I Am Going Dutch
I Am Going Dutch
If you have read the previous article, you will be aware that my third week of English Premier League testing was one that was somewhat below average in terms of performance and with us being in the midst of the international break, it means all live testing is on hiatus.
Which is a bit of a shame on one hand, although on the other it does allow to setup more predictive modelling and roll out the testing of another league. With that in mind, it is time to unleash ‘NedBot’ onto the world.
Why ‘NedBot’ you ask? Because I am going to have a crack at predicting the Dutch Eredivisie and I like the first two competitions that I’ve tried to predict (with differing levels of succees), I am going to start with a control week first.
The model will be based on version r2.0 and this means that I will be using the home team’s last six results and comparing that against the away team’s last six – instead of just tracking performance from each of their last half dozen appearances.
Which means although the round of fixtures is still a week or so away, there is no harm in getting our first round of picks in early and they are something that looks like this:
As you can see, the competition consists of nine matches per week and with both Ajax and PSV Eindhoven being at home, I should at least get two of these correct (of course there is no guarantee of this)
Ultimately, there’s not a lot else to add to this because it is a control week, so this is nothing more than a bit of ‘international break’ fodder but over the course of the next few weeks, I should be able to add more tweaks to this.
In theory the more different league models I can have at any one time, the more I can test and learn the best approaches for each league. With the question being, do different leagues need different kinds of modelling?
I would imagine that may be the case, as I think one of overriding set of logic might be not only impractical but also inflexible. While the end game in all of this, is too create a ‘SuperBot’ and one where it can spit out all manner of results across all manner of leagues.
However, we are still some way away from this but it is good to have a long term plan in all of this and if/when that comes to fruition, then we can put a lot more leagues to the test – which also means we can start rolling out the financial element further.
Anyway, that’s your lot in this one as I know look to fill a bit more time during this international break. Then again, at least this is the last one until March 2020 and for that I guess we can all be very grateful.
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.