2014 - Round 26 : Semi-Interested

With two short-priced Home team favourites this week, Investors were never going to face the same level of wagering activity that they endured or enjoyed (you choose which) last weekend.

The Head-to-Head Fund was this week prohibited from wagering by its $1.50 minimum price rule and so is bereft of wagers, while the Line Fund thinks that Hawthorn is being asked to give too much start (20.5 points) to justify a wager but that the Swans giving 27.5 points start represent value. And then, of course, the Margin Fund remains as oblivious as ever, this week wagering on both games and presenting us with twin Chasms. At least we're getting an unusually high $8..50 price tag on two of those bets.

In aggregate, the five wagers we have represent less than 1% of the original Recommended Portfolio, the smallest proportion of the Portfolio we've had at risk in any single round this season.

A win by Sydney by a margin of anywhere between 50 and 59 points would see the week's largest possible gain of 1.3c while a win by 19 points or fewer would materialise the week's largest possible loss of 0.6c.

In the other game, anything but a Hawks win by between 20 and 29, or between 40 and 49 points will result in a loss of 0.3c.


There's no surprise at all in the unanimity of the Head-to-Head Tipsters this week, the consequences of which are that there'll be no changes on the MoS Leaderboard at the end of it.

The Margin Predictors, though equally resolute in their opinions about the more probable victor in both games, are again surprisingly broad in their estimations of the likely sizes of those victories.

In the Swans v Roos game the range of their opinions spans over 6 goals, from C_Marg's low of an 18 point victory for the Minor Premiers to Combo_NN_2's 55 point prediction for victory by the same team.

The range of predictions in the Hawks v Port game is almost as large, coming in just 0.3 points under 6 goals. In that game, Combo_NN_1 has the smallest predicted victory margin of 12 points while Combo_NN_2 here too sets the high-water mark at 48 points.

In the broader context of season-long Mean Absolute Prediction Errors (MAPEs), the key aspects to note from these predictions are that Combo_7's and Bookie_LPSO's margins differ, in aggregate and absolute terms, by about 7 points, which is about one-half the current difference in their overall MAPEs. So if, say, Sydney were to win by 5 goals and the Hawks by 4 goals, Bookie_LPSO would be within about 7 points of Combo_7 going into the GF.

Bookie_9, who sits third on the MoS Leaderboard and who trails Bookie_LPSO by over 7 goals, is now too far behind and insufficiently diverse of opinion to hope to bridge that gap.

From an historical viewpoint, one aspect of this week's activity that's particularly interesting is the fact that we'll see the team finishing 2nd on the end of home-and-away season Ladder meeting the team finishing 5th in a Semi-Final for the first time in the current century. In fact, no team finishing 5th on the Ladder has ever progressed this far in that time period, regardless of the team they'll face.

We also see, in the other game, 1st versus 6th, which we've witnessed only twice before since 2000, those two games representing the only occasions on which the team from 6th has reached this stage of the Finals. In both of those previous encounters the Minor Premiers emerged victorious.

Overall, in Semi-Finals, Minor Premiers currently have an 11-3 record,  Runners-Up have a 9-5 record, 5th-placed teams are 0-0, and 6th-placed teams are 0-2.

Returning, lastly, to MoS' Probability Predictors we find that, from a Head-to-Head perspective there's no doubt about who wins, but some debate about the relative likelihood of those results.

In Friday night's game, opinion varies about the Swans' merits from a C_Prob low of a 70% probability to a WinPred high of 92%. Saturday's game sees a low assessment of the Hawks' chances at 71% from ProPred and a high assessment of 95% from WinPred. Outliers aside, the opinions of most Probability Predictors are quite similar though, and have the Swans as 77-84% chances, and the Hawks as 71-78% chances.

The Line Fund algorithm rates the Swans as 62% chances of prevailing in the Line market but also rates the Hawks v Port game as something close to a coin toss.


There are no shenanigans with ChiPS this week in relation to Home team status. and we wind up with a low-end estimate of the Swans' victory margin (18 points) and a mid-range estimate of the Hawks' victory margin (21 points) in the context of MoS' Margin Predictors.

Those margin forecasts translate into victory probabilities of about 70-75%, which is towards the low end of the assessments of all Head-to-Head Probability Predictors and too low to justify a wager at prevailing TAB prices.