Rennes, Auxerre and the late scorers made a mess of the model’s neatest ideas

The round started with a few tidy reads and then spent most of its time undoing them

Fipster found three correct calls in a nine-fixture round, which is enough to show the model was not guessing in the dark, but not enough to pretend it had the room under control. The wins it did land were narrow ones: Auxerre over Nice, Rennes over Marseille, and Lille over Estac Troyes. That matters because the misses were not all of one kind. Some were simple scoreline errors. Others were worse: the wrong side of a draw, or a home win that never arrived because the match was decided by a scorer nobody had put on the board.

Auxerre over Nice was the cleanest call Fipster made

This was the one where the model’s basic read held up best. It backed Auxerre at home, and Auxerre won 1-0 through Cameron Archer’s second-half finish from Christ Makosso’s assist. The useful part is not just that the result landed; it is that the shape of it matched the pre-match logic without needing any embellishment. A tight home win stayed tight, and the decisive moment came from a player who had not been called before kick-off. That is still a miss on scorers, but it is a decent outcome for the match view itself: Fipster got the direction right even if it did not name the man who supplied the final touch.

The worst miss was Rennes against Marseille, because the model trusted volume over restraint

This was where Fipster over-weighted a high-scoring home-win picture and under-read how little room there would be for it. Rennes did win, but 1-0 rather than anything like the 4-2 shape predicted. Adrien Thomasson scored the only goal, and both goalkeepers had proper work to do: Brice Samba with five saves for Rennes, Jeffrey de Lange with six for Marseille. That tells you why this one went wrong. The model leaned into an open game; the actual match was controlled by defences and keepers until one moment broke it open. When a game finishes with those save totals and only one scorer on either side of your forecast ledger, you have not merely missed by a margin — you have misread what kind of contest it would be.

The round belonged to players Fipster never called out before kick-off

The clearest pattern across the goals was how often an unforeseen scorer settled things. Cameron Archer did it for Auxerre. Adrien Thomasson did it for Rennes. Ferran Torres opened Paris Saint Germain’s win at Brest. In Le Mans against Lens, A. Bourabaa and Lucas Calodat both scored for Le Mans while Abdallah Sima and Franjo Ivanović answered for Lens; none of those names had been flagged pre-match either. Lorient’s Ibrahima Baldé also scored without a prior call, as did Noah Mbamba-Muanda in that same draw with Toulouse. Even Monaco’s equaliser came from Paris Josua Brunner after Strasbourg had gone ahead via an own goal from Sadibou Sané. The round’s scoring was therefore less about modelled favourites doing their expected work than about individual moments arriving from outside Fipster’s shortlist.

That does not mean every surprise was wild luck in disguise. Matvey Safonov’s 8.5 rating in Paris Saint Germain’s win at Brest shows why some low-scoring away results can still be built on one outstanding performance rather than attacking fireworks; he helped make Fipster’s defeat there look more decisive than chaotic. Dominik Greif did something similar for Lyon in Paris FC’s goalless draw, which is another reminder that some misses are really about goalkeepers turning a forecast into paperwork. The next round will ask whether Fipster can separate sturdy low-event matches from games where one uncalled attacker is waiting to ruin everyone’s afternoon.