In Simon Rolfes, the Bundesliga and its partner AWS have an expert who knows the Bundesliga from A to Z. Bayer 04 Leverkusen's Sporting Director, Rolfes wore Die Werkself's colours 288 times between 2005 and 2015, and represented Germany on 26 occasions. Rolfes is the ideal man to write a regular column for the 'Bundesliga Match Facts Zone' on bundesliga.com, analysing current trends and giving unique insight on the Bundesliga Match Facts.
This week, Simon Rolfes turns the xG model on its head and, rather than focusing on it from an attacking point of view, he considers the number of goals teams would have been expected to concede.
By Simon Rolfes
The Expected Goals (xG) model has become an integral part of football reporting. The shots teams are taking and the probability that each will end in a goal are compared with the actual number of goals teams are scoring. Well, today we want to turn things around and look at xG from another perspective: how many goals would the Bundesliga teams have been expected to concede based on their opponents' shots on their goal, and how often have goalkeepers actually had to fish the ball out of their own nets?
Thanks to xG Against (xGA), the following table shows quite clearly which teams have been conceding the most opportunities in front of their own goal, and the difference between xG Against and actual goals conceded.
Quite a few things catch the eye immediately: the by far lowest xGA value belongs to RB Leipzig (24.8). This means that the hardest team in the Bundesliga to score against are the side currently sitting in second place in the standings. What is also interesting is that Leipzig have conceded just 25 goals, which is almost identical to their xGA. In fact, eight of the 18 Bundesliga teams have an xGA within three goals of the actual number of goals they have conceded, which shows how good the model works.
Yet once again, exceptions confirm the rule and the largest difference can be seen with the already relegated FC Schalke 04. They have conceded almost 16 goals more than would have been expected of them, and that can easily account for five or six defeats. Hertha Berlin and Borussia Dortmund could also have been in a much better league position when their xGA is considered.
Watch: xG - an explainer
VfL Wolfsburg and FC Augsburg both cut fine figures in terms of xGA. Oliver Glasner's men are fourth in this classification, yet they have conceded the second fewest number of goals in the league this season and therefore have the largest positive difference when compared to xGA. Augsburg have also evidently been particularly good at scuppering their opponents' best opportunities at the last minute.
So what are the reasons for this? One explanation could be that, in Koen Casteels and Rafal Gikiewicz, they both boast extremely good goalkeepers among their ranks. Reducing xGA just to the ability of the goalkeepers would not do full justice to the complexity of the model, however. Indeed, it is the behaviour of the whole defensive line which contributes to a team's success or failure in terms of xGA.