No More Draws

No More Draws

If you read my previous article, you will know that my pursuit for data driven betting excellence has been reignited and with a second week of Premier League action in the books, it is time to reflect on just how gut feel fared last time out.

Before we do that though, I want to look at an early statistical phenomenon which is hampering my ability to conjure up correct data driven football predictions and that is the fact that there are seemingly no more draws.

18 Premier League matches have been played this season and not one of them has ended in a state of stalemate. Why is happening exactly? To be honest, I have no idea as to why and although it is no help to you, it is of no help to me either.

Will this quirk expand itself all the way across week 3 of the season? I would be very surprised if it did. Then again, nothing surprises me when it comes to betting but with that said, if I bet against the draw this week, I’m sure one would pop up from somewhere.

More on that theory will come next week, but for now it is time to finally reflect on how I fared in terms of gut feel and the groundwork for whether it can eventually get the better of data (in a sense I hope not, otherwise this would have all been a huge waste of time)

HOW DID WE DO?

With the first full weekend of the season, 10 matches were made available to predict and here are more choices with their eventual outcomes:

Everton – Y
Leeds – Y
Manchester United – N
Arsenal – Y
Southampton/Tottenham Draw – N
Newcastle – N
Liverpool – Y
Leicester – Y
Aston Villa – Y
Manchester City – Y

7 out of 10, not at all bad to be honest. A top four finish in table terms, but not enough to bring the jackpot home. That means I am on 12 out of 18 for the season and have a srtike rate of 66.6% – a figure that if my data driven predictions could spit out come next May, I would be over the moon.

Of course, I also took a blind sample of gut instinct from social media and here is how the ramdom picks fared last weekend:

Everton – Y
Leeds – Y
Manchester United – N
Arsenal – Y
Tottenham – Y
Newcaslte/Brighton Draw – N
Chelsea/Liverpool Draw – N
Leicester – Y
Aston Villa/Sheffield United Draw – N
Manchester City – Y

This week ‘social media’ was punished by being draw heavy and only returned 6 out of 10 by comparison, the same number of current picks compared to last week (albeit from 8 selections). Which means a total of 12 and also a 66.6% strike rate thus far.

DEAD LEVEL

Which means my gut instinct and the combined random nature of social media’s are dead level at the moment. With that said, there is little to read into this at present and if we are say 6 weeks in and still in deadlock, that would mean the picks are relatively in tune.

While when it comes to intune picks, we should also look at how the results stacked up against the pre-match bookmaker odds:

Tomorrow, 19 Sep			1	X	2	B’s
11:30	Everton – West Brom		-185	+334	+542	13
14:00	Leeds – Fulham		-154	+299	+457	13
16:30	Manchester Utd – Crystal Palace		-323	+468	+877	13
19:00	Arsenal – West Ham		-189	+355	+511	12
20 Sep 2020			1	X	2	B’s
11:00	Southampton – Tottenham		+212	+236	+139	13
13:00	Newcastle – Brighton		+167	+218	+188	13
15:30	Chelsea – Liverpool		+220	+277	+118	13
18:00	Leicester – Burnley		-159	+304	+471	13
21 Sep 2020			1	X	2	B’s
17:00	Aston Villa – Sheffield Utd		+174	+223	+175	13
19:15	Wolves – Manchester City		+515	+338	-182	13

The bookmakers got 8 out of 10 correct, which means they have got 13 out of the 18 matches correct in terms of pre-match odds and are therefore just shading mine and social media’s gut instinct by 5.6% – 72.2% overall.

To summarise this:

Bookmakers  – 72.2% correct
Gut Instinct – 66.6% correct
Social Media – 66.6% correct

Which also means that in terms of modelling, we know that on average 7 out of 10 picks will go with the grain and 3 will go against. This is something that we will need to factor in to the ‘PremBot’ model going forward. Although with more data, this figure could drop to say 5 or 6.

For that we will have to wait and see, but the early gut feel has been good and if it continues, then my data driven football predictions are going to have their work cut out. Until then, it is back to the grindstone and I will be back with a week 3 piece on Friday.

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)

Follow me on Facebook at Dan The Stat Man
You can also check out my Premier League Podcast on Soundcloud