Keep It Quick

Keep It Quick

If you read my previous article, you would know that the pursuit of data driven betting excellence is picking up pace but unfortunately not performance and with another full round of Premier League results now in the history books, it is time to quickly see how we did.

As before, it is a case of gut instinct and data driven football predictions, first up lets take a look at how the gut instinct picks did:

Chelsea – N
Manchester City – N
Southampton – N
Liverpool/Tottenham Draw – N
Leicester – N
Leeds/Newcastle Draw – N
Fulham – N
West Ham – N
Aston Villa – N
Manchester United –  Y

An absolutely pathetic 1 out 10, so moving on very quickly – lets have a look at home the first model or “PremBot” fared:

Just two out 10

And what about the other model in play:

Also just two out of 10

Something that can only signify an incredibly bad day at the office and with that in mind, here is how the “league table” of sorts looks at the moment

Pure Gut Instinct: 60/127 (47.24%)
PremBot 1: 25/59 (42.37%)
PremBot 2: 21/49 (42.85%)
Hybrid Model (Combining First 68 matches and then PremBot only): 62/117 (52.99%)

It’s slipping to ropey territory, which means I am going to run this until the end of the first half of the season and then carry out some analysis, in order to tweak things for the second half of the campaign, as this is not quite working as expected.

Admittedly this is a quick one, but we are in that period where there is more football than anyone knows what to do with and therefore, it is better to use this more as a developer article/data dump than writing war and peace on more incorrect data driven football predictions.

Happy punting and thanks for reading – to discuss this or any other projects you have, feel free to message me at either @dantracey1983 on Twitter or email me: [email protected]

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