2011 Round 14: Commencing The Slow Climb?

Eight bets this week, five from the Head-to-Head Fund and three from the Line Fund. Four of the Head-to-Head bets are on short-priced favourites and the fifth is for only about 2% of the Fund on Port Adelaide at $2.65 in the last game of the Round. The largest wager is for nearly 15% of the Fund and is on the Round's shortest-priced team, the Cats, at $1.04. In total the Head-to-Head Fund has wagered a little over one-third of the Fund, mostly on the wagering equivalent of blue chip stocks though with decidedly more of a post-GFC than a pre-GFC risk of default.

Two of the Line Fund's wagers are on the same teams as those wagered on by the Head-to-Head Fund - Fremantle and Geelong - which both consolidates Investors'  risk and leaves then without a diversifying wager in two of the fixtures. The Line Fund's big bet is on Richmond giving Melbourne 2.5 points start and priced at $2; it's for almost 8% of the Fund. Combined the remaining wagers are for only just over 2% more.

Here's the detail:

A Richmond victory by 3 points or more is the only result promising Investors a gain of more than 2c in a single game, while wins by Port or Carlton, or by Fremantle by 34 points or more each offer another cent or two of upside. Favourable results from the Hawks or the Cats would also produce positive returns albeit small ones.

Wins - or even draws - by Adelaide, Sydney, the Lions or Melbourne would make profitability an unlikely outcome across the weekend, while Roos or Dons victories would also be unpleasant though not profit-preventing.

Amongst the Head-to-Head Tipsters only Home Sweet Home (twice) and Easily Impressed (once) prevented there from being unanimity in all but the Richmond v Melbourne clash this week.

The Margin Predictors are in fact unanimous in all but the Richmond v Melbourne game where the majority favour the Tigers 7.5 to 5.5. In that game, however, Combo_7 is tipping the Tigers by just five ten-thousandths of a point - which makes splitting hairs look like cleaving mountains - and five more tipsters are predicting margins of less than half a point. In aggregate, the average predicted margin is just half a point and the median is at 0 points. I'd call that a split result.

This week I'm offering a new way to look at the difference between each Margin Predictor's margin predictions and those of the herd. To create the following table I calculated a standardised difference (by subtracting the all-Predictor mean predicted margin from the Predictor's predicted margin and dividing by the all-Predictor standard deviation of the predicted margins for that game) for each Predictor for each game and then averaged these standardised differences for each Predictor for each round. As you do.

What the standardisation does is measure the difference between a Margin Predictor's margin prediction and the all-Predictor average in units of standard deviations rather than in raw points. This serves to lessen the influence of games where there's a lot of difference amongst the Margin Predictors which might otherwise tend to dominate the average difference calculation.

A Predictor with a negative score for a round tended to tip a smaller (standardised) victory margin for Home teams in that round than did the average Predictor, and vice versa.

What's unusual about the predictions for Round 14 is that Bookie_3, which has generally tipped smaller Home team victory margins this season than other Predictors, has been even more Home team averse. The same can be said to a lesser extent for Bookie_9. Another notable departure from the norm is the margin predicting of Combo_NN_2, which this week is relatively pro Home team in sharp contrast with its usual habit of being mildly to strongly pro Away team.

Finally, to the Probability Predictors, where there's not a lot to comment on this week. Only for the Richmond v Melbourne contest - seemingly the only source of significant disagreement this week - do these Predictors think differently about which team will win and even then the difference is very slight with H2H having the Dees as tiny favourites.

In the Port Adelaide v Kangaroos game ProPred has Port as 49% chances, which is as close as we come to disagreement in any other game.