Third Time Lucky
Third Time Lucky
If you read my previous article on trying to beat the English Premier League via a data driven betting approach, you will be familiar with the slight improvement that was made last weekend and although I missed out the midweek action, it’s time to go again.
Last week saw the predictive model otherwise known as ‘PremBot’ record five correct picks from the possible ten on offer and where it not for some relative shocks at both St James Park (Manchester City draw) and Stamford Bridge (Chelsea loss) then the overall outcome could have been even better.
Of course, there are always going to be the odd shock along the way and on the evidence of last weekend, you would have to say that it was a good weekend for the bookmakers and this week, I’m hoping it swings back in my favour.
Therefore with that in mind, let’s have a look at the predictive picks for GW16:
Now there are a couple of things to take into account here and that is that three clubs have sacked their managers in the past weeks, with Marco Silva joining both Quique Sanchez Flores and Unai Emery on the proverbial scrapheap.
While these three departures, does lend itself to a rather pertinent question:
SHOULD I FACTOR IN NEW MANAGER BOUNCE
When Jose Mourinho was installed as Tottenham manager, the lift the club had was almost instant and that meant they came out on top across local rivals West Ham, something that ‘PremBot’ did not factor in at all.
While although hindsight is a wonderful thing and that change should have been made, at the same time when you look at the clubs that have recently handed out their P45’s, there is little to suggest that even a new man at the helm would make much difference.
Take Arsenal for example, they’ve installed Freddie Ljungberg as an interim manager and the former Swedish international has picked up just one point from the last six on offer, while facing Norwich and Brighton to boot.
Therefore, I’m not in a rush to start adding extra values to Arsenal, Everton and Watford have a new man at the helm. Although going forward, there may need to be some element of form of form override, where quite simply new manager = win.
Then again and as proved in the past few days, that is by no means a guarantee either and therefore, because I saw improvement in ‘PremBot’s performance last week and returned five correct picks out of ten, I’m going to leave things unchanged.
Part of this is due to stubbornness, the other due to not really having that eureka moment, that would necessarily make the model more intuitive. Therefore, with that in mind, I’m going to spout the cliche, if if ain’t broke don’t fix it.
After five weeks of trying to model the Premier League results correctly, I’m on a 48% hit rate, admittedly that’s not a success level you can take to the bank (that is if we were playing for money) but it’s not been a total washout either, especially with the project in its infancy.
If I can keep improving my success week on week (therefore the target for this week, is a minimum of 6) then I know that we are getting closer to the promised land, although if things go the other way this week, I will look to implement some changes before the busy Christmas period.
Now all we have to do is sit back and enjoy another feast of Premier League football and hope my data driven betting approach is one that is a certified winner. We live in hope people, we really do live in hope.
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@example.com. 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.