2014 - Round 10 : Just Lines and Margins

With four of the home teams short-priced favourites this week, the Head-to-Head Fund had only two decisions to make and, in both cases, it decided "no". Neither the Giants' $3.85 price tag nor the Blues' $2.15 offering were deemed sufficiently generous to merit a wager.

The Line Fund also passed on these two games, and on two more, eventually venturing just two wagers, one on the Cats giving the Roos 14.5 points start and the other on Port giving the Hawks 15.5 points start. The Margin Fund has two wagers on each of the Cats and Port Adelaide too, as well as two more on the Pies and two on the Suns.

All of which means that we've just 4.5% of the Portfolio at risk this weekend - the least we've had active since Round 2 - a relatively high proportion of it attached to just two teams, Geelong and Port Adelaide.

Despite the high concentration of wagering activity on these two teams, it's Collingwood and the Gold Coast who carry the greatest upside. Wins by 20 to 29 points by either team will see the Recommended Portfolio grow by 3c. The Cats and Port then represent the remainder of the upside, each promising 2.4c gains with propitious outcomes - in the Cats' case a win by 15 to 29 points, and in Ports' a win by 16 to 19 points.

Geelong and Port Adelaide do carry the largest and equivalent downside of 1.8c however, while Collingwood and Gold Coast can do no more damage than losing 0.5c each.

The range of possible outcomes this week spans from a gain of almost 11c to a loss of almost 5c.


Richmond, Port Adelaide and the Gold Coast are all either unanimous or near-unanimous selections for the Head-to-Head Tipsters this week, Home Sweet Home the only Tipster plumping for the Giants, and C_Marg alone in choosing the Hawks.

Both the Cats and the Pies enjoy 3:1 majorities in their respective games, in each case enjoying supporters of, on average, superior relative calibre.

In the sixth and final game, Carlton v Adelaide, the Crows enjoy 18-10 favouritism. They're also the favourite with the TAB.

Yet again this week, despite high levels of general agreement amongst the Head-to-Head Tipsters, there are even higher levels amongst the Margin Predictors. The Cats, Giants, Pies and Suns are all unanimous selections, and the Hawks are favoured by C_Marg alone (probably no great thing for them given C_Marg's season so far), leaving only the Blues v Crows game where there's any significant uncertainty.

Courtesy mostly C_Marg, the Port Adelaide v Hawthorn game carries the greatest range and highest standard deviation of the associated margin predictions. The Cats, Giants and Pies games are the next three in terms of the range of predictions, the Predictors in these games offering margins spanning about 3 to 4 goals. For the Suns and the Blues games the range is only about a goal and a half.

Notwithstanding its obvious eccentricities, C_Marg has managed to be high or low Margin Predictor only in that Port Adelaide v Hawthorn game, leaving Pro_Pred_7 and Bookie_3 to assume that position in three games each, Combo_NN_1 and Combo_NN_2 to do the same in two games more, and Win_7 to snap up the sole remaining opportunity.

As you might expect, being low or high Predictor significantly enhances a Predictor's chances of being nearest to the actual result. So far this season, in 55 of 75 games the Predictor nearest to the actual result has been the low or high Predictor. As a matter of mathematical certainty however, either the low or the high Predictor must be the furthest from the actual result in every game, so it's a high-reward, high-risk strategy to assume either outlying position.


It's also only in the Port v Hawks and Blues v Crows games that the Probability Predictors break ranks, C_Prob, of course resolutely and solely behind the Hawks in that first game, and the Predictors split 5-3 in favour of the Crows in that second contest. In the four other contests the Predictors are all siding with the TAB favourite.

There are no 90% probability assessments from the Line Fund algorithm this week, but there is a 68% assessment for Port Adelaide, a 60% assessment for West Coast, and a 59% assessment for Geelong, making possible another round of high probability scoring for the algorithm. The three other games the algorithm pretty much rates as coin tosses.


For five of the six games ChiPS has ultimately selected as the likely winner the team with the higher base Rating, and in four of those games that selection coincides with the TAB favourite. The exception is the Port Adelaide v Hawthorn game where ChiPS, as already discussed, fancies the Hawks, but the TAB rates Port Adelaide about 65 to 70% favourites.

The only game in which the combined benefits of Form Difference, Home Ground Advantage and Interstate State allow the team with the lower base Rating to nonetheless be tipped by ChiPS to win is the Carlton v Adelaide game where ChiPS foresees a very narrow Carlton victory.

After converting the margin predictions to probabilities, no team is assessed by ChiPS as representing value at its current TAB price.