Fiorentina and Napoli played out a 1-1 draw in a match that didn't quite follow the script. Napoli went ahead early through Gilmour in the 23rd minute, set up by Politano, but Fiorentina pulled level after the break when Jimenez finished in the 51st minute following an assist from Ndour. That's how it stayed—a share of the spoils neither side could quite break.

Our model came into the match leaning toward Napoli with a 40% win probability, and had predicted a 1-2 scoreline. A draw sat at just 26% of our expectation, so we didn't call the result direction. Before kickoff, the expected goal model favored Napoli (1.76 to Fiorentina's 1.13), which fed that lean toward the visitors. In the end, though, both teams managed to find the net once each, and Fiorentina held firm enough to take a point rather than fold.

It's a solid example of why probability matters—the draw was always plausible, just not the most likely outcome. Our model underweighted this particular scenario, and while Napoli's underlying advantage showed up in that early goal, Fiorentina's second-half response meant the visitors couldn't capitalize on it. Sometimes the less favored result is exactly what happens.