Brest Is Best
Brest Is Best
If you read my previous article, you will be fully aware that the Bundesliga is once again ready and open for business and although my pursuit of data driven betting excellence may now have a more teutonic feel to it, my good friends in Belarus certainly won’t be forgotten.
That’s because another weekend of Belarus Premier League action is instore and that means, it is time to once again serve up what at present, have been nothing short of an absolute shocking set of data driven football predictions.
Now regular readers (of which there are some) will know that I’ve tried to think a bit more outside of the box when it comes to the Belarus Premier League and this come in the form of A/B testing across two models.
In which a short summary would be:
The standard variant of the model has been rancid
The smaller variant of the model has been less so
The standard sized ‘BelaBot’ is working worse than had I just picked the games at random, while it’s younger and slimmer brother, redeemed itself with 4 out of 8 correct picks – so at least there is something to work on that front.
However, the focus fundamentally lies with the model that looks at form other the last six matches (this will eventually become home vs away – but for now, it is just the last six) and as mentioned a few days ago, tweaks are going to have to be made.
With the performance being so bad, it allows us to throw a little more caution at the wind and that means we can try ideas, that may or may not have been risked on competitions such as the Premier League or La Liga.
Before we get to that point, I once again need to lay the groundwork for this week’s data driven football predictions and that comes with the Belarus Premier League form guide itself:
DOING THE ADMIN
Now first up I need to do a little admin that comes in the way of marking the scorecard, which this week actually one has seven matches to choose from and with FC Minsk vs Neman Grodno postponed, there is one less opportunity for this to go wrong.
With that in mind, my usual saviours BATE Borisov will have a tougher task this weekend, as we are treated to a top of the table clash on Saturday afternoon – which I might watch half an hour of before the Eintracht Frankfurt vs M’Gladbach game.
In addition to all of this, I also carried out some deeper research and looked at the teams that got promoted to the league this season. Something that offers a little more knowledge, although with said, this was meant to be a pure data driven project and this knowledge may create bias.
Admittedly its not the most pertinent amount of knowledge, but now I know that Smolevichy and Belshina Bobruisk were promoted, it sort of stands to reason wht they are alreay in danger of being cast adrifrt at the foot of the table.
Does this mean, that outcomes will be tweaked to always suggest a defeat, perhaps but to be honest that would not be in the spirit of letting the data decide and with my concious now cleared, it’s time to reveal a first major tweak to my model ‘BelaBot’
PUTTING ON SOME WEIGHT
With the past two weeks being far from impressive, I’ve finally decided to throw my weight around and that’s exactly what I’ve done for the home teams in this week’s model, in the hope that it adds a bit more empahsis on the home teams,
Admittedly home advantage is not that pertinent in this current climate, but the mentalitty of a home fixture cannot be understated and it is perhaps something that I’ve neglected across this and other leagues.
Of course, the beauty of this is not just the opportunity to beat the bookies but also use it as a test and learn process and things such as this, are exactly the kind of thing that you should be trying when you have nothing to lose.
While because of this change, it has removed a lot of draws that had crept up and if you read last week’s offering, you will know that ‘BelaBot’ spat out no draws at all and that means, the balance in criteria is not quite – although hopefully it can be fixed with a little bit of numerical ballast.
Therefore with these seven picks now in play, I’m going to put on a very small accumulator to test the power of these predictions, it won’t be life changing but if I reveal the odds that a 7/fold would bring around it is: 1295/1, which means that fundamentally I am going against the grain.
To give a bit of context, if I went with the favourites only the odds would be: 40/1, so the formbook is certainly not being followed in this aspect. Then again, that’s probably what we want, because I’ve be very surprised if there were 7 wins from all 7 favourites.
Which means, there could be the potential to win relatively big, or we could be so wide of the mark, that the bookmakers once again. The fate of ‘BelaBot’ will be decided on the other side of the weekend, but of course, we need to look at the 3 game model also.
After getting four picks correct last week, there’s little reason to change and also a sense of not knowing what direction to turn to, therefore I’ll use the adage ‘if it ain’t broke don’t fix it, although it has thrown up an interesting results.
Even though the two sets of calculations are completely different, the results are nearly almost the same and the only difference is that Torpedo Zhodino are marked down as a win in this game and not the previously suggested draw.
Which either bodes well on both fronts or it could be absolute disaster, but again there’s not a lot more to say at this point except it is time to get the bets on and prepare for another weekend of Belarus based action.
(THESE ARE NOT TIPS PER SE, THESE ARE JUST DATA DRIVEN PREDICTIONS IN A PURE TEST AND LEARN MODELLING CAPACITY)
If this has grabbed your interest and you would like to discuss/feedback then please feel free to drop me a message at [email protected]. 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.
Follow me on Facebook at Dan The Stat Man
You can also check out my Premier League Podcast on Soundcloud