Everton vs Sunderland

Sunderland's clinical performance at Goodison Park delivered a convincing 3-1 victory that proved considerably more comfortable than the pre-match data suggested. Marcus Rohl's 43rd-minute opener gave Everton hope of the draw we'd anticipated, but the visitors responded swiftly through Bruma Brobbey's 59th-minute equaliser. From there, Sunderland's attacking intent overwhelmed a home side that had appeared vulnerable throughout. Emil Le Fée added a second in the 81st minute before Will Isidor sealed matters in the 90+1st, leaving Everton with no realistic route back into the contest.

Our model predicted a 1-1 stalemate with only a 15% probability assigned to a Sunderland win—a significant miss. The prediction was grounded in reasonable foundations: both teams' mid-table positioning suggested limited attacking motivation, Everton's leaky home record paired with Sunderland's defensive solidity pointed toward a low-stakes draw, and recent head-to-head history supported the outcome. What we didn't capture was Sunderland's capacity to be genuinely dangerous when given space. While their away form showed limited scoring output (0.93 goals per game), they proved far more clinical here, converting chances when created rather than the predictable stalemate the data favoured.

Everton's failure to build on Rohl's early advantage exposed defensive vulnerabilities that the pre-match metrics underestimated. Sunderland's second-half intensity was the decisive factor—a tactical intensity that Everton couldn't match. The gap between model prediction and actual result highlights how situational momentum and in-game adaptation can override statistical baselines, particularly when one team shows greater willingness to press their advantage.