A Solid Platform
A Solid Platform
If you read my previous offering on the English Premier League, you will have been aware that this last weekend was the first weekend in which I put my predictive model to the test, in a bid to guess the correct results in any given gamweek.
Now with the weekend’s action out of the way, it is now time to see how I fared but before I do so, lets have a look at the predictions first:
As you can see I got six out of ten correct and if it weren’t for Lucas Digne puttng through his own net for Everton, then we would have been one better. However, I’m not going to look at the ifs and buts all season, because there will undoubtedly be a fair few of those.
Therefore, let’s look at the actual successful picks, a half-dozen that were as follows:
Leicester Win
Manchester City Win
West Ham/Sheffield United Draw
Chelsea Win
Liverpool Win
Arsenal/Crystal Palace Draw
To be honest of those six, I would have gone with the first five in terms of ‘gut instinct’ but I referenced the fact that I thought the model was incorrect for the Arsenal/Crystal Palace pick – but little did I know, the model seems to know more than me at present.
Ultimately as there is no money on the line, we are looking at this from a pure percentage strike rate and with six out of the first ten trial picks being correct, it does not take a genius to work out that we are currently at 60%
Therefore what changes if any need to be made for this coming week? To be honest, when you’ve had an above average week in terms of predictions, it’s almost counter-intuitive to then pick things apart.
With that in mind, it might be better to see if last weekend was not a fluke and see if it can return something similar this time around. While if it doesn’t that is when I will start to tinker under the bonnet and make the necessary changes.
What is interesting though, is that in a sense this could be an easier league to model because of the perceived disparity between the ‘Big Six’ and the rest of the division (although that has somewhat dissipated over these past few weeks) therefore some of the games almost pick themselves,
At the same time, that is where previous bias may see us come unstuck because you are building around what you belief rightly or wrongly will happen and therefore, it might be easier to work on a league where you have no notions of good/bad/ugly.
That point can be countered by the fact, that you may end up working on something that is harder than it needs to be and therefore why not rely on the depth of knowledge that you have on the English Premier League?
There’s not necessarily a right or wrong answer, although a lot will depend on what results come in but if we can get somewhere in the 70% region for the rest of the season, then that is a lot better than the 45% I recorded when I did a whole season of predictions on blind picks alone.
So for this coming week, I think the plan is to leave things unchanged and hopefully we can use this a solid platform in which to build on. While if it doesn’t bring in the right results, it will be back to the drawing board once again.
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 [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