Five goals and not a single real chance

For an hour, at the Mapei, almost nothing happened. Then in the last half hour five goals arrived, Juventus turned the deficit around and then saw it turned around again deep into stoppage time, and everyone left the stadium talking about a crazy match.

The numbers say something else, and for once it is something very simple: nothing happened for an hour because nothing was happening, and the five goals came from finishes that on a normal day all end up wide or in the goalkeeper's arms.

The verdict, before the numbers

The expected goals for the match are 0.96 for Sassuolo and 1.18 for Juventus. A total of 2.14: a poor afternoon of chances, spread over 28 shots.

5 goals came out of it. It is the highest ratio between what was scored and what was created across the entire European matchday this weekend, and there aren't even any matches close to it.

The number: 0.32

Zero point three two is the value of the most dangerous shot of the match. Berardi hit it shortly after the half hour of the first half, with his left foot, and the goalkeeper saved it.

In the whole game nobody else went above that threshold. Sassuolo's two goals from open play are worth 0.30 and 0.09, Juventus's first 0.31, the second 0.05, and the Emilian winner in stoppage time 0.12. Adding up the five shots that went in you get to 0.87: less than one expected goal for five real goals.

A big chance, in our count, starts at 0.30. Here there are three in ninety minutes, and they are three borderline chances, not the kind that get the crowd on their feet. The match everyone will remember for the flood of goals is, measured with the yardstick of chances, one of the poorest of the day.

How we got to that hour of nothing

Sassuolo played the match it had in mind, and its choice can be read in one number: PPDA 25.4. That means it let Juventus make 25 passes before going to bother them, which is to say almost never. Low block, narrow lines, no intention of going to win the ball in the opposition half.

Juventus, on the other side, pressed at more than double that, 12.6, and had the ball as much as it wanted. The problem is where. Eighteen shots for 1.18 expected goals means an average of 0.07 per shot: shots from distance, deflected efforts, headers from a tight angle on corners. Juventus never found the door in.

The entry-into-the-box figure confirms it on its own: 7 against 4, and it is the hosts who have more, despite having played while waiting. A team that waits and still gets into the box more than the other is a team that has prepared its counterattacks well.

When Sassuolo took the lead, shortly after the hour, it did it exactly like that: ball won back, immediate ball forward, shot from the edge of the six yard box. Zero point three zero, the most that match was willing to give.

The last half hour, and Juventus who don't come back

From there the script broke three times.

Juventus equalised with a left-footed finish after an aerial duel, 0.31, their most dangerous shot of the ninety minutes. Then Sassuolo put their head back in front with a shot worth nine hundredths. Then Juventus equalised again on the counter, 0.05, five hundredths, meaning one shot in twenty of that kind goes in. Then, deep into stoppage time, Sassuolo won it with a right-footed shot worth 0.12.

Put bluntly: in the last half hour four finishes went in that, added together, are worth half an expected goal. This is not a judgement on the two teams, it is a snapshot of how anomalous the evening was. And anomalous evenings have one characteristic: they don't repeat.

There is one thing, though, that is not random, and it concerns the periods in which the two teams produced. Sassuolo built almost all of its 0.96 in two moments, the siege in the first half around the half hour and the final quarter of an hour. In between it let Juventus move the ball around without danger: a PPDA of 25 means exactly this.

The season context

Juventus come out of matchday four with 7 points from 9.60 expected. It is the third biggest negative gap in Serie A, and it all comes from the last two weeks.

There is one thing that makes the case even more curious. In four matches the bianconeri have conceded 1.62 expected goals, the lowest figure in the league: Roma, who have conceded just one goal, sit at 2.97. Juventus's defence is the most solid in Serie A for the amount of danger allowed, and meanwhile it has shipped 4 goals. Almost everything that went in was, on paper, a shot worth nothing.

Sassuolo, by contrast, have 7 points from 5.78 expected and 8 goals scored from 7.04. It is a team that concedes a lot, 6.62 expected goals in four matches, but that has found a way to hurt teams vertically. Its problem, if anything, is the opposite: after two months of matches like this one, the bill comes due.

The honest point

First, the opposite reading, which is legitimate. Anyone who says expected goals don't see the quality of the player has a serious argument right here: five difficult finishes out of five went in, and it didn't happen by cosmic accident. Some players strike the ball better than others, and a model that assigns five hundredths to a shot on the counter is telling you what the average player does, not what the one who hit it does.

Second, the limit of the sample. Four matchdays are not a verdict. Saying Juventus have 2.60 points fewer than what they have produced means saying they have been unlucky so far, not that their season will be good: if they keep conceding 0.40 expected goals per match, those points come back on their own.

Third, what we think. This match says almost nothing about Sassuolo and says quite a lot about Juventus, who pressed high, had the ball, never got into the box and lost to a team that gave them everything except the twenty metres that matter. The solidity is there. What's missing is a way in.

On Sunday at 18:00 Juventus host Atalanta, the other Italian team with an open account between real points and performances. Sassuolo go to Monza on Friday evening.

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's the Nerd Zone.