From Russia With Love
From Russia With Love
Admittedly I’m not sure how many Russian based puns I’m going to be able to use as article titles, but if you had read the previous offering you will have seen me reference that I was going to offering up the first raft of weekend bets.
First up lets look at the infograph for this latest round of Russian Premier League fixtures:
As you can see Lokomotiv Moscow and Zenit could be a tough one to call while CSKA Moscow and FC Krasnodar will be confident of earning wins this weekend against Ural and Arsenal Tula respectively.
Now looking at the above infograph would give you a good basis in terms of the necessary information required to pick a set of winners, but could you pick all eight correct outcomes across the gameweek?
Ultimately there’s nothing stopping you or I from doing so, but let’s see if a model can do the work for us instead. The question you are probably wondering, just how are we going to get the model to spit out eight outcomes?
THE FORM GUIDES
As I mentioned in the previous article, in this initial stage the form guides are going to become the bedrock for the predictive model and if you read that piece, you will also be aware of the form index I have created.
There are 729 different six-match combinations, from six straight wins being rank 1 and six straight losses being rank 729, with everything else getting a value in between.
Therefore depending on each club’s last six matches, they have all been assigned a rank and that rank will then be compared against each other – with the difference between the two, being the vital indicator.
I’ve written a small bit of code (one that can be adjusted as we go) to highlight certain difference thresholds offering up different results, so if it hits one threshold = home win, another = away win and another = draw.
With the final outcome for this week, being what you see in the image above and this will be the eight bets that I place for this upcoming gameweek. With each game being worth 1 unit (50p) and the eight game combo being worth 2 units (£1)
I did a blind control test last week and picked the outcomes by gut instinct and made a small loss, so lets see how I fare when I put my blind faith in the model and see what return I get this time around.
This means that my total outlay will be 10 units each week (£5) and will probably make sense to track the single game P/L rather than the combo bet record – only for the simple reason that we are shooting for the moon.
While this week’s moon based target is the above 8-fold coming in at a odds of 723/1.
So settle in and let’s see what’s on the other side of the weekend, when we’ll take stock of the results and see what tweaks or additions need to be made to the model.
Happy punting and thanks for reading, Dan.
If this has grabbed your interest and you would like to discuss/feedback then please feel free to drop me a message at firstname.lastname@example.org. While I am always looking for new football/data projects to work on and if you feel that my skills would be of use, I can be contacted at the same address.