In the course of the present Premier League season, Pep Guardiola’s Manchester City have been imperious. Having lighted upon a perch at the top of the league, it is looking less and less likely that any other club will knock them off it this season. Unbeaten through 14 games in all competitions, it appears that the Spanish manager has finally cracked the Premier League.
But just how good are they? Jon Mackenzie introduces the concept of Expected Goals and explains how it reveals that Manchester City are well worth their place at the top of the English league pyramid.
A Sense of Injustice
If you have ever been to a football match, you will probably have had the feeling at one point or another. You know the one – that feeling of injustice. Your team lost but it felt instinctively like they should have won or, vice-versa, your team won but you can’t help but thinking that you were a little bit lucky.
Usually what we’re doing when we go through this process is totting up all the chances that have occurred during a game, assessing the likelihood of them being scored, and comparing them against the opposition.
Of course, with such an inexact methodology – ‘what I think I saw’ – the chances are that your assessment won’t be that accurate. But in general, the average fan has a good chance of being able to go through this operation.
But what about over the course of a few games? Of a whole season? And how do you iron out the subjectivities that sneak into our assessments?
Fortunately for us, a number of football statisticians have developed a statistical measure that allows us to go through this process with a little more objectivity. They call it Expected Goals (which is often shorted to xG – a figure you might have noticed on the bottom of your screen on Match of the Day this season).
How good is a chance?
How do you decide how good a chance is? There are a number of factors to take into account. For example, if Sergio Aguero was taking the shot, you might feel as though the chance was better than if James Tarkowski found himself in the same situation.
Beyond this, you would have to factor in whether or not the player was taking the shot with their weaker foot, their head, whether the ball was crossed in or passed to them, whether there were any defenders in front of them and, perhaps most importantly, where abouts on the pitch the shot was taken from.
By taking all of these factors into account, football analysts have been able to develop a metric which assigns a percentage value to each chance. A shot with an xG value of 0.7 means that, on average in the Premier League, you would expect the goal to be scored 70% of the time. On the other hand, a less clear-cut chance would have a lower xG value.
Having assigned a number to chances, you can now go through the process of assessing a game by totting up all the chances made by a team with more accuracy, giving some sort of mathematical clout to your post-match ruminations.
What is more, though, you can now aggregate the xG values from a single team across a whole series of games to see how good the chances they are creating stack up over the course of the season.
Who’s the best?
Applying this process to the Premier League, Understat have drawn upa table of all the teams in the league comparing their combined xG values for (that is, the quality of the chances that they have created in their games), their xGA values (that is, the quality of the chances that they have faced during their games) and their xPTS values (that is, the number of points they could have hoped to have got based on the quality of the chances created and the chances faced).
Taking these values, so the logic goes, you should be able to see which team has created the best chances in the season, the team who has faced the best chances in the season and the team who has the biggest gap between the two.
Looking at the table, you can see that, according to the Expected Goals column (xG), Manchester City have the highest value so far – a figure of 26.83. This indicates that, statistically, you would have expected City to have scored 26.83 goals this season. Given that they have 32 goals in the Premier League so far, you can see that they are performing slightly above average in their finishing.
At the bottom of the table, perhaps surprisingly, you find Burnley who have an xG value of 5.38. With only 8 goals to their name, this shouldn’t be as shocking as it seems although, once again, you’ll notice that they are still finishing slightly above their xG rate.
Looking at the next column, Expected Goals Against (xGA), we should now be able to see which teams have been performing best and worse defensively in the Premier League.
Obviously, given the nature of this statistic, the table has been reversed and now Crystal Palace find themselves top of the pile with an xGA value of 17.26. Compared to their actual goals conceded of 19, they can count themselves slightly unlucky not to have conceded slightly fewer.
Once again, though, Manchester City are the best performers with an xGA value of only 4.99 which maps onto their goals conceded figure of 4. According to these statistics, then, City have had the best attack and defence in the league to this point.
Turning to the final column, you can now see the way the Premier League table would look if the games had gone the way that the chances created suggested.
This is very interesting when you look at the bottom of the table. On this metric, Crystal Palace find themselves in 13th place with an xPTS value of 10.31, a fact that suggests they are mightily unlucky to find themselves bottom of the table.
With this being a full 7 points more than they are currently on, Roy Hodgson will take comfort from the fact that, with a little luck in their favour, his team could turn it around and find themselves climbing up the table again.
However, once again, Manchester City are the clear winners. Top of the table with 22.85 xPTS, they could count themselves lucky to have a points total of 25.
So there you have it. Applying statistics to football shows that Manchester City are deserved league leaders. It took Pep Guardiola a year to get there, but it appears from the stats that it was well worth the wait.
What do you think of the use of statistics in football? Let us know in the comments below
Want to share your opinion? Why not Write For Us?