2016 Round 4 : The Dons Conundrum

This week Investors have another opportunity to mull over the wisdom of MoS' willful blindness to team composition.

When the Dons opened with the bookmaker at $8 early in the week, wagers on both the head-to-head and line markets seemed likely, and when they drifted to $13 head-to-head and received 65.5 points start in the line market, that likelihood crystallised into a 2.5% wager from the Head-to-Head Fund and a season-high 6.3% wager from the Line Fund.

That Dons Line Fund wager dwarfs the three others in play this week, two of which are on favourites and one on an underdog. None is larger than 1% of the Line Fund.

In the head-to-head market, the Dons wager is joined by four more, all sized at 2.5% of the Fund but at prices ranging from $1.22 for the Eagles to $3.55 for the Brisbane Lions.

As I write this, the Overs/Unders markets are yet to be posted. I'll post a supplementary blog once they have been.


This week sees a return to more usual levels of disagreement amongst the Tipsters and Predictors, the Head-to-Head Tipsters, for example, turning in an Index of 24%, the second-highest of the season.

All of that disagreement, however, is confined to just three games, with the Tipsters favouring Collingwood over Melbourne 7 to 4, Sydney over Adelaide 6 to 5, and GWS over Port Adelaide also by 6 to 5. In the six other contests the Tipsters are unanimously behind the favourites.

Viewed on a tipster-by-tipster basis, disagreement levels are remarkably similar across the whole set, the exception being Home Sweet Home, which is just as often in disagreement as agreement with a randomly chosen peer this week.

The all-Predictor average mean absolute difference (MAD) is back up to 5 points per Predictor per game this week, which is more than 10% higher than last week's figure and only a fraction less than the figures for Rounds 1 and 2. 

Saturday's Essendon v Geelong game is a significant contributor to that increase, the predictions in that game spanning a seven-and-a-half goal range and carrying a 14.5 points per Predictor MAD. Only one other game has an above-average MAD, the Carlton v Western Bulldogs game where it's 6.7 points per Predictor.

C_Marg, which is one of the Tipsters at the low end of the projected Cats' victory margin, has the round's highest MAD of 8.1 points per game, ahead of MoSSBODS_Marg's 7.1 points and Bookie_3's 6.2 points. Bookie_LPSO has the round's lowest MAE of just 2.9 points per game.

Turning next to the Head-to-Head Probability Predictors we find slightly higher levels of disagreement this week compared to last, the Index rising to 4.1% points per Predictor per game, up from last week's 3.4%, but still below the 5.5% and 5.6% figures from Round 1 and Round 2. 

Four games have above-average MADs, though it's the Dons v Cats game with by far the largest MAD of 8.7% points per Predictor. As for the Margin Predictors, it's the ChiPS- and MoSSBODS-based Predictors that have the highest MADs, here 5.9% points per game and 5.1% points per game respectively. Bookie-OE, narrowly, has the round's lowest MAD, though it, Bookie-RE and Bookie-LPSO all have essentially the same figure.

MoSSBODS Details

MoSSBODS has only four teams enjoying significant "home" venue advantage this week (West Coast, Adelaide, the Kangaroos and Hawthorn), three of those mostly or entirely because of the interstate nature of their matchups. In the other games the notional home team has only a fractional advantage or, in the case of Carlton a not insignificant disadvantage of over 2 Scoring Shots.

Gold Coast are predicted to produce the round's highest score, their expected haul of 108 points just slightly ahead of the Western Bulldogs' 106, and West Coast's and the Kangaroo's 102. The week's low-scorers are projected to be Carlton (57), St Kilda (62), and Richmond (70)

The biggest total is projected for the Lions v Suns clash, where the expectation is for 200 points to be registered. Both the GWS v Port (157 points) and Hawks v Saints (158 points) games are expected to generate fewer than 160 points in aggregate. 

So far this season, MoSSBODS has correctly identified the highest-scoring team in Round 1, and the highest-aggregate game in Round 2. It's yet to correctly identify the lowest-scoring team or the lowest-aggregate game however.

Let's hope it's made the right call about Essendon ...