What does xG predict for the rest of the season?

With more than half the season gone, we have enough stats to get a good idea of how well Premier League teams have been playing and therefore how things will likely unfold in the months ahead.

Expected goals tells us how well teams do the basics of football by creating lots of good chances up front and limiting their opponents to few, difficult ones at the back.

With the figures based on how Premier League teams have performed in the past, history shows that the higher a team’s xG difference – the more they dominate games – the more chance they have of success.

So what do the stats tell us? If sides keep performing at around the level they have so far, then…

The good news for Arsenal fans is that the team with the highest xG difference has won the Premier League in three of the past four seasons.

The exception was in 2023-24 when Manchester City won the title despite Arsenal having the best underlying numbers.

In the race for the Champions League, it might surprise you to see Manchester United rank so highly in xG given that Ruben Amorim has just been sacked and his 32% win rate in the Premier League has been well publicised.

If we compare teams’ xG difference with the points they’ve taken this season, we can clearly see in the chart below that the majority of sides’ results align closely with their general performance.

Aston Villa and Sunderland have almost certainly achieved their season objectives already, with Opta giving Villa a 96% chance of Champions League football and the Black Cats a 99% chance of survival.

This isn’t about whether the two sides ‘deserve’ to be where they are in the table. The fact that they’ve scored and kept out the goals they have means they have earned those positions, thanks in part to Villa’s elite long-range finishing and Sunderland keeper Robin Roefs being the best shot-stopper in the league.

But what it does show is that this season’s success is built on shaky ground and there’s still work to be done if Villa wish to establish themselves at the top of the Premier League long-term and Sunderland are to avoid suffering second-season syndrome next season.

Villa fans will rightly point out that their team has regularly overperformed their xG under Unai Emery, while Sunderland fans will claim that you can’t measure tenacity and togetherness.

At the other end, while Wolves are playing like a team worthy of relegation, they have been far better than the two-points tally they had until recently so it’s no real surprise their results have begun to improve.

Two examples of why teams can only cheat their xG for so long are Tottenham and Leeds this season.

Early results misleadingly suggested that Thomas Frank’s team were strong Champions League contenders while Daniel Farke’s side were in for a long, tough struggle.

But Tottenham’s xG was far worse than their points tally implied, while Leeds’ was much better.

Expected goals isn’t perfect and there are aspects of football that can never be captured on a spreadsheet but history shows that it has a great track record for predicting a team’s long-term success.