2016 - Round 25 (Finals Week 2) Results - Barking Up the Right Tree

Experience has taught me, more than once, that it's unwise to crow too loudly or too long when your statistical modelling approach has proven prescient. Nonetheless, history shows that simple stats modelling here on MoS has highlighted, for example, Freo's fatal flaws last year well before they bowed out of the Finals, and similarly has questioned the credentials of the Roos this season after they went 9 and 0.

Read More

2016 - Round 24 (Finals Week 1) Results - Almost Perfect

Investors enjoyed collects on five of the weekend's six wagers, missing out on a perfect record only because of the narrowness of the Cats' victory margin.

Those five wins and a loss added 2.2c to the Head-to-Head Fund, 0.4c to the Line Fund, and 3.6c to the Overs/Unders Fund, lifting the Overall Portfolio by 1.8c and putting it now 14.7c in profit for the season. That profit has come from a 7.6% ROI and a 1.9 turn on the funds in the Portfolio.

Read More

2016 Round 23 : Lopsided Lines?

So far this season, the average Line Fund wager has been for almost exactly 1.5% of the original Fund value, so it's a mite disconcerting that the Fund has chosen Round 23 with all the associated uncertainties about team motivational levels to make three bets each of 2% or more of the Fund. It, of course, knows nothing about the distorting influence of preferred draft picks.

It's also plumped for a fourth wager of just over 0.5% on the Saints giving the Lions 50.5 points start.

The Head-to-Head Fund, as has been the case all season, has narrowed its focus to consider only the same teams as the Line Fund, this week doubling-down on the Line Fund's larger wagers in three of games, passing only on the opportunity to get behind the Saints at $1.06. I am, I'll admit, very grateful for that.

With the larger-than-usual line bets, Investors face some sizeable swings between the best- and worst-case outcomes. The spread is largest for the Dons v Blues game where it spans over 5%, a Dons win promising a 3.4% gain and a loss by over 22 points threatening a 1.8% loss.

The spread in the Roos v Giants game is almost as large (4.6%), in the Crows v Eagles game a little smaller (3.4%), and in the Saints v Lions game smallest of all (0.4%).

In aggregate, the seven wagers put 6% of the original Overall Portfolio at risk for a maximum return of 7.6%. So, regardless of the weekend's results, MoS will definitely finish the home-and-away season in profit.

Last year, by way of painful contrast, Investors entered the Finals down by almost 33%.


Were it not for Home Sweet Home, this week would see the MoS Head-to-Head Tipsters in complete unanimity across all nine games.

As it is, the Overall Disagreement Index is just 8%, HSH responsible for its own Disagreement index of 44% and for the 4% figures for all other Tipsters.

Amongst the Margin Predictors, it's MoSSBODS_Marg and its derivative Predictors that this week have the highest mean absolute deviations (MADs), they and Bookie_3 the only Predictors with MADs above the all-Predictor average of 4.9 points per game per Predictor.

Only three games have MADs above this average. The Hawks v Pies game has by far the largest MAD of 9.1 points per Predictor, forecasts in that game ranging from a 6-point Hawks victory (ENS_Greedy_MoSS) to a 34-point Hawks victory (Bookie_3).

The two other games with above-average MADs are the Crows v Eagles game (5.5 points) and the Swans v Tigers game (5.3 points).

MoSSBODS_Prob's head-to-head probability assessments are the most divergent from the all-Predictor game averages this week, its MAD of 5.9% points per game almost 1.5% points larger than the all-Predictor average of 3.5% points per game. It differs most from all other Predictors in its estimates of the Crows' (89%), Roos' (49%), and Dockers' (18%) chances.

Still, I suppose, it's not going to drag itself off the bottom of the ladder by offering up conformist probability assessments.


The various MoSSBODS-based forecasters have been informed by the following view of this week's matchups.

Adelaide's projected 41-point victory stems from a significant underlying Offensive superiority (+6 SS), bolstered by a +4 SS Net Venue Effect, 3 SS of which stem from the interstate nature of the clash, and the rest from the teams' historical under- or over-achievement at the venue.

Similarly, according to MoSSBODS, the Roos' relatively elevated chances come entirely from a +3.4 SS Net Venue Effect, this fillip only falling short by 0.2 SS of having the Roos rated as favourites.

In the other games, MoSSBODS' projected winners are the teams with the superior assessed underlying ability, the Net Venue Effects merely reinforcing or slightly diminishing the expected victory margin for those teams.

Four teams are expected to score or exceed 100 points this week, St Kilda (128), Adelaide (111), Sydney (101), and Port Adelaide (100), while Richmond and Fremantle are expected to duel for the mantle of the round's lowest-scoring team.

The Saints v Lions game is projected to produce the most points (200), and the Dockers v Dogs game the fewest (145).