As the title suggests, I’m not sure my ratings and systems can do much better than they did this weekend. It’s one thing to select a list of games where there is value in backing one of the sides and in my opinion, this isn’t anywhere nearly as difficult as people imagine but to narrow these games down and select the very best games that offer value, this is the thing that separates a good set of ratings from a great set of ratings. As I’ve shown over the past 18 months though, my ratings can do this effectively (some of the time!) but when it works, it doesn’t half look spectacular.
This weekend, the top 8 teams on Algorithm one (system 7) returned 7 winners. The odds of the winners were 4.77, 4.50, 3.20, 3.13, 2.62, 2.00 and 1.80. The loser was Newcastle at odds of 11.40. I can forgive the algorithm that one!
The top 4 teams on algorithm two (system 22) didn’t do too badly either. 4 selections, 4 winners. The odds were 4.81, 4.50, 3.13 and 2.62. Not a bad performance.
Not to be outdone, the top 3 teams on algorithm three (system 33) had 3 selections, 3 winners. 4.50, 3.13 and 2.00 were the odds of these selections.
One great thing about having so much data is that I can look at losing runs and determine how unlikely the run was to occur and it helps keep me sane. Well, it also works the other way too. Looking at my historical results quickly this morning, the systems are on their 3rd or 4th best run historically and the weekend past was maybe the 4th or 5th best weekend ever in terms of ROI, given the number of selections. It’s always hard to judge good/bad spells as pts won/lost doesn’t always tell the story as it depends on the number of bets too.
As someone said to me in an email this weekend, to see the backtested results of my systems is very good. To see it happen in practice when they are following the bets week in, week out is something very special.
Now, if I wind the clock back 12 months, the returns on the systems were much in line with where they are now this season. Things were looking great and I thought I had built something really special here. We all know what happened last season after the systems peaked at Christmas and therefore, with over 70% of the season to go, I know only too well that there is a long, long, long way to go this season. However, you have to enjoy moments like these when they come along as they don’t come along too often and they never stay around for long at all!
Anyway, onto the results from this weekend.
Algorithm one had 22 bets and hit 9 winners. A profit of 9.47pts for H/A betting and a profit of 2.94pts for DNB betting. (Bear in mind that system 7 took these 22 bets, selected 8 of them and hit 7 winners.)
Algorithm two had 14 bets and hit 8 winners. A profit of 15.36pts for H/A betting and a profit of 9.42pts for DNB betting. (Bear in mind that system 22 took these 14 bets, selected 4 of them and hit 4 winners.)
Algorithm three had 14 bets and hit 6 winners. A profit of 6.41pts for H/A betting and a profit of 1.85pts for DNB betting. (Bear in mind that system 33 took these 14 bets, selected 3 of them and hit 3 winners.)
Overall, knowing what you know above, it’s clear that the systems were going to have a great weekend.
The established systems from last season had 73 bets, hit 46 winners and made a profit of 87.98pts for H/A betting. DNB betting lagged behind a little but a profit of 50.05pts from 73 games isn’t bad!
The new systems from this season had 49 bets, 24 winners and a profit of 35.46pts for H/A betting and a profit of 19.90pts for DNB betting.
Overall, all the systems combined had 122 bets, hit 70 winners (57%), made a profit of 123.44pts for H/A betting and a profit of 69.95pts for DNB betting.
As I said at the start, weekend’s like that don’t happen too often. Maybe once a season. Enjoy it. :)
There are no bets this midweek. Only a couple of games take place in the Championship and although Burnley are a marginal value bet, they don’t appear on any system. Dunfermline are also a value bet in the SPL but appear on no systems. The systems can have a midweek off.