Villarreal’s early afternoon encounter with Girona this weekend might have only produced one actual goal, but its 21 shots, whether curling, dipping, scuffed or skied, helped to illustrate some interesting points about the expected variety.
Love it or hate it, xG is a useful indicator of a team’s attacking threat. The Yellow Submarine’s healthy reading of 2.51 on Sunday, in contrast to Girona’s 0.45, paints a fairly accurate picture of a game dominated by the Champions League-chasing hosts.
Two Villarreal penalties, however, each representing a value of 0.79, tell us that Sétien’s men were limited to just 0.93 xG from open play, perhaps suggesting some difficulty for the home side in carving out dangerous goalscoring opportunities. Sofascore’s shot map, showing nine efforts from outside the area, could also be used as further evidence to that claim.
What simple xG fails to take into account, however, is the quality of the shot that follows. Even if a player takes on a strike that 1/1000 players would score, what if he catches the ball so well, sending the ball flying towards the top corner, that only 1/1000 goalkeepers would save it?
This is where “Post-shot Expected Goals”, or “Expected Goals on Target” (xGOT), can change the picture dramatically. Combining the simple xG value of Villarreal’s four on-target, long-distance shots on Sunday, for example, yields a value of just 0.12. Take into account the quality of the actual attempts, however, and that tally rockets to 0.66.
Such a distinction is particularly useful for players like Álex Baena, who are not afraid to take on low-probability shooting opportunities. In fact, having fired in 30 shots this season, accumulating a simple xG value of just 1.73, only five LaLiga regulars have a lower average xG per shot than Villarreal’s all-action midfielder this season.
As such, while using simple xG might suggest a wasteful player, frequently attempting to score from unrealistic distances and angles, xGOT can help to quantify a player’s ability at turning low-probability scoring opportunities into higher-probability chances with the quality of their shooting.
In Baena’s case, his six shots and 0.22 simple xG generated against Girona translate to 0.82 xGOT, meaning that the 21-year-old was able to create almost four times as much danger than his shooting positions would suggest. And, extending this data to reach across the entire season, 1.73 xG turns into 3.60 xGOT, more accurately portraying the long-distance threat that his wicked ball-striking presents.
Comparing these two values can give us an indication, then, of a player’s shooting ability. Having increased his scoring probability by a multiplier of 2.104, it could be suggested that Baena has been the third-most dangerous shooter in LaLiga this season, proportional to the quality of his chances:
While this might all look convincing, then, there are still many problems that xG and xGOT are working to iron out. Goalkeeping coach John Harrison, who uses xGOT to measure shot-stopping performance, highlighted Kaoru Mitoma’s goal this weekend, to underline some of the issues with this ever-developing metric.
For our purposes, though, the emergence of Expected Goals on Target can help to tell us a few very useful things. Not least that players like Aleix García and Gonzalo Plata, as shown in the table above, are perhaps not as bad at shooting as their combined 37 shots and zero goals might suggest.
And so, while Álex Baena might not have scored this weekend, it might be wise for future defenders to close him down quickly, even if he is 30 yards from goal.