Data metrics explained: Expected Goals (xG)

D1 Arkema Text graphic for The Data Scout by Marc P. Lamberts.

You only have to open up Twitter, read an article online or watch football on the TV and you can’t miss it — expected goals.

Often used by analysts and pundits, yet the majority either don’t use in an appropriate way or do not fully understand it. In this article, we will try to explain what the expected goals metric is, and how we can use it when analysing games, clubs or players.

Expected goals explained

The Analyst explains it very well.

“Expected goals (or xG) measures the quality of a chance by calculating the likelihood that it will be scored from a particular position on the pitch during a particular phase of play. This value is based on several factors from before the shot was taken. xG is measured on a scale between zero and one, where zero represents a chance that is impossible to score and one represents a chance that a player would be expected to score every single time.”

We see this in game after game. Team A has an xG of 2,1 and Team B has an xG of 1,73. Based on the quality of the chances, the likelihood of scoring was bigger at team A than it was with team B.

This however, doesn’t mean that Team A should have won. Expected goals measures the quality of the chances and the likelihood of scoring. However, it does not predict the expected outcome of the game — that’s a whole different metric.

It’s not as simple as saying that a player ‘should have scored’ a particular chance because the xG was high. Similarly, if the chance is missed, the player is not necessarily putting in a bad performance. Expected goal consists of different factors. These include distance to the goal, angle, one-on-one, big chance, body part, type of assist, and pattern of play.

Based on these variables — and others — we can calculate the expected goals of a certain player

How do we work with it?

It’s all about the narrative and the bigger picture. If we are looking at one game, you can see how much a team was likely to score from the quality of the chances — interesting, but hardly convincing. On another day the team would have overperformed or underperformed their xG.

This could benefit the writing or audio-visual media. If we look at it from a coaching or analyst point of view, single match xG doesn’t really tell us a lot about the long-term performance of a team.

Analysis within clubs deal a lot with trends. If a club is scoring more than their xG might suggest over a period of more than 10 games, you can conclude that the ones taking the shots have quality. And in doing so and converting those chances.

However, you can also anticipate that at one point the scored goals will be lower because the quality of the chances suggest that. The same goes for a struggling team that is underperforming. In the long run, they will catch up with their xG and start converting more chances. Another important detail to add is you have to look at the number of shots and the xG per shot. You can have 5 shots that equal 1,80 xG or 30 shots that equal 1,80 xG. It’s all about the context.

The expected goals metric is a great tool to measure the likelihood of scoring a chance. It gives you an idea of the shots tha’should’ be scored. It also gives an idea of how your attack/defence is doing in terms of a longer period. And, of course, whether the performance matches the end product. But all of this shouldn’t be taken as gospel. It’s a tool that needs to fit a narrative, a context and the playing philosophy.

MORE from Her Football Hub: