Model Performance - Round 1

How Did the Model Perform?

Throughout 2023 I will be comparing my model to those in the Squiggle model competition with hopes of competing in it next year. 

The Squiggle competition is a very well respected group of modellers who all predict the outcome of every match in the AFL season and then analyse their performance using various methods. In this post I explained how these were but for those in need of a refresher we have bits (confidence based picks), MAE (the average of how many points you were wrong by) and just general tips (how many games did you pick the winner).

On Squiggle the top models by bits are The Footycast who leads the way with 0.79, Elo Predicts who is second on 0.59 and AFLalytics a close third on 0.58. My model finished round one in a would be 17th place with -0.07 bits. Not to worry, 23 rounds of AFL football to go.

Looking at MAE now and we have The Footycast again in first with 27.13, The Cruncher in second with 28.33 and AFLalytics again in the top three on 28.63. The PPP model comes in a would be 12th place, just above the s10 model which is the average of last year's top 10 on the Squiggle website.

And for tips we are sitting on 5 right (the draw counts as a win) which is... not great, in the bottom half of the table but again a long way to go for this to even out.

Both bets suggested by the model also lost with Hawthorn and the Doggies going down (by a lot), however the closing line was a long way lower than the price we got, another tick for the model judging from a large amount of CLV.

The model also competes in the FMIplay competition which is a promotion/relegation style competition where 5 games are randomly picked every week and points are awarded against another player for getting both closest to the pin and the correct team right, bonus points for being within 3 points of the final margin and a final bonus point for getting the exact margin. Our PPP model lost round one of Division 7 by a score of 5 to 1 against The Footycast which was a very tough draw for us. 

Next week we play TheAFLBluster who drew their round one match against Sprenstama (both aren't models).

So overall it wasn't a great round one for the model but we can take solace knowing we are in the top half of MAE standings going into week two.

Comments

Post a Comment

Popular posts from this blog

Round 9 Tips and Bets

Round 8 Tips Richmond v Geelong (72%, 25pts) West Coast v Gold Coast (73%, 29pts) Sydney (68%, 9pts) v Fremantle North Melbourne v Port Adelaide (79%, 21pts) Hawthorn v Melbourne (80%, 38pts) Brisbane (74%, 20pts) v Essendon Carlton v Western Bulldogs (56%, 9pts) Adelaide v St Kilda (55%, 10pts) Collingwood (85%, 37pts) v GWS Not the greatest round of football coming up outside of two relatively close games near the end of the round. We should be expecting 7 at a minimum here. Head to Head Bets Season performance so far: 8.5 bits, 26.25 MAE, 48 tips (67%), -1.48 units, -6.6% ROI Only one head to head bet this round, 2 units on St Kilda to beat Adelaide at $2.30. This is a real tester of St Kilda's real standing at the moment and could equally expose any shortcomings Adelaide may have. Expect a well coached St Kilda side to look to defend the skinny Adelaide Oval and shut down the Crows in front of their home crowd. Futures A juicy 7 bets this week to get us going into the second th...
Read more

Moving On Up - Post Round 10 Model Analysis

Model Success, Short Lived or Sustained? After back to back successful weeks for the PPP model we are moving up the Squiggle leaderboards in bits, MAE, tips and profit on our bets. It felt like every game in round 10 went the way of the model for both long term and short term outcomes. How long this will continue is unknown but we will enjoy the pleasure when it is given to us. Bits We gained 2.17 bits in round 10 to go with the 1.79 from round 9 to move to 12.45 on the year. This has moved us up on Squiggle into the top 20 with a 20th place spot. We still sit 5 bits off the top spot which looks a bit of a long shot but continuing to move in the right direction is nothing but promising. MAE A couple very good MAE results have helped to lower the edge over the competition in the MAE fight too. We sit on 27.21 MAE for the season, sitting 20th as well in this ranking on Squiggle. Again this is a rise up from 23rd the round before and we look to continue making strides in this ranking too ...
Read more

Can't Stop, Won't Stop - Round 18 Model Update

Can't Stop, Won't Stop It's been a while since our last update of the PPP model so it is time to see where we stand with a quarter of the season remaining. We've had a few upsets lately but mainly games that have been close enough to 50/50 to call them a coinflip. Unfortunately for us, it feels like we've called every flip wrong, nearly for the entire season. Most recently the model was extremely bullish on Port Adelaide to beat Carlton at Marvel stadium but the 'Baggers come out swinging after McKay was subbed out and gave the Power their first loss in over 10 games. We had the Bulldogs 53%, the Swans won. We had Brisbane 52%, Melbourne won.  While these close games don't determine a lot of a model's success, it does feel like we have not been on the winning end of one of these all season, but I digress. Bits We have slowly accumulated our bits as the season progresses with 21.84 post round 18. Comparatively to Squiggle models, it is not at the pace we ...
Read more