Manchester City ground out a 1-0 win over Coventry in a match that proved far tighter than the pre-kickoff script suggested. Erling Haaland broke the deadlock in the 26th minute with a finish from Antoine Semenyo's assist, and that single goal proved enough to settle it. City controlled possession for large stretches but couldn't find the cutting edge to run up the kind of scoreline that looked plausible on paper beforehand.

Our model had leaned toward a 3-1 City win, backed by a 72% probability of a Manchester City victory. The result direction came off—City did win—but the actual scoreline was far more subdued than expected. The prediction flagged Coventry's likely approach of sitting deep and absorbing pressure given their precarious league position, and that defensive posture clearly worked better than anticipated. City's home form had looked prolific in the buildup, yet they couldn't convert that into the volume of chances our model expected, while Coventry's shaky away record didn't translate into the defensive collapse the numbers had hinted at. The gulf between the prediction and reality highlights how margin calls at the sharp end of football are genuinely hard to nail, even when you've got the winner's direction right.