22nd – 28th April Bookings Model Update
This week, I want to do a bit of a deeper analysis into how the MLS & Brazilian leagues are getting on given that we’re fast approaching the end of the European season. Will be a slightly more concise write-up this week as the aforementioned tweaks have been added, so it has just been a case of tracking. Still, another 107 games added to the bank of data which is ideal. Let’s start off with the…
Weekend Results
Results viewable as a pdf here – Bookings Model Tracking – Week Recap (4)
Lifetime Results
In general, this week was a pretty significant drawdown, a week of negative variance, directly after a week of positive variance. The gains of last week were pretty much entirely wiped out, which is not ideal, but not something to particularly worked up over – a bad week is nothing out of the ordinary in the betting world.
My confidence remains high in the model, because of it’s accuracy. It is predicting totals to an accuracy of 95.5%, and splits to 98.2% overall.
As it stands, April is over 12 units up at an 8% ROI.
For me personally, total units returned is the most important thing. A high ROI is of course optimal, but secondary to total units returned. Because of that, volume is key. Ultimately, I’m happy with the total volume of bets this month, but will need to look into ways of improving the ROI – above 10% should be the standard.
With that being said, I will be having a look into the numbers, and potentially adjust the strategy slightly.
Update on the Summer Leagues
The projects, MLS and Brazilian Serie A have been getting along nicely in the background – but are they ready for predictions to be shared? Lets have a look.
MLS
Last week, predictions were certainly a bit wayward, particularly on the handicaps. This week, things look slightly better, a lot of unders recommended which is ideal. Handicap predictions are also looking more accurate, so I think a couple more weeks and this league could be ready.
We are at gameweek 10 in the MLS, so hopefully by week 12 predictions will be good to go.
Brazil Serie A
The Serie A predictions are a bit difference, because they are based off data from last season – so we’re expecting predictions to actually diverge and become worse as the weeks go on.
After 6 weeks, predictions are still relatively close to market lines, which probably suggests an inefficiency more than anything. The strategy going forward will probably be to start basing predictions off current-season data, and compare which ones look more accurate and provide better results.
I’ll use these two sets of data to try and determine when it is best to use current season data, and when to lean on the numbers from the previous campaign for the the 25/26 season starts in Europe – this is another important thing to work on.
Thank you for the support on this project, your interactions are greatly appreciated.
