Great Expectations

A fortnight ago, we introduced the idea of a goal expectation model, using shots as our main variable. The aim was to highlight players who were scoring fewer goals than expected given the quantity and location of their shots.

We can now apply this methodology at a team level.

The Premier League season might be in its infancy (be aware of sample size error) but we can start to build a match result expectation model based on a team’s number of shots per game.

If we treat every game as separate event and run a Monte Carlo Simulation, we can assign a probability for each result (win-draw-loss) based on the number of shots each team had during the game.

By treating home and away games separately, we have created the following expectations for this weekend’s match.

Crystal Palace v Chelsea 18 October

Key
shots = Actual shots per game
ExpG = Expected goals from shots (home/away team)
goals per g = Actual goals per game
home supremacy = ExpG (home) minus ExpG (away)
win Kelly = Difference between forecast & actual odds

When examining the results, just because there is a positive Kelly figure doesn’t mean we should automatically run to the nearest bookmaker to place a wager. Statistics are the starting point of our research not the end game.

To be successful we need to analyse what the model is saying.

Crystal Palace v Chelsea
Model Recommends: Crystal Palace win at 15/2

Chelsea have started the season in a rich vein of form but much of their success has been down to the prolific Diego Costa; his current strike rate of nine goals in seven games sounds unsustainable.

There are currently seven players who average more than Costa’s 3.1 shots per game and although the Spaniard takes most of his shots from inside the penalty area (86%) his 41% scoring rate is insane. 

Compare him to the hotshot Daniel Sturridge of last season and that strike rate is ready to regress. Sturridge managed 3.4 shots per game, 65% inside the penalty area but with a ‘modest’ 21% scoring rate. 

As regards Crystal Palace, they are difficult opponents on their home patch. The Eagles won as many as they lost last season (eight) and this campaign they are showing a positive home supremacy of 0.23.

With Chelsea benefiting from Diego Costa’s unsustainable goal scoring form and Crystal Palace no pushovers, a home win is overpriced at 15/2 (BetVictor) with Palace also a fair bet at 21/10 (Boylesports) in the double chance market.