Bookie_9 and Combo_NN_1 are still tied in first, a tip ahead of BKB, with Combo_NN_2 one tip further back.

The average victory margin this week was very low and the largest victory, that of the Roos', was also a low 31 points. This served to keep Margin Predictor MAPEs for the round quite low too - the average of 21.61 points per game was the 3rd-lowest average this season, and every Margin Predictor broke 30 points per game for the round.

Best MAPE for the round went to Win_3, which came in at just 18.85 points per game, and worst MAPE went to Combo_NN_1 though it was a still-creditable 27.19 points per game. Bookie_3, which still leads the all-season table, scored 26.99 points per game, while Combo_7, which remains 2nd, scored 21.89 points per game to reduce its deficit to Bookie_3 across the season to just over 0.5 points per game.

Five Margin Predictors have sub-30 MAPEs and have now set up a gap of over 1 point per game to the remaining Predictors. It's interesting to note that these same five tipsters all have considerably lower average biases than the remaining Predictors too, ranging from Bookie_3's virtually zero bias to Combo_NN_1's 2.7 point average Home team bias. The next smallest bias amongst those Predictors not in the top 5 is 4.6 points per game, recorded by H2H_Unadj_10 and H2H_Unadj_3.

The line predictions of the Margin Predictors were generally better than chance again this week, the fourth straight week that this has been the case. Indeed, the last 3 weeks have produced the 2nd, 3rd and 4th best averages of the season, including this week's average of 4.8 from 8. Best score for the round went to Bookie_9, which scored 7 from 8, and worst was Bookie_3, which scored 2 from 8. Bookie_3 still has the best season-long performance though, which now stands at 55%.

All of the Head-to-Head Probability Predictors turned in sub-zero probability scores for the round, the TAB Bookmaker being the best amongst them. He still leads out from H2H (Adj), which narrowly heads H2H (Unadj) and ProPred, with WinPred some distance back in 5th.

The Line Fund algorithm returned to its sub-zero probability score producing ways too, hurt in particular by the Cats' inability to cover the spread against the Lions.