(Over)performing Lions: Can England keep defying data and beat France at World Cup?
It is easy to see why bookmakers have anointed France, the reigning champions and home of the best player at this World Cup, Kylian Mbappe, as favourites for Saturday’s quarter-final against England.
But should they? Based purely on performances in Qatar over the last fortnight, England can make a strong case of their own. They are unbeaten, no side has scored more goals, and their defensive record is better than France’s.
That reading could be regarded as a little superficial, however, and a deeper dive into the underlying metrics – using data supplied by analytics company Sporting Risk – reveals a more complex puzzle for Gareth Southgate.
The most glaring wrinkle in the numbers is that England are massively outperforming their expected goals (xG) at this World Cup. In layman’s terms, they have scored many more times per game (3) than they ought to have (1.58) based on the quality of chances created.
To a lesser but nonetheless significant extent, the same is true at the other end. England have shipped an average of 0.5 goals per game in Qatar versus an xG conceded of 1.0 per game; in other words, Jordan Pickford has been beaten half as many times as might be expected.
There are a few possible explanations for this divergence. It may be that England have been incredibly efficient, and other data suggests that is certainly true in attack, where their shot conversion rate of 27.9 per cent is the best of the 32 teams.
On the other hand, luck or poor performance by opponents can play a part, and that is a more likely explanation for their defensive success. Either way, outperforming xG to a large degree tends to be unsustainable so it may catch up with England against France.
Perhaps more concerningly, England’s underlying metrics are actually worse than France’s in both attack and defence. Although they have scored fewer and let in more goals, Les Bleus have better xG at both ends of the pitch at this World Cup.
France are particularly dangerous when crossing and counter-attacking, the data shows. They rank first of the 32 teams for headed goals, accurate crosses and accurate cutbacks, and second for goals from counters, ball carries and dribble distance.
Those who fear that Mbappe will punish England’s defensive deficiencies will not find much reassurance in the numbers. Southgate’s men rank 30th of 32 for interceptions and saves, and dead last for successful defensive actions per game.
If England are to expose a weakness in the French rearguard then it may be through pressing. They rank much higher than the holders in that respect, and for the frequency of winning the ball in their opponents’ final third of the pitch.
Opportunity could also knock at set-pieces, where France rank fifth worst for goals conceded, and down England’s left flank. Didier Deschamps’ side concede more attacks down their right, where Ousmane Demebele offers scant protection to makeshift full-back Jules Kounde.
On balance then, the data indicates that France’s favouritism is probably justified after all. They have played better, even if they have less to show for it so far. But a World Cup is a small sample size, so some caution is advised and it is equally possible that England ride the crest of their current wave and continue to outperform underlying metrics all the way to the final.
Sporting Risk is a sports analytics company whose expertise lies in predictive analytics, forecasting and betting on football. The company leverages the predictive modelling of its data science team to generate outputs for the media, gaming and professional football sectors. www.sportingrisk.com