A 5-0 at home, against a historic club of the German league, is the kind of result that ends up in headlines and in video reviews. Freiburg built it with nineteen shots, a dominance that was never in question and five goals that arrived almost all from situations that, normally, end up elsewhere.
The numbers tell of a wide and deserved win, but of a different size: 2.15 expected goals against 0.27. Freiburg played a good match and closed it with a scoreline from a match that will not repeat itself.
The verdict, before the numbers
Expected points say 2.72 against 0.17: with that performance the win was almost a given. Entries into the box are 6 against 3, total shots 19 against 5.
The 5-0, though, is not in any of these figures. It is in the fact that Freiburg converted five difficult shots, and that Gladbach in ninety minutes never once managed a shot with more than 0.09 of probability.
The number: 1.05
One point zero five is the sum of the five expected goals of the five goals.
Taken one by one: a header from a corner worth 0.042, a right-footed shot from the edge of the box worth 0.074, a header worth 0.710 in the second half, a left-footed shot worth 0.088, a right-footed shot from outside the box worth 0.138.
Four of the five shots that went in are below 0.14, that is below a probability of one goal every seven attempts. Only one, the header a quarter of an hour from the end, was a chance you would expect to be scored. Freiburg scored five goals with the material of a match worth one, maybe two.
Be careful how you read this figure, because this is where people usually get it wrong. It does not mean Freiburg were simply lucky: nineteen shots are a lot, and whoever shoots nineteen times sooner or later finds the evening when five go in. It means the 5-0 is not the measure of the difference between the two teams, it is the measure of an afternoon.
Gladbach, and the seventy-five minutes in which they did not exist
The 0.27 of the visitors is the truly alarming figure of the German matchday.
Five shots in the whole match, the best one worth 0.082, a header from the first corner of the afternoon. Then nothing more: two shots from outside the box that were blocked and two attempts in the final minutes when the result was already settled.
PPDA says 13.3 for Gladbach and 17.0 for Freiburg. These are high values for both, but they tell of two different situations: Freiburg did not need to go and press the opponents because the opponents were not coming out of their own half, while Gladbach tried to disturb without ever winning the ball back in a useful position.
After three matchdays the team has zero points, 12 goals conceded and 8.95 expected goals conceded. The difference between those two numbers, three goals, is the part of the season that depends on episodes. The rest, that is almost nine goals of chances conceded in three matches, is structure, and it does not get fixed with a save.
Freiburg, and the bill that is starting to grow
On the other side Freiburg have 9 points out of 9, 10 goals scored and only one conceded. The numbers underneath are good but less spectacular: 5.71 expected goals produced and 1.95 conceded.
Ten goals from 5.71 of chances means the team is scoring almost double what it creates. It is a real advantage in the table and a credit that the league, statistically, asks back. Expected points say it calmly: 6.85 against the 9 real ones, a gap of +2.15.
The interesting thing is that the defensive side holds up in the comparison: 1.95 expected goals conceded in three matches is a figure of a solid team, not of a team fleeing its own box. If up front the output drops towards the average, Freiburg do not collapse: they slow down.
The honest point
First, the sample. Three matchdays of Bundesliga are thirty or forty shots per team: every shot weighs enormously, and a header in or out shifts the whole picture. Everything written here should be read as a first impression, not as a verdict.
Second, the opposite reading, which deserves respect. You can argue that scoring five goals from difficult shots is a quality, not a coincidence: Freiburg have players who strike the ball well from outside the box and the model, which only looks at position and type of shot, does not know that. If that quality is real, the gap between actual goals and expected goals will still be there after ten matches too.
Third, what we think. This is a real win inside an exaggerated scoreline, and the figure that interests us most is not the 5, it is Gladbach's 0.27. A team that in ninety minutes cannot build even a tenth of an expected goal per shot has a problem that concerns neither the goalkeeper nor bad luck. It concerns the way it tries to come out of its own half.
T.L.
Do you want to see how much each team really deserves its league position? The Meritometro compares real points and expected points. For those who want models and not headlines, there is the Nerd Zone.