Scoring 20+ goals in the premiership – part II

Part I was posted on Saturday and concluded that to score 20+ goals in a premiership season a player has to be good and to some extent lucky. Firstly they have to be healthy – the group averages almost 35 games per season. Secondly their goals per game ratio shows a marked increase from the seasons before and after. My theory is that this improvement is largely due to an unsustainable increase in shooting percentage, partially explaining the goals per game ratio being lower the seasons before and after.

Now I’m going to split the players who’ve scored 20+ into two groups. The first will be the players who managed to score 20+ in back-to-back seasons and the second is the players who did not follow up one 20+ season with another. They happen to split perfectly, with 17 seasons worth of data for each group.

The first group comprises of undeniably excellent players (Shearer, Hasselbaink, Henry, van Nistelrooy and Tevez*). Realistically there probably isn’t much that this group will teach us about the rest of the playing population as they’re simply in a class of their own. Below is outline of their performance the seasons prior to and following a 20+ season.

 Year Games Goals Goals per game Season prior 30.9 19.4 0.65 Season of 20+ 34.5 24.2 0.70 Season after 29.4 19.1 0.65

There isn’t much year-to-year variation. An increased number of games played coupled with a slight improvement in the goals per game ratio and a season of almost 20 goals the year before becomes a 20+ season next time round, or vice versa.

It’s the second group intrigues me more. Scoring 20+ is a genuine achievement, is there anything in the previous years performance of these players that we could use to predict them having what is essentially a career year?

 Year Games Goals Goals per game Season prior 32.0 10.2 0.32 Season of 20+ 35.1 24.1 0.69 Season after 26.5 9.2 0.35

Now we’re getting somewhere interesting. The differences are pretty remarkable, there is no way of predicting what these players are about to do from looking at their performance the season prior. Sure these players benefit from being healthy during their 20+ seasons compared to those prior and following but they also double the rate at which they score goals. There’s two explanations I can think of and I think both probably have an effect. Firstly an outrageous and unsustainable (over a two-three year period) shooting percentage. Whilst it isn’t sustainable to have a high shooting percentage over a full season the smaller sample size of a single player means there are more likely to be individual outliers. The second explanation that comes to mind is playing on a team where the tactics depend on getting the ball to that particular player.

I think this data highlights a couple of things we can learn. Firstly if you have a Shearer, van Nistelrooy etc. then keep hold of them. Guaranteed goals in the premiership aren’t easy to come by and they perform consistently. Generally managers seem to have got this right Henry was sold to Barcelona and did little from that point onwards, same for van Nistelrooy and Hasselbaink. The Tevez situation was unique and Shearer didn’t want to leave, so no manager would have the bottle to sell him.

Second if you are the manager of a player who’s come out of nowhere to score 20+ it is a great time to cash in and move on (see Saha as one example). The odds are that they overperformed during that season and they’ll regress to their career performance given time.

Finally if you play any fantasy football, steer clear of those ‘shock’ high scorers from the year before, no matter how tempting they are. They’re guaranteed to be overpriced because everyone else is expecting (at minimum) a repeat performance.

*strictly speaking Tevez needs 2 this season to make it back-to-back but barring injury he’ll get them so I’ve included him in this group