Matter of Stats

View Original

2018 - Round 3 : Lots More Disagreement

The tipping carnage in Round 2 was always a possibility this week given the general slightness of favouritism in each of the nine contests.

This week sees a significant increase in the average quality-gap between competing teams, with six of the favourites priced under $1.50 at the TAB, and the average line handicap coming in at 17.5 points per game, more than 4 points higher than we saw in Round 2.

That said, in an historical context, an average handicap of less than 3 goals per game is quite small - since 2012, only one season has had a smaller average for Round 3 - so it's not as if we should be hoping for a slew of 9 from 9 head-to-head tipping results.

TIPS AND PREDICTIONS

One thing that is distinctly different this week is the level of disagreement across the forecasters.

Amongst the Head-to-Head Tipsters we have particularly high levels of contrarianism from Consult The Ladder, C_Marg, MoSSBODS_Marg, as well as the perennial dissenter, Home Sweet Home. The average Disagreement Index - the probability that a randomly selected Tipster would disagree with another randomly selected Tipster in a randomly selected game - has come in at a season-high 30%.

Amongst the Margin Predictors it's Bookie_3, C_Marg and the MoS twins that are elevating the overall mean absolute deviation (MAD) most, with MoSSBODS_Marg the Extreme Predictor in five contests, C_Marg in four, and MoSHBODS_Marg and Bookie_3 in three each.

Looking across the games, the forecasts are most disparate (in a MAD sense) in the Melbourne v North Melbourne, and Gold Coast v Fremantle games, though the forecasts in the latter contest span only a 16 point range. (And, by the way, how do you estimate a sensible Home Ground Advantage for a Suns v Dockers game where the Suns are the designated home team and the contest is being played at Perth Stadium?)

C_Prob and the MoS twins are also heavy contributors to the overall average MAD for the Head-to-Head Probability Predictors this week, with C_Prob the Extreme Tipster in five contests, MoSSBODS_Prob in four, and MoSHBODS_Prob in three.

Four games have high MADs, all of them above 6% points per Predictor: Melbourne v North Melbourne, West Coast v Geelong, Sydney v GWS, and Richmond v Hawthorn.

We should expect then to see some significant movements on all three Leaderboards this week.

WAGERS

With the Line Fund a little more active this week than last, and the Head-to-Head Fund just as active, Investors find themselves with a larger aggregate risk this week from their three line and four head-to-head bets .

(Please click on the image below to access a larger version.)

None of the four head-to-head wagers is on a favourite, and the largest is on Gold Coast facing Fremantle where MoSHBODS has the Suns as only the slightest of underdogs despite (or, more accurately, completely oblivious to) their $2.80 price tag. Together, the four head-to-head bets represent 8% of the Fund.

The largest of the line wagers is 3.6% on the Roos, who are getting just under 4 goals start in taking on the Dees, and the next-largest is about 1% point smaller and on the Eagles getting just under a goal's start in taking on Geelong. In total, almost 8% of the Line Fund is also at risk across the three wagers.

The ready reckoner for these wagers is shown below.

Success versus failure for the head-to-head and line bets on the Roos represent the difference between a 3c gain and a 2c loss - a 5c overall swing. The next-highest potential swing comes in the Suns v Dockers game where the difference between a Suns win and a Suns loss is equivalent to 3% of the original Combined Portfolio. 

Across the nine games, about 5.5% of original funds are at risk for a maximum upside of just under 8%.

To finish, here are MoSHBODS' and MoSSBODS' opinions about likely team scoring.

Commentary about these forecasts will come in the week's Overs/Unders blog.