1. FC Heidenheim vs FC St. Pauli

Heidenheim made their early dominance count against St. Pauli, securing a decisive 2-0 victory with goals from Bashkim Zivzivadze in the third minute and Eren Dinkci in the 82nd. The home side's setup proved more efficient than the underlying numbers might have suggested, converting their limited chances while St. Pauli failed to create genuine scoring opportunities throughout the contest.

Our pre-match model predicted a narrow 1-0 Heidenheim victory, correctly identifying the likely winner but underestimating the margin by one goal. The Poisson framework generated xG figures of 1.15 for Heidenheim against 1.12 for St. Pauli—a near-parallel projection that reflected the statistical closeness of the matchup. What the model captured was the directional advantage; what it missed was the clinical finishing that would separate the sides beyond a single goal, particularly through Dinkci's late-match strike that settled the contest beyond doubt.

The early breakthrough proved decisive, with Zivzivadze's third-minute finish setting an aggressive tone that St. Pauli never fully answered. Heidenheim controlled possession in the middle periods without extending their advantage, but their defensive shape held firm throughout, limiting the visitors to peripheral efforts. The gap between expected and actual goals was modest here—roughly half a goal per team—reflecting a match where clinical execution at both ends mattered more than chance creation. For a prediction model operating on narrow probability margins with equal three-way confidence, identifying the correct winner represented the core success, even if the final scoreline required one additional data point to fully capture.