Seven Matchweeks In: A Team's Numbers Are Only as Good as the Opponents Behind Them

Every team average in the Eredivisie is really two things at once: how a team played, and who it played against. After seven matchweeks, the second half of that sentence has started to matter more than the first.

Matchweek 7 made the point four times in four separate matches. Each time, a number that honestly described a team's recent weeks failed to describe the next ninety minutes. The number was not wrong. The opponent was different from the ones that had produced it.

Some of those numbers carried FairEleven's own previews, which makes them the clearest place to see the mechanism: each preview wrote down in advance what it expected, and why.

The international break is a good moment to take stock. Seven rounds is enough to sharpen the questions. It is not yet enough to turn an attractive pattern into a fact about a team.

Willem II and Fortuna: the shape that broke where we said it would

Fortuna Sittard spent the first weeks of the season with very little of the ball. In their three matches of the first four rounds played eleven against eleven, against PSV, Cambuur and Groningen, they had between 24 and 36 per cent possession, as our four-matchweek review set out at the time.

Our preview of Willem II v Fortuna built its central position on that shape: "Willem II take more of the attempts and create more; Fortuna take fewer, better-struck ones." It also named, before kickoff, the one thing that would break it: Fortuna having more of the ball than expected.

That is exactly what happened on 19 September. Fortuna had 53 per cent of the ball, took 17 shots to Willem II's 15, and won 1–0. The chance count was close, 1.42 expected goals (xG) to 1.32, so a narrow away win was not a surprise in a match we had called open. The shape was.

The lesson is not that Fortuna have become a possession side. One match cannot show that. It is that their low possession figures described Fortuna against PSV, Cambuur and Groningen. Against a Willem II side with two points from its first five league matches, the same team had the ball for most of the evening. Whether that was Fortuna's choice or Willem II's concession, the figures do not say.

A possession share is not a fixed property of a team. It is an agreement between two teams, and this was a different pair.

Related analysis: Willem II v Fortuna Sittard

ADO Den Haag and PEC Zwolle: attacks measured against a different kind of opponent

ADO Den Haag arrived at Matchweek 7 with one of the clearest-looking numbers in the division. In six league matches they had never reached 1.20 xG; their best was 1.17. Our preview called that floor "the most durable thing in the fixture".

The same preview also listed where the floor came from. ADO's six opponents were, at the time, second, ninth, eighth, third, seventh and fifth in the table. Every one of them was in the top nine.

Cambuur were a different kind of opponent. At home on 19 September, ADO had 62 per cent of the ball, 18 shots to eight, six on target to three, three big chances to none, and 1.82 xG to 0.67. That was higher than in any of their previous six matches. They drew 1–1.

ADO's floor never showed that ADO cannot create. It showed that ADO had not created against the top half of the table. Those are different claims, and only the second one was supported.

The draw deserves the same caution in both directions. Cambuur took a point from a match in which they created almost nothing, and that point says very little about Cambuur. ADO created plenty and scored once: a note about finishing on one afternoon, not a new attacking level.

PEC Zwolle showed the same thing from the other side. Before Matchweek 7, their attack had looked respectable on the road, at 1.26 non-penalty expected goals per away match. At Groningen on 18 September it produced 0.58 xG, nine shots and a single attempt on target, in a 3–0 defeat.

An away average from a handful of matches is an average of those particular hosts. Groningen at home, with 17 shots and seven on target of their own, was a harder test than the ones that had built the number.

Related analysis: ADO Den Haag v SC Cambuur · FC Groningen v PEC Zwolle

FC Twente and PSV: when the sample is one match

The FC Twente v PSV preview had tried to do exactly what this article argues for: it looked at how each team performed against a particular kind of opponent. Its key observation was that "for both clubs, the match with the least possession was the worst defensive match of the season".

The problem was the size of the sample behind it. For each club, the lowest-possession match was a single away game. The preview said so, and then leaned on it anyway.

On 20 September FC Twente had 40 per cent of the ball against the league leaders and won 3–2, having trailed 2–1 at half-time. They out-shot PSV 21 to 11, with five attempts on target to four. Whatever that one away match had shown about FC Twente, it did not describe FC Twente against PSV.

A note on the figures: the data provider revised this match's expected-goals and big-chance numbers after we first recorded them, so they are left out here. Goals, shots, shots on target and possession did not change, and they are enough to tell the story.

This is why looking at the opponent is necessary but not sufficient. A split built on one match is not a pattern with context. It is an anecdote with a label.

What the match supports is narrow: FC Twente showed they can produce a heavy volume of attempts without the ball, against strong opposition. It does not yet show that they are a better side across the board.

Related analysis: FC Twente v PSV

What held up

Not every early number fell over. The ones that held tended to be the ones that did not depend on the opponent playing a particular way.

Ajax against Excelsior on 19 September is the clearest case. Ajax had 81 per cent of the ball, 33 shots to four and 3.53 xG to 0.59, and drew 2–2. Excelsior put two of their four shots on target, and those two were worth 1.39 expected goals on target (xGoT, which rates only the shots that reach the goal, taking into account where they were placed). Two goals from that is finishing far above the chances, not a change in either team.

What the match did confirm is a pattern that has now repeated: Ajax's goals keep trailing their chances. Their chance creation was never the issue here. Turning it into goals, and closing matches out, is. Ajax, like Willem II, have played six league matches rather than seven, because their meeting from Matchweek 3 has not yet been played.

SC Heerenveen's 4–0 win at Sparta was a real performance, and a smaller one than the score. Heerenveen had 23 shots, nine on target and 1.86 xG, while Sparta did not put one of their eleven attempts on target. But Heerenveen's shots on target were worth 3.47 xGoT: their finishing that afternoon ran well ahead of the chances they created. The direction is real. The scoreline overstates its size.

FC Groningen's 3–0 win over PEC Zwolle is the quieter example. The preview's direction and its moderate confidence held for the reasons it gave, on 1.59 xG to 0.58. It was a good evening, not a new level.

The same discipline applies to the league as a whole. Over the first five rounds, away sides scored 59 per cent of the Eredivisie's goals; in Matchweek 6 it was 57 per cent, and in Matchweek 7 it was 41 per cent. The share of goals scored after half-time went from 58 per cent to 63 and then to 41. Three samples, three different readings. Neither pattern is a property of this league, and nothing in the opposite direction should replace them.

Related analysis: Ajax v Excelsior

What changes before Matchweek 8

The break ends with Matchweek 8 on 9, 10 and 11 October. Three things from Matchweek 7 should shape how those matches are read.

A recent average should come with its opponents attached. When a team's figures come from a narrow mix of opponents (only strong sides, a few away trips, a single match), the next opponent has to be measured against that mix before the number is used.

The scoreline is the least reliable part of Matchweek 7. Heerenveen's win, Excelsior's draw in Amsterdam and ADO's point against Cambuur all say more about finishing on one afternoon than about the teams involved.

Some questions are now sharper. Fortuna Sittard, whose early results mostly came with little of the ball, host FC Twente, who have just beaten PSV with 40 per cent of it. How the ball is shared in that match will say more than either side's possession average. ADO Den Haag, whose attack was measured all season against top-nine sides, visit Telstar. PEC Zwolle, held to one shot on target at Groningen, host Cambuur.

Team news from the international break is not covered here. That picture will only be clear closer to the matches.

Seven matchweeks have not told us who is good. They have told us which question to ask of a number before believing it: against whom?

Data: Sportmonks, as recorded after Matchweek 7 on 21 September 2026. Analysis and interpretation: FairEleven.