Brighton absolutely dismantled Arsenal at the Amex, running out 3-0 winners in a result that caught our model well short of the mark. We'd leaned heavily towards an Arsenal victory, assigning them a 53% win probability, while Brighton came in at just 27%. Instead, the hosts delivered a dominant display that our pre-match numbers didn't anticipate.

The goals came in quick succession either side of the break. Pascal Groß opened the scoring in the 31st minute with an assist from Charalampos Kostoulas, then Kostoulas himself got on the scoresheet just before halftime when Diego Gomez set him up at 45'. By the time Chema Andrés added the third in the 57th minute—Groß supplying the assist this time—the contest was already done. Before kickoff, our model had expected something much tighter, built on the expectation that Arsenal would dominate possession and create chances while Brighton would sit deep. Form suggested Arsenal's xG of 2.29 would prove decisive against Brighton's 0.7. That clearly didn't materialize. Brighton were sharper, more clinical, and far more threatening than the pre-match data had suggested they'd be. There's not much more to it than that—sometimes a team simply outperforms its underlying metrics. Our model saw this as an outlier outcome at just 27%, and Brighton proved us wrong.