WoSHBODS’ assessments of the relative merits of a number of teams continues to deviate from the bookmakers’, which has resulted in its forecasting generally larger margins including by as much as 3 goals in some cases.

Below is a summary of the latest simulation results for the AFLW (produced using a version of the methodology described in this blog post).

So, roughly speaking, WoSHBODS currently groups the teams as folows:

• Tier 1 (90%+ chances to play Finals): Melbourne, North Melbourne, Adelaide, Gold Coast, and Geelong

• Tier 2 (80% chances to play Finals): Brisbane

• Tier 3 (50% chances to play Finals): Richmond and Essendon

• Tier 4 (35-40% chances to play Finals): Collingwood, Carlton, and Sydney

• Tier 5 (2.5-8% chances to play Finals): Fremantle, Port Adelaide, Hawthorn, and St Kilda

• Tier 6 (1% or less chances to play Finals): Western Bulldogs, GWS, and West Coast

Three weeks in now, and WoSHBODS has tipped 78% of the winners, registered an 18.8 points per game margin MAE, and facilitated a 23% ROI on wagering. Be hard not to be happy about that.

After a second good week on forecasting, WoSHBODS returns with its thought on nine new games, which include forecasted margins of two goals or more in all but two games. They also include three forecasted Totals of more than 100 points after Round 2’s average of just under 90 points per game.

As foreshadowed last week, this week sees large movements in WoSHBODS’ assessment of the 18 teams’ Finals chances, as recorded in the table below (produced using a version of the methodology described in this blog post).

(We should also expect fairly large movements again next week.)

It was fairly easy to predict winners in Round 2, but much harder to predict Margins and Totals, which is why WoSHBODS picked 8 of the 9 winners, but ended the round with a 19.9 Margin MAE and a 19.3 Total MAE.

In this blog there are the first ever WoSHBODS simulations for the remainder of the season and for the Finals.

They use the same hot sims methodology that I’ve employed over on the men’s site and that I’ve described in this blog post, with suitably reparameterised negative binomial and beta binomial models for simulating individual game results.

They are also subject to very high levels of uncertainty because they’re based on relatively few historical games, and because the women’s game is developing rapidly and at different paces across teams.

After an excellent start to the 2023 AFLW season, WoSHBODS is back for Week 2 after having made some sizeable changes to individual team ratings, which is its habit now in the early rounds of new seasons.

Back in the dimest of darkest of ages. when I was younger yet less wise, I wagered on men’s AFL for the first time, and, in 2023, I did the same thing on AFLW and emerged with a small profit (despite eschewing my “never bet head-to-head on away teams” ethos.

I’ll definitely take that.

I also emerged with a 16.4 Margin MAE, and a 17.8 Totals MAE, which I will also take.

We’re back again for the 2023 AFLW season with a re-optimised WoSHBODS model and a willingness to put that model to the test against the bookmakers.

WoSHBODS finishes the season with an incorrect tip, an excellent Margin Absolute Error, and a quite large Total Absolute Error.

Below are WoSHBODS’ forecasts for the third week of the AFLW Finals.

Very close to a perfect round from WoSHBODS this week as it tipped both winners and finished with a 2.5 points per game Mean Absolute Error for the Game Margins, and an 11.3 points per game Mean Absolute Error for the Game Totals.

Below are WoSHBODS’ forecasts for the third week of the AFLW Finals.

WoSHBODS tipped both winners this week, one by too little and one by too much, perhaps suggesting that its approach to Venue Performance Values isn’t too outrageous. It finished with a Margin MAE of 15.8 points per game, a Totals MAE of 41.9 points per game (thanks largely to the Roos’ 74 point haul), and yet another positive log probability score, its seventh in succession for a full round.

Below are WoSHBODS’ forecasts for the second week of the AFLW Finals, which again include some margin-expanding Venue Performance Values. Interestingly, it finds the Punt Road venue enhances rather than diminishes the Roos’ chances.

WoSHBODS was definitely too heavy-handed with Venue Performance Values (VPV) this week, ending up with a Margin MAE of 25.4 points per game as a result, but it did tip all four winners and, such was its confidence, also returned a highly positive log probability score. Its Totals MAE was a very respectable 12.5 points per game, although the VPV values used meant the Away Team Score forecasts were relatively poor.

Below are WoSHBODS’ forecasts for the first week of the AFLW Finals.

Another pretty good week for WoSHBODS, with a Margin MAE just above the season average at 19.6 points per game, a Totals MAE of 21.4 points per game, a fifth consecutive positive LPS, and a 7.5 from 9 result on accuracy, all of which has led to the table below.

Below are WoSHBODS’ forecasts for Round 10 of the current AFLW season.