Chelsea, Tottenham, Manchester City, Liverpool, Arsenal and Manchester United – The Top 6. And then Leicester – who recently had their title-winning season, followed by a return to reality. Everyone knows Leicester’s fairy-tale, where they won the Premier League in 2016, followed by a more ‘normal’ 12th place finish this year. What links these 7 teams is that they were all in the top 10 for conversion rates in the last 5 seasons.
Whilst most of the former teams feature regularly in the list of highest conversion rates in the last 5 seasons, Leicester do not. It’s well known that conversion rate isn’t repeatable year to year and that it fluctuates highly, so I thought it’d be a good time to remind everyone. It’s annoying to still see tweets like this, when much better analysis could be performed:
Tweets like these can provide fun ‘nuggets’ of information, but if a player has a good/bad conversion rate, it may be due to the quality of his shots (which can easily be measured) or he is getting lucky or unlucky. So the only real use of conversion rates is to help identify if a team or player is particularly lucky or unlucky.
Back to the original point, Leicester had a conversion rate of 12.8% in their title-winning season. In the last 5 seasons, the average conversion rate has been 9.8%. In the following season, their conversion rates dropped to 10.6%, closer to the average. ‘Big’ teams can usually post high/above average conversion rates consistently, but this is because they will tend to create more high quality chances as they have better players.
This shows the conversion rates of teams in consecutive seasons from the last 5 seasons in the Premier League, showing that a team’s conversion rate one season is essentially independent of their conversion rate the following season. This is because conversion rates are not repeatable.
All of this also applies for shots on target. The average shot on target conversion rate is around 30%. This is touched on a lot more in these wonderful youtube videos by Joel (Twitter – @MessiSeconds) :
Lets look at conversion rates from teams last season to see if they can tell us anything. Chelsea had the highest conversion rate of 14.3%, and the 3rd highest in the last 5 seasons. This high conversion rate was probably most evident when Chelsea won 13 games in a row, and thus helped push them towards the title. At the other end of the conversion rate table, Southampton were bottom with a conversion rate of 7.1%. They still finished 8th, but next year may see them compete for the Europa League more seriously, and perhaps Claude Puel will have felt unlucky to lose his job. More interestingly, Bournemouth had a high conversion rate of 11.7% last season, which helped propel them to 9th at the end of the season. To demonstrate this. Below is the highest conversion rates from the last 5 seasons. This includes most of the usual suspects and Bournemouth from last season, plus the previously mentioned Leicester.
This is what I believe conversion rates should be used for in analysis, to see if a team is getting particularly lucky or unlucky, as if they are their conversion rate should return towards the mean. Conversion rate is not a good indicator of the quality of a team as it is not repeatable. I know this is generally well-known, but I thought this post would be a good reminder.