Matter of Stats

View Original

2020 - Round 20 (Week 2 of the Finals) - The MoS Twins Get Brave

Both MoS Team Rating Systems have parameters that determine how Venue Performance Values (VPV) are estimated. Two parameters are key:

  • The default assumed VPV for a Venue (which is 0 for a venue in a team’s home State, and negative for venues outside their home State)

  • The number of games a team needs to have played at a Venue (in the past 6 years) before we override the default and, instead, come up with a value based on how the team has actually performed at that Venue

Those assumptions turn out to be crucial for this week’s Geelong v Collingwood clash at the Gabba, because the Cats have played too few games there in the relevant time period to override the default VPV assumptions under either System, which consequently sees them carrying a non-trivial net VPV deficit.

You could certainly make the argument that this year is different so one-off adjustments should be made, but it’s been a principle here on MoS that adjustments are not made mid-season, except for obvious errors. The parameter values have been chosen because they’re optimal for the 123 year history of the V/AFL, so maybe they’ll be okay in this its 124th year. We will, as they say, see.

And so it is that MoSSBODS is tipping the Pies by 14 points, and MoSHBODS tipping them by a slightly less startling 4 points.

You can see what the other MoS models think in the table at right, which shows that those two MoS tips are the only contrarian ones made by any of the Head-to-Head Tipsters.

You can also see that the MoS twins are somewhat divergent in their opinions about the Tigers v Saints game where MoSHBODS has the Tigers winning by almost 5 goals, and MoSSBODS has them winning by less than 2 goals.

On the Leaderboard, MoSSBODS_Marg currently trails Bookie_9, who’s in top spot, by less than 3 points. The absolute difference between MoSSBODS_Marg’s and Bookie_9’s forecast margins is about 24 points, so some favourable results will see MoSSBODS_Marg atop the ladder at the end of the round. MoSHBODS_Marg, who trails by about 19 points, differs in its forecasts compared to Bookie_9 by 23 points, and so could also do some climbing this week.

Amongst the Head-to-Head Probability Predictors the probability estimates span almost 20% points in the Tigers v Saints game, and over 25% points in the Cats v Pies game.

It seems fairly likely that, even this late in the season, we’ll see some movement on, at least, the Margin Predictors and Head-to-Head Probability Predictors Leaderboards.

RECENT FINALS HISTORY

The table at right below summarises the results from the last 20 years of men’s AFL Semi Finals based on the final home-and-away ladder positions of the teams, with the cells shaded blue reflecting the matchups we have this year.

Teams finishing 3rd on the ladder overall have a 10 and 4 record in Semi Finals, with 3 of the losses and 6 of the wins coming from clashes with teams finishing 6th.

Teams finishing 4th on the ladder have a slightly better 11 and 3 record in Semi Finals, and an impressive 5 and 1 record against teams finishing 8th.

Taking a whole-of-Finals approach, the table at left reveals how teams have fared across all weeks of the Finals, and shows the superior records of teams from 1st, 2nd, and 3rd, but also the near 50-50 records of teams from 6th.

That said, 14 of the 19 wins recorded by teams finishing 6th have come in Week 1 of the Finals. From Semi Finals onwards, their record is a less-impressive 5 and 13.

WAGERS

As the saying goes, when the MoS get divergent, Investors get going, which is why they face two largish wagers - one head-to-head, and one line - this weekend.

The Combined Portfolio maths of those two wagers works out as follows:

RICHMOND v ST KILDA

  • Richmond wins: 5.6% x 0.36 x 30% = +0.6c

  • Draw: 5.6% x 0.36 x 30% = -0.5c

  • Otherwise: -2.1% x 30% = -1.7c

GEELONG v COLLINGWOOD

  • Collingwood lose by 3 points or less: 3.1% x 0.9 x 60% = +1.7c

  • Otherwise: -3.1% x 60% = -1.9c

Finally, here are the details behind the MoS twins’ forecasts