Lyon and Monaco were the tidy reads; Toulouse and Marseille did the damage

Three calls held, and they all pointed to the same habit: trust the stronger side when the game state looks ordinary.

That was the useful part of the round. Lyon did not just beat Rennes; they did it with enough control that a 3-1 home-win read looked sensible before kick-off and even a little modest by full time. Monaco also justified a home lean against Lens, although only just. Paris FC over Strasbourg rounded out the clean work. None of those were glamorous calls. They were simply the sort of reads a model wants to get right without making a song and dance about it, which is usually how you know they were worth making.

Lyon was the best kind of hit: boring in advance, convincing afterwards.

The model liked Lyon at home because the scoreline profile pointed that way, and the match followed it almost too neatly. Ernest Nuamah Appiah scored twice, Corentin Tolisso added a goal and an assist, and Lyon never had to chase their own plan. That matters because this was not a case of one lucky bounce rescuing a forecast. The pre-match read on Lyon’s control held up in the data after the whistle: early lead, repeated pressure on Rennes, no need for heroics from anyone else. A prediction can be right for lazy reasons. This one was right for proper ones.

The miss that stung most was Toulouse against Le Havre.

Fipster leaned away from Toulouse and towards a narrow Le Havre win, which is exactly where caution became over-caution. The model overweighted the away-win shape and under-read how quickly Toulouse could turn the game once Casper Tengstedt got involved. He scored twice, including a penalty, Thomas Jørgensen added both an assist and the winner’s touch later on, and Le Havre still found two replies through Joshua Erowoli Orisunmihare Oluwaseun and Mbwana Samatta. In other words: this was not a neat tactical correction by the away side or some freak finish spree from nowhere. It was a bad read on who would seize control once chances started arriving in volume.

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

That is where this week really moved. Ernest Nuamah Appiah was everywhere for Lyon and had not been flagged pre-match. So did Casper Tengstedt for Toulouse, whose two goals changed one of the round’s biggest forecasts in a hurry. Ferran Torres decided Marseille against Paris Saint Germain with a goal that arrived from nowhere in Fipster terms; Marquinhos then finished off PSG’s win with another uncalled contribution. Elsewhere, Mohammed Amoura scored for Nice after an own goal had already tilted things their way, while Ayase Ueda gave Lille something back too late to matter for the forecast. The pattern is plain enough: Fipster was often reading team strength correctly, but several decisive scorers were not on its shortlist at all.

The next round needs less faith in tidy away wins and more respect for who is actually finishing chances.

The model did enough this week to stay honest rather than embarrassed: three correct calls in nine is not much comfort, but it is better than pretending every miss came from bad luck. The better lesson is narrower than that. When Ligue 1 gives you a strong home-side profile, take it if the underlying shape supports it; when it starts throwing up unfamiliar scorers like Ernest Nuamah Appiah or Casper Tengstedt, do not assume they are there to make up numbers. They usually have other ideas.