Groningen 2–3 Fortuna Sittard: AI predicted 2–1 ✗

Synthesis: Groningen vs Fortuna Sittard

Where the Experts Agree

All four models converge on a 2-1 home win for Groningen, and the convergence is more than superficial. Each arrives at the same scoreline through independent reasoning: Groningen's home scoring reliability (2+ goals in both home matches, both ending 2-1), Fortuna's defensive fragility (2.71 xGA per match, no clean sheets), and the suspension of Ole Romeny gutting Fortuna's transition attack. The 2-1 scoreline is not a compromise — it's the natural product of a home side that scores freely but concedes, against an away side that can nick a goal but cannot keep one out.

Resolving the Confidence Gap

The models disagree meaningfully on confidence, ranging from GPT-5.6's 52% to GLM-5.2's 62%. This is not noise — it reflects a genuine analytical divide. GPT-5.6 argues that Fortuna keep games close by design, citing their 2-2 draw at PSV (despite only 0.98 xG) and Branderhorst's shot-stopping ability. This is a legitimate concern: Groningen conceded 2.12 xG to Utrecht at home, and their centre-back pairing was exposed at PSV.

This calibration aligns with the structural warning — these models tend to be favourite-leaning, so when all four cluster on the same home win call, the true probability sits at or slightly below the consensus.