Estoril and Arouca played out a goalless draw in a match that saw neither side break through, leaving the pitch locked at 0-0. It was a cagey affair that failed to produce the goals our model had anticipated heading into kickoff.

Before the match, our model leaned toward an Estoril win with a 2-1 scoreline, giving the draw just a 26% chance. That probability wasn't negligible—it was a plausible outcome among several—but it wasn't where the numbers pointed. The underlying metrics favored Estoril slightly on expected output, with the Poisson model projecting an edge in their xG profile. The data suggested both teams would create chances and that goals would come. They didn't. The match unfolded in a way that sat well outside our main expectation, and while stalemates happen in football, the model had weighted a more open game.

So we got it wrong on both the result direction and the exact score. That's the reality of prediction in football—even when the underlying metrics tell a story, the ball doesn't always cooperate. Estoril's xG advantage on paper didn't translate into chances converted, and Arouca's defensive shape held firm. It's a reminder that low-probability outcomes land regularly enough to keep everyone honest, and 0-0 draws, though less common than wins or losses, remain part of the sport's texture.