With the new Premier League season only less than a month away, we look back at the previous campaign and use the expected goals metric to see which teams overperformed the most and what it means for next season.
Expected goals, also referred to as xG, “measure the quality of a shot based on several variables such as assist type, shot angle and distance from goal, whether it was a headed shot and whether it was defined as a big chance”, according to football statistics database Opta.
The xG values are measured using historical data points from millions of shots, which is how, for instance a penalty attempt is typically valued as 0.76 xG. Over the course of a season we can see how teams and players perform relative to their xG and xGA (expected goals allowed). So with the help of Understat’s expected points table, we’ve discovered which teams were the most fortunate to finish the season with as many points as they did.
There is an element of “if my aunt was my uncle, but she’s not” to this, but it is a decent indicator of a team’s overall performance and perhaps a good predictor for how they will do in the future.
5. Crystal Palace
Actual position: 14th – 43 pts
Expected position: 16th – 38.26 xPTS
Finishing marginally above Burnley and Sheffield United in the xPTS table, Crystal Palace’s placing probably has a lot to do with their post-lockdown form. The Eagles were the worst team in the league after the restart, losing seven of their nine matches. According to Understat they finished with 4.74 points more than they should have expected to.
Roy Hodgson did a terrific job once again to keep Palace not only safe, but punching above their weight. At one point they were looking odds-on to finish in the top half of the table and better their best ever finish, but slid back down to 14th. They will have to refresh their aging squad if they are to avoid a relegation battle next season.
Actual position: 8th – 56 pts
Expected position: 9th – 50.15 xPTS
Last season Arsenal finished a full 11 points above where they should have, according to the underlying numbers. The Gunners felt that predicted regression and more this year, as they won 14 less actual points compared to 2018/19. And it could have easily been worse.
The xPTS table shows that Arsenal finished 5.85 points above their expected points total, which is a little jarring given last season was their worst campaign in 25 years. The scored 56 goals from 50.82 xG and conceded 48 from 57.25 xGA – so they were basically bang average this season while also over performing.
The North London outfit will be sorely hoping Mikel Arteta really is the man to rejuvenate them in the long-run, because their numbers have been bleak these last couple of years.
Actual position: 6th – 59 pts
Expected position: 12th – 49.26 xPTS
According to the xPTS table, Tottenham Hotspur should have finished below their North London rivals, not above them. In fact they shouldn’t have even ended up in the top half apparently.
Spurs were abject in the final few months of Mauricio Pochettino’s reign, playing well below the standard they had set in the previous five years. Jose Mourinho’s appointment saw an improvement in terms of results, although not everyone is pleased with the style of play. He has a massive job ahead of him to return this team to the top four.
2. Newcastle United
Actual position: 13th – 44 pts
Expected position: 20th – 31.92 xPTS
No club had a bigger discrepancy between their actual and expected positions in the league this year than Newcastle United, who should have been relegated according to Understat.
Steve Bruce has clearly worked some sort of wizardry to get as much out of this squad as he did, considering they looked so weak in attack for a large part of the season, while their defence didn’t exactly look resolute either. Finishing a full 12 points above their expected tally, the Magpies could find themselves regressing into a relegation battle next season.
Actual position: 1st – 99 pts
Expected position: 2nd – 74.28 xPTS
And now we come to the real anomaly and outlier of this list, Premier League champions Liverpool, who supposedly finished a whopping 24.72 points better off than the numbers suggest they should have. Such a swing would have Manchester City the title winners for a third year running.
Does this mean that the Reds are somehow undeserving of their title or simply more lucky than good? Not even a little. It is not unusual for great teams and players to outperform their underlying statistics – that’s what makes them truly great! Their level of elite finishing, incredible defending, or world class shot-stopping are what separate them from the rest. Add in a bit of luck and you get a winning formula.
Still, a difference of almost 25 points is absolutely massive, so what gives? As Statsbomb founder Ted Knutson recently said, it’s entirely possible that Liverpool are gaining advantages that aren’t being picked up by the xG model.
“At a certain point when you have had the second and fourth-best Premier League seasons ever and the expected goals models don’t really reflect that, maybe you’ve got some stuff going on that is beyond the model,” Knutson told the Guardian.