2011 Round 24: A Late Dive at the Line

Prior to running the Fund algorithms for this week I was concerned that they might both turn bet-shy which, coupled with the substantial possibility of pronounced inactivity during the weeks of the Finals, would have left Investors with a loss and no hope of mitigating it. I need not have worried.

The Head-to-Head Fund found four games to its liking this week, making two sizeable wagers on the Eagles (13% at $1.15) and the Swans (11% at $1.14), both of which I'm feeling fairly confident of landing, and a moderately sized wager on the Blues (6% at $1.65) and a 'for-interest-only' wager on the Pies (2% at $1.30), on neither of which I'd stake anything material (other than my and fellow-Investors' money). 

The Line Fund's also found four wagers, three of them on the same teams as the Head-to-Head Fund - including a bold 10% on the Pies giving 20.5 start - and the fourth, a tiddler, on the Dogs giving Freo 46.5 points start.

Any material wager on the line market exposes Portfolios to commensurately material gains and losses, so the 10% wager on the Pies is of necessity associated with a large potential gain and a large potential loss. Significant head-to-head wagers on short- and shortish-priced favourites, however, share all of the downside but much less of the upside as evidenced by the situations in the Swans v Lions and Eagles v Crows games.

A round of perfect results will add just over 12c to Portfolios moving them close to breakeven yet again, while a round of perfect disasters will knock over 25c off Portfolio prices and hasten my progress towards MAFL retirement.

If the majority Head-to-Head Tipster views prevail any such thoughts of retirement will be entirely premature, as Collingwood, Sydney and Carlton, the Dogs and the Eagles should all prove victorious, although whether they'll win by enough to secure the line bets is not a topic on which the Head-to-Head tipsters are qualified to pass an opinion.

Cue the Margin Predictors. Six of them see the Pies as covering their 20.5 points spread, while two see the Dogs covering their 46.5 points spread, twelve see the Blues covering their 6.5 points spread, and twelve see the Eagles covering their 34.5 points spread. That's sound endorsement for the two medium-sized wagers on the Blues and the Eagles, but troublingly equivocal support for the chunk of cash stapled to the back of the Pies.

Combo_NN_2 and Combo_NN_1 have chosen this week as the week to go strongly pro-home team, in Combo_NN_2's case, easily to the largest extent it has done so in any single round this season.

Since the measure of home team bias that I'm using is zero-sum in nature, Combo_NN_1's and Combo_NN_2's pro-home team bias must be offset by the collective pro-away team bias of other Predictors, which it is most notably by Bookie_3, Bookie_9, ProPred_3, WinPred_3 and WinPred_7.

On probability prediction, all Head-to-Head Probability Predictors concur on the choice of favourite for every game this week, significantly disagreeing only on the extent to which the Blues should be fancied over the Saints.

The Line Fund algorithm's rating of the Pies as 76% chances is the fourth-highest rating it has bestowed upon any home team this season and caps off a season in which it has regularly showed great confidence in the Pies' ability to win on line betting, especially when they're playing at home.

As you can see from the table at left, on average the Line Fund has rated the Pies 68% chances to win on line betting when they're playing at home. In reality they've won 80% of the time when playing at home, so this confidence has been well-placed.

Other teams that the Line Fund has, on average, rated as strong chances to win on line betting have been the Hawks when playing at home (a well-founded assessment) and St Kilda when playing at home (also well-founded).

In terms of actual line betting performances, Geelong at home have the best record (82%), followed by the Pies at home and the Hawks away (both 80%), then the Roos and St Kilda at home (both 73%). The overall mild skew towards home team wins on line betting has persisted this year, with 52% of them winning on line betting.

Converting the Line Fund's probability assessments and the actual line betting results into probability scores provides an indication of how well-calibrated the Line Fund algorithm has been in relation to each team.

As you can see from the table at right, the Line Fund algorithm has been especially well-calibrated in relation to Port Adelaide and Collingwood home games and, to a lesser extent, to Fremantle away games.

Conversely, it's been relatively poorly calibrated for Melbourne, Collingwood, Geelong and Sydney away games.

Overall, the algorithm's been best calibrated for Port Adelaide's line results and most poorly calibrated for Melbourne's, Richmond's and Sydney's line results.