2011 Round 10: Things Get Interesting

Investors will be hoping that Home Sweet Home bags a hatful this weekend. If the only home team loser for the round is Gold Coast, their Portfolios will rise by over 25%. With reward though, generally comes risk, and if the results are all contrary to Investor wishes, Portfolios will fall by 30%.

Here's the week's action - and as you've probably gathered, there's a bit of it:

A Melbourne win on Friday night would secure almost one-half of the maximum possible increase for the Round, though at $4 and playing Carlton on the other side of a narrow loss to the Cats, secure's probably not the right word to use. Other sizeable increases would accrue from a Dogs victory over the Hawks, a Pies victory by 35 points or more over West Coast, or a Richmond victory by 29 points or more. 

Painful portfolio losses loom in the shape of defeats for the Pies or the Tigers, a loss by Melbourne by 29 points or more, and in a loss by the Dogs by 23 points or more.

Here's the Ready Reckoner for the week (both in the relatively new pictorial format and the old text-only format so you can use whichever you prefer):

Amongst the Head-to-Head tipsters, Carlton, Fremantle, Geelong, Richmond, Collingwood and Hawthorn are all clear favourites, and in the remaining two games, Sydney is favoured 9-4 over the Roos, and Adelaide are favoured 8-5 over the Lions.

Our Margin Predictors agree with the Head-to-Head tipsters in 6 of the games - though in one of these games they're far more solidly behind the Crows than are the Head-to-Head tipsters - outright disagreeing only in supporting the Roos over Sydney, and the Dogs over Hawthorn. In this latter game, the margin predictions are strikingly variable, ranging from a 4 point win by the Dogs to a 32 point win by the Hawks.

There's similar variability in predicted margins for the Melbourne v Carlton game where they range from about a 1 point win by the Dees to a 43 point win by the Blues.

Notwithstanding this high level of cross-Predictor variability in the margin predictions for these two games, the average absolute distances between the Margin Predictors this week is about average. Here, firstly are the distances for this week ...

... and here's how different those distances are from the all-season average distances (where positive values represent larger-than-average distances).

Amongst the higher-performing Margin Predictors the week's largest distances are those between Combo_NN_1 and Combo_NN_2 (about 13 points per game), Combo_NN_1 and Combo_7 (about 9 points per game), Bookie_9 and Combo_NN_2 (about 8 points per game), and between Bookie_9 and Combo_NN_1 (also about 8 points per game).

Lastly then, the Probability Predictors.

Notable differences between the probability assessments of the TAB Sportsbet bookmaker and our three head-to-head Probability Predictors are H2H's assessment of Melbourne as a 51% chance; ProPred's, WinPred's and H2H's assessment of the Roos as between 54% and 64% chances; and ProPred's, WinPred's and H2H's assessment of the Dogs as just over 50% chances.

So bold is H2H's assessment of the Dees it triggered the 25% rule, which limits the difference between the probability used for Kelly-staking and the bookmaker's probability to a maximum of 25%. Had this rule not been tripped, the Head-to-Head Fund would have wagered a little more on the Dees.

The Line Fund's probability predictions are near to or over 60% for three games this round. The Cats are rated 64% chances to cover the 65.5 point spread, the Pies are rated 62% chances to cover their 34.5 point spread, and the Tigers are rated 59% chances to cover their 28.5 point spread. As a consequence of these confident predictions, Investors have wagers ranging between 3% and 5% of the Line Fund on each of these three games.