Over The Threshold

Over The Threshold

Over The Threshold

In my previous article I moved away from the quest to model the outcome of English Premier League matches and instead used a data driven approach to work out which of these weekend’s games will finish with either over or under 2.5 goals…….and then it hit me


With some many different over/under markets available and a ream of data available to me, it begged the question as to why not look at other thresholds and see if some other bigger value recommendations cannot be found along the way.

The question is though what level of probability is good probability?


If we take the Over/Under 2.5 goals model, I work on the basis that 50% is the marker for the over/under and from there I build in a 5% margin of error – this means that an over bet has to be at least 55% probability and an under bet has to be at least 45% probability for it to be recommended.

Therefore if we were to use that basis for Over/Under 1.5 goals – then I would end up recommending each and every Premier League game (pretty much), which would fall down on two fronts:

a) The use of a data driven approach would be pretty much futile
b) The odds being offered by the bookmakers would be relatively small.

If we were going to bet on over/under 1.5 then you would be looking at 90% probability as the benchmark, but to be honest it is such a futile bet, it is hardly worth sweating over on any given Premier League weekend.

Therefore we need to think a bit bigger and work out the best way to highlight games that may have a better chance of ending up with over 3.5 goals or more and the way to do this, will be on a single game basis rather than a gameweek.

Admittedly you could also offer an over/under 3.5 goals table, but again the odds on offer for games to finish under 3.5 would be in and around 1/4 (1.25 for all you decimal fans) and hardly worth the effort, especially if you are willing teams not to score.

The reason that over/under 2.5 is the betting benchmark is due to the fact that the average goals per season is approx 2.8 and that’s why this is the perfect balance between value of bet/likelihood of it getting over the line.

Whereas with anything either side of 2.5 goals, it does end up becoming more a game of risk =/= reward and although they do say when the FUN stops, stop. I’m going to advocate for more risk this time around.

Which means what is the trigger point for a game to recommended as over 3.5 – if we worked on the 50% principle, we could remove the margin of error and turn it into a literal flip of a coin (especially if the probability was exactly 50%)

If that is the case, then there is one game that hits that threshold this week and it is the Manchester City vs Manchester United clash at the Etihad, on the basis of all their league performances there is a 50/50 chance of this happening.

Which when you consider that the odds are 6/5 for this to happen, it means there is a slight hint of value in the Manchester Derby and if you were inclined to go bigger, then this stands out as the game for you.

To be honest, this one is more of a theory piece than solid tips/recommendations but I’d be interested to get other people’s feedback on this and work out the best way to use all of the goal data that I have available to me. If you have any ideas, please get in touch via the email below.

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 dan@realfootballman.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.

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

Leave a Reply

Your email address will not be published. Required fields are marked *