Samuel Surridge and Cavan Sullivan gave the round its best arguments, while the model spent too much time admiring the wrong forwards

The round had a clear idea of where goals would come from. It just kept choosing the wrong address.

Fipster was broadly right about the league’s attacking mood: this was not a week for cautious scorelines or tidy little 1-0s. The problem was selection. It found four correct scorer calls across the round, which is respectable enough in isolation, but too many of the other picks were built on reputation and recent volume rather than the specific matchups that actually mattered. That is how you end up with a model that sees danger everywhere and still misses the sharpest knife in the drawer.

The cleanest hit was Samuel Surridge doing exactly what the pre-match case suggested he might do.

Inter Miami against Nashville SC was the best call because it was built on something sturdier than name recognition. Samuel Surridge had been flagged as a live scorer, and he backed that up with both Nashville goals, including the late equaliser that turned a likely home win into a draw. The model also had Lionel Messi Cuccittini down as a threat, and he scored too. That matters because this was not lucking into one obvious star while missing everything else; it correctly identified the two players most likely to decide Nashville’s end of the story. The own goal in between only made the shape of it messier, not less convincing.

The miss at San Diego came from trusting the wrong kind of form.

This was where Fipster over-weighted San Diego’s usual attacking profile and under-read Philadelphia Union’s actual edge on the day. Anders Laustrup Dreyer and Milan Iloski were both in the pre-match frame, but neither scored; instead Cavan Sullivan took over with two goals, Bruno Damiani added another, and Philadelphia ran away with it. The model had Cavan Sullivan in its notes, which at least means it saw him as part of the picture, but it did not give him enough weight to explain a result that became lopsided very quickly. That is the difference between spotting a threat and understanding who is about to take charge of a match.

The names that mattered were not always the names Fipster liked best before kick-off.

Samuel Surridge was one of the round’s clearest vindications: called pre-match, then rewarded with two goals and a standout rating. Lionel Messi Cuccittini also justified his billing by scoring for Inter Miami. Marco Reus did his part for Los Angeles Galaxy by converting from the spot, which at least kept one Galaxy call alive when their broader home win prediction did not. Christian Ramirez deserves a mention too: Fipster named him for Vancouver Whitecaps against Austin, and he answered with both Austin goals. That one stings slightly more because it shows the model can identify an individual threat even while getting the result badly wrong; useful knowledge, just deployed in service of defeat.

There is still something useful here if it stops flattering familiar scorers.

The round’s lesson is not that Fipster should guess less boldly. It is that boldness needs better target selection. When it trusted Samuel Surridge or Marco Reus or Christian Ramirez in the right context, it looked sharp enough. When it leaned on bigger attacking names without enough regard for how those games were actually unfolding, it drifted into expensive optimism. Next week should be simpler: keep backing players whose role matches their form, keep listening when an under-the-radar scorer keeps turning up in the data, and stop assuming every familiar forward will do his bit just because he usually does.