Just One Away
Just One Away
If you read one of my articles from earlier in the weekend, you will know that the data driven betting project continued into the New Year and with that, it meant another raft of English Premier League Over 2.5 goals predictions were served up.
While with the final whistle blowing at Villa Park and the hosts being nothing short of massacred, it means it is now the perfect time in which to review who ‘GoalBot fared over the course of this extended weekend.
To do that, we need to first take stock of the six games that were recommended as bet friendly and without further ado, here they are:
HOW DID WE GET ON?
With Aston Villa shipping three to Manchester City before halftime, we know that one was in the bank and also with the first two fixtures of this weekend’s Premier League schedule ending Under 2.5 goals, we can tick them also.
This meant that there were three games still to play for on Saturday and by the end of the 3pm kick offs, the bet was still very much on as there were plenty of goals at both the King Power Stadium and Stamford Bridge
PARKING THE BUS
Unfortunately though, it was once again my beloved Tottenham who let me down and not only in terms of final outcome, as the thrashing that everyone was perhaps fearing before kick-off never materialised.
A win is a win for the champions elect Liverpool, however there was not a bundle of goals and that meant that ‘GoalBot’ did not return the clean sweep, meaning only five of the six recommended bets got over the line.
Admiittedly this is frustrating, as it has cost me a payout and although that in itself is galling, there is at least one crumb of comfort. That’s because the individual correct lines from this weekend, have at least bumped the overall success percentage.
From the 42 matches that I have tested this data driven model on, 32 have come back correct and that means a success rate of 76.1% – which is also an increase of 1.1% before the weekend’s Premier League action.
THE SAMPLE INCREASES
Therefore with the sample size getting bigger and the success percentage holding firm at more than three quarters, we can say that the model is working to a certain degree. However, we haven’t quite turned that into a financial conversion.
The reason for this, is that if you were to bet on all 42 lines individually, then the returns would all be miniscule ( that said, an overall profit would have probably been returned due to the odds of each game being on average 1/2 and we are at over 75% success)
However, the real trick is try and get the right combination and turn that into accumulator and this perhaps begs another question. Are we trying to big in terms of accumulator size, does it need to be scaled down.
Well if we were to try the old Countdown method of ‘2 from the top and 2 from the bottom’ then we would have still be caught out by Tottenham on Saturday, as they had the second highest Over 2.5 probability, so I guess like with any losing bets, we just have to put this down to ‘bad luck’
Still the show will go on and with the overall percentage rate increasing, it means the fundamental methodology is correct and for that reason we will peversere with another round of Premier League picks next weekend.
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 firstname.lastname@example.org. 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.