Felix Nmecha and Jonathan Burkardt gave Bundesliga Round 9 its sharpest correction
Seven calls landed, two did not, and the round’s clearest lesson was that the model was better at reading the broad direction than the exact scorer. Felix Nmecha’s brace for Borussia Dortmund and Jonathan Burkardt’s late work for Eintracht Frankfurt were the names that changed the tone of the weekend.
The round mostly went where Fipster pointed, but not always by the route it expected.
The broad call was sound: home sides were still favoured in most places, Bayern München were backed to edge a tricky away trip, and the stronger squads generally did what stronger squads tend to do. But this was not a neat parade of exact scorelines. Several matches landed a goal or two away from the forecast, which is usually how a decent model looks in practice — close enough to be useful, not so smug that it starts asking for a statue.
Borussia Dortmund was the cleanest hit, even if the scorer list went missing
The best call came at Borussia Dortmund against SV Elversberg. Fipster had Dortmund winning comfortably and, more importantly, had read the match as one where their attacking depth should tell over time rather than in one early burst. That part held. The scoreline finished shorter than predicted, but the shape of the game still matched the pre-match view: Dortmund controlled enough of it to win without needing a frantic finish. The model missed every named scorer there, which is awkward on paper, yet the result itself still sat inside the logic of the forecast. Sometimes you are right about the match and wrong about who gets the applause.
The miss was Köln against Bayer Leverkusen, and it came down to trust in the wrong end of the pitch
The hardest miss was 1. FC Köln against Bayer Leverkusen. Fipster leaned toward a home win and built that case around Leverkusen’s attacking focal points doing enough damage for Köln to be edged out. Instead, Leverkusen took control through Aleix García early, then let Patrik Schick settle it late before Said El Mala made it look closer than it had been. The model over-weighted Köln’s ability to turn this into their sort of game; what actually decided it was Leverkusen’s ability to score first and then keep enough composure to survive until Patrik Schick could finish it off. That is a different kind of control entirely.
The round belonged to players Fipster did not fully see coming into view
The player story was where this round got interesting. Felix Nmecha scored twice for Borussia Dortmund after being absent from Fipster’s pre-match scorer list; that is not just a tidy correction, it is a reminder that midfield runners can ruin an otherwise sensible attacking map. Jonathan Burkardt did something similar for Eintracht Frankfurt with two goals at SC Paderborn 07, again without being called beforehand. Christopher Nkunku also turned up as an unforeseen scorer for RB Leipzig and added an assist besides, which rather underlines why he remains such an irritatingly useful footballer for anyone trying to pin him down in advance.
The next round should reward patience more than bravado
The useful takeaway is not that Fipster was perfect — it plainly wasn’t — but that its strongest reads still travelled well across a busy nine-fixture slate. When it leaned on team strength and match control, it usually found something close to truth. When it tried to name every scorer in advance, football did what football does best: handed the credit to someone else. That is no disaster. It is just a reminder that prediction works best when it stays humble enough to leave room for Felix Nmecha.
Who decided the round
Called by Fipster pre-match vs who actually scoredPatrik Schick
Bayer Leverkusen
1goalCalledPhillip Tietz
FSV Mainz 05
1goalCalledYuito Suzuki
SC Freiburg
1goalCalledFelix Nmecha
Borussia Dortmund
2goalsUnforeseenJonathan Burkardt
Eintracht Frankfurt
2goalsUnforeseenChristopher Nkunku
RB Leipzig
1goalUnforeseenJoshua Kimmich
Bayern München
1goalUnforeseenChristoph Baumgartner
RB Leipzig
1goalUnforeseen