The Other Bundesliga
The Other Bundesliga
If you read my previous article, you would be aware that the predictive modelling production has been ramped up somewhat and once you on something of a creative groove, it seems silly just to stop.
Therefore, I’ve decided to turn my focus to another of Europe’s top level leagues and with qualification to next Summer’s European Championships confirmed, there is undoubtedly something of a party mood in Austria.
A mood that I want to tap into, as I unleash ‘AutBot’ on the world and see if I can garner success from predicting results in ‘The Other Bundesliga’ – which if you don’t know is also the name of Austria’s top division.
Not only that, but it also the name of an incredibly good Austrian Football podcast, and if you want to know more about one of Europe’s footballing more hidden secrets, the guys can be followed on Twitter.
Why the Austrian Bundesliga?
Previous readers may also be aware, that I had a crack at predicting results from this competition on pure gut-feel alone and it is fair to say that my success was not all that impressive (just two out of six correct). Something which on further inspection, wasn’t all my fault.
After some more investigation, it turned out to be ‘one of those weeks’ and ultimately it was just a bad week in which to try and climb this particular Austrian mountain. However, this previous failure has not deterred me and now I am back for more of a data driven approach.
First up, let’s look at the infograph that will follow in next week’s fixtures:
Now looking at these six fixtures I would have to say that my gut-feel picks here would be:
Salzburg – Home Win
Wolfsberger AC – Away Win
Lask – Away Win
Admira/Austria Wien – Draw
Hartberg – Away Win
Rapid Wien/Sturm – Draw
Now let’s look at what ‘AutBot’ has conjured up:
As you can see, we have an exact match – therefore the modelling logic and the gut-feel are running parallel. Now this can either be a good thing, or we are both set to crash and burn come this time next weekend.
Not only that but when you consider how dominant Salzburg and Wolfsberger have been (especially over the past six weeks) that should suggest that two bankers are on the way and we only have to really get four other results correct.
However, there are a couple of things to take into account here:
1) All runs must come to an end
2) The other four matches will be a lot more competitive and the likelihood is that at least one of these will go against the grain.
The second point is great for a league point of view, but something of a nightmare from a betting/predictive modelling stand point and more often than not, that is why trying to predict all of the fixtures becomes far harder than it looks.
In regard to the model build, it is based on version r2.0 and this one that compares the home team’s last six results at home and the away team’s last six results on the road, while also adding an element of bias to the home team.
Therefore, this model will be used as the first version of ‘AutBot’ and this particular week will be used a control model. From here, we will analyse what went right and wrong and see what changes then need to be made.
Anyway, that is me not only covering the ground in terms of a control test but also getting my excuses in early, if it does go wrong again. While on the other side of the international break, I will review just how successful this new sub-project has been.
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.