Test And Learn
Test And Learn
If you read the previous article in the series, you would have seen me explain just why the Russian Premier League was the competition of choice and although this does offer a more analytical viewpoint in terms of trying to build a model, it has led me to think that more can be done.
Now as I’m not prepared to bankrupt myself at any point, I’m only using a small outlay of £5 per week (10 units), which in itself is not going to make me rich anytime soon. Unless, I manage to get ‘StatBot’ to produce a 8-fold and return something in the region of £1000.
That would be nice, but it is going to take some time or perhaps a huge slice of luck in regard to being able to scoop such a big prize. However, that time can be cut down to a certain degree, if I look at it from a different view point.
In the previous piece, I referenced the fact that I wanted to take the emotion out of predictions and not get sucked in by previous biases. That is still the case, but I can actually still test my theories on competitions such as the Premier League and all the biases that will come with it.
How can that be done?
At present, I have a model that is working on predicting Russian Premier League fixtures and I’ll be testing and learning along the way. However, this development may come across at a slower pace than is necessary.
How can we get around this?
We could use the same framework for the English Premier League but make more advanced tweaks and more importantly, only test the percentage of success than opposed to actually playing for hard earned money.
Why do that?
Because if you can find a model that works, you can then scale it up and across all the multiple leagues. Build something that keeps working week in and week out and then when you’ve finally got the end result, pull the trigger and bet for real.
So the Russian Premier League is still going to be as an ongoing series, however I’m going to expand operations in the background, to see just how successful this really can be – not only but it also allows me the ability to compare and contrast.
Twice the work – perhaps, but twice the content – fantastic.
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.