Albian Ajeti and Thomas van den Belt made a mess of the map

The round started tidy enough, then spent most of its time undoing that impression.

Two STAR calls landed cleanly, which is the sort of thing that keeps everyone honest. The rest was less cooperative. Fipster leaned on a few obvious scoring patterns — home edge, established finishers, and the sort of player profiles that usually travel well in this competition — but League Stage - 2 had other ideas. Several matches turned on scorers who were not in the pre-match notes at all, and a couple of the model’s better-looking attacking reads were simply crowded out by different names arriving first. Not a disaster, but certainly not a gentle evening for anyone who likes their probabilities to behave themselves.

The Ajax call was right for the right reason

Ajax against Atalanta was one of the cleaner reads. Fipster called Toluwalase Arokodare as a scorer, and he did exactly what the model wanted: he scored late, after Ajax had already spent most of the night looking like they might need someone to prise the game open. That matters. The point was not just that his name appeared on the sheet; it was that the pre-match logic around him held up. He had been flagged as an efficient finisher with impact off the bench, and his goal arrived in precisely that kind of late-game window where those calls tend to pay off. The second Ajax goal came from Abdellah Ouazane, which was not forecast, but it did not change the basic verdict: the model had identified one of the players most likely to matter, and he did.

The miss came where Fipster trusted reputation too much

Getafe v FC Lugano was the one that hurt most because it looked so plausible before kick-off. Fipster backed Getafe to edge it and even found Kevin Behrens as a scorer, which at least gave the model one foothold when FC Lugano ran riot. But it over-weighted Getafe’s ability to control a close game and under-weighted how quickly FC Lugano could turn it into something else entirely. Albian Ajeti scored twice, Yanis Cimignani added another, Uran Bislimi joined in, and Borja Mayoral’s reply for Getafe came far too late to rescue anything except dignity. The real lesson is simple enough: when a side has several live attacking routes, betting on one neat home script can be expensive. Football does enjoy making tidy forecasts look faintly decorative.

The round’s decisive names were not always the ones Fipster circled first.

Some of the best individual returns belonged to players who were already on Fipster’s radar. Toluwalase Arokodare scored for Ajax after being called pre-match; Cyriel Kolawole Dessers did likewise for Panathinaikos with two goals; Michele Šego also delivered for HNK Hajduk Split. Those are useful hits because they show the model can still find a live scorer when form and role line up properly. But this round also belonged to players outside the original shortlist. Albian Ajeti was not called and still scored twice for FC Lugano. Thomas van den Belt was not called and scored twice for Twente, while Wout Weghorst added another for good measure. Daouda Weidmann got in on Twente’s act too, again without being flagged beforehand. Elsewhere, Noah Holm rewarded belief with three goals for Brann after being named pre-match, while Paris Josua Brunner kept Monaco moving with an opening goal before Lamine Camara and Ansu Fati finished SC Freiburg off.

+3 on one side of the ledger is useful; 11 misses on the other is harder work. Still, there is something usable here for next time: when Fipster gets a scorer right in this competition stage, it tends to be because role and recent output are aligned rather than because reputation alone looks shiny on paper. The next round will ask for more patience with late scorers and less faith in neat scorelines that assume one team will keep things orderly for 90 minutes. That is rarely how these evenings go, which is rude but efficient.