Chelsea and Hull City played out a thrilling 2-2 draw at Stamford Bridge, defying our pre-match forecast of a comfortable 3-1 home win. Rogers opened the scoring for Chelsea in the seventh minute, assisted by João Pedro, but Hull hit back hard. Belloumi levelled just before the half-hour mark—a neat finish from Ryan Giles's assist—then doubled his tally four minutes later, this time teed up by Oliver McBurnie to leave Hull 2-1 up at the break. Chelsea equalised through João Pedro in the 66th minute, Palmer setting him up, but couldn't find a winner despite the pressure that followed.

Our model assigned just a 21% probability to the draw, backing Chelsea to dominate with 66% win odds. That misjudgement came from overweighting Chelsea's home record and historical dominance in this fixture—eight previous wins with an average of three goals per game—while underestimating Hull's ability to stay compact and hit on the counter. The gap between our 3-1 prediction and the actual 2-2 scoreline tells the story: Hull showed far more attacking threat than the pre-match data suggested, particularly through Belloumi's clinical finishing, while Chelsea created chances without clinical efficiency. Both teams' attacking potential was on display, but neither could land the knockout blow. It's a decent reminder that historical head-to-head records and home advantage, while valuable indicators, don't always account for the specifics of how a single match unfolds.