A Wacky Week

A Wacky Week

If you read my previous article, you will know that once again there was another attempt at the pursuit of data driven betting excellence and with the international break now thankfully behind us, it is time to have a much delayed look at week four of this season’s Premier League.

WHAT A WEEK IT WAS

It may have been almost a fornight ago, but we cannot forget what was an incredible Sunday of action as even though West Ham picked up a shock away win at Leicester, this slip up for the Foxes would go somewhat under the radar.

Because not even Tottenham’s 6-1 win over Manchester United would manage to earn top billing just a couple of hours later. Instead it would be Aston Villa’s sensational 7-2 win over Liverpool and although that is great for those Villians supporters, it reeks of the one thing we don’t want.

THE COUPON BUSTER

Shocks are great in football, but not when you are betting on them and this means the overall return from week four of the season, might be less than one would have hoped for after laying out the 10-fold accumulator.

As before, it is a case of me versus social media and the bookmakers and now it is time to see just how we all fared, to recap here are my picks updated with the outcomes:

Chelsea – Y
Everton – Y
Draw (Leeds vs Manchester City) – Y
Draw (Newcastle vs Burnley) – N
Leicester – N
Southampton – Y
Arsenal – Y
Wolves – Y
Draw (Manchester United vs Tottenham) – N
Liverpool – N

6 out of 10, not too bad and I think a special pat on the back has to go to the prediction of Leeds and Manchester City ending all square, then again lets see how that stacks up against social media’s random picks:

Chelsea – Y
Everton – Y
Draw (Leeds vs Manchester City) – Y
Draw (Newcastle vs Burnley) – N
Leicester – N
Southampton – Y
Arsenal – Y
Wolves – Y
Tottenham – Y
Liverpool – N

Social media actually beat me by 7 picks to 6 in week four of the season, so a tip of the cap to the last set of random suggestions, while of course we also need to remind ourselves of the pre-match odds set by the bookmakers:

Tomorrow, 03 Oct			1	X	2	B’s
11:30	Chelsea – Crystal Palace		-227	+378	+642	13
14:00	Everton – Brighton		-120	+261	+363	13
16:30	Leeds – Manchester City		+661	+438	-256	13
19:00	Newcastle – Burnley		+149	+218	+211	13
04 Oct 2020			1	X	2	B’s
11:00	Leicester – West Ham		-145	+315	+391	13
11:00	Southampton – West Brom		-130	+289	+362	13
13:00	Arsenal – Sheffield Utd		-179	+313	+544	13
13:00	Wolves – Fulham		-192	+309	+630	13
15:30	Manchester Utd – Tottenham		-106	+274	+296	13
18:15	Aston Villa – Liverpool		+868	+446	-303	13

While the bookmakers also got the favourite right on 6 different occasions – which means when we look at the predictive league table it is as follows:

Bookmakers – 60.5% correct (23/38)
Gut Instinct – 57.8% correct (22/38)
Social Media – 57.8% correct (22/38)

As you can see there is not a great deal in all of this and my gut instinct is arguably just about where one would it expect it to be, although there is a modicum in comfort regarding that the Bookmakers are not all that far away.

We can also see that the bookmakers are averaging a 60% strike rate (taking into account that they would not necessarily opt for a draw like you or I would, so there overall success does come with a slight caveat.

Then again, if we use their success as a guide for how to build a model, it does suggest that the bookmakers are not going to get it right all the time and therefore, we should look to layer in an element of counter intuinity when we get to build “PremBot v2” in just a couple of weeks.

Admittedly, this wrap up article is way overtime and to be honest, it is just another recap one but once again we need to start laying the land for the more data driven football predictions that will be unleashed in just a couple of weeks time.

Happy punting and thanks for reading.

(THESE ARE NOT TIPS PER SE, THESE ARE JUST DATA DRIVEN FOOTBALL PREDICTIONS IN A PURE TEST AND LEARN MODELLING CAPACITY)