Fipster’s play-off read was lively, but the round kept changing its mind

The broad shape was there, just not the detail.

Fipster leaned towards home wins in most of the round, and that part of the picture was not absurd. Three of those calls landed cleanly enough to count as ticks. The trouble was that this stage of the Champions League is rarely polite about margins. Several ties drifted away from the exact scoreline, and a few were turned by names the model had not put on the page. That is how you end up with a round that feels half-right and half-embarrassing at once. Very football.

The best call held because the scorer profile held.

Sabah FA against Hapoel Beer Sheva was one of the cleaner reads. Fipster backed Sabah to win and singled out Joy-Lance Mickels among the likely scorers. That part came off. Mickels scored, and his performance was strong enough to sit among the standouts as Sabah eventually got through after extra time. The model did not get every twist of the tie right, but it did identify a forward who mattered and a home side with enough attacking threat to keep asking questions. In other words: useful scouting, even if the match refused to stop at 90 minutes like a well-behaved fixture should.

The miss was mostly about trusting the wrong shape of game.

The clearest wrong turn came in Lyon against Fenerbahçe. Fipster expected a narrow home win for Lyon and built that around Fenerbahçe attackers such as Talisca, Ainsley Maitland-Niles and Tyler Morton contributing in front of goal. None of those calls landed. Instead, Vedat Muriqi and Archibald Brown scored for Fenerbahçe before Malick Fofana pulled one back for Lyon. The model over-weighted Lyon’s home edge and missed where the decisive goals would come from. That is annoying in hindsight, but also instructive: sometimes the match does not reward your neatest theory; it rewards whoever turns up with sharper finishing.

The round belonged to players Fipster only partly saw coming.

Among the names that shaped things, Veljko Simić was one of the few pre-match calls to pay off across both legs for Sabah FA, while Joy-Lance Mickels also justified his billing with a goal in the extra-time win. Beyond that, though, plenty of damage came from outside the forecast: Igor Zlatanović scored twice for Hapoel Beer Sheva without being called; Mason Greenwood scored for Fenerbahçe; Haitam Aleesami scored for Bodo/Glimt; Fredrik Bjørkan and Sondre Auklend both found goals in Bodo/Glimt’s second-leg win; and Malick Fofana decided Lyon’s side of matters when Fenerbahçe had been expected to carry less threat. The pattern is clear enough. Fipster found some of the right attackers, but this round was often decided by secondary scorers stepping into primary roles.

What comes next is simple enough to say and harder to do.

'Play-offs' are exactly where tidy probability can look foolish very quickly. The model still had enough structure to find three correct outcomes and a couple of useful scorer reads, which is better than guessing with confidence and calling it analysis. But it also showed where low-confidence home-win logic can become too blunt when knockout ties start bending after 90 minutes or when unexpected scorers take over. The next round will ask for more restraint on exact scorelines and more respect for late-game chaos. Football remains committed to being awkward on purpose.