Bayern München 7–0 Union Berlin: AI predicted 3–0 ✓

A mismatch the market has fully absorbed

All five models converge on a Bayern win, but what matters is why they converge — and the reasons here are unusually concrete rather than reputation-based. Union Berlin arrive at the Allianz having conceded 10 goals in three league games (0-4 at Leverkusen, 1-3 at home to Schalke, 3-3 with Frankfurt), sitting 16th with a single point. Critically, GPT-5.6's underlying numbers show this is not a finishing accident: Union have allowed roughly 7.00 expected goals in three matches, meaning opponents are repeatedly reaching high-quality chances. Bayern, by contrast, have generated about 7.07 xG while conceding only 2.49, and have scored 13 in their last three home league games, including 5-1 over Stuttgart and a 5-0 Champions League win over Bodø/Glimt.

Where the experts disagree — and how it resolves

The only genuine fault line among the five reads is the clean-sheet question, which separates the 3-0 camp (Grok, Kimi, GLM, GPT) from Claude's 3-1. The case for Union scoring: they have found the net in each of their last five competitive matches, created 2.21 xG even in defeat to Schalke, and Bayern's defence has been sloppy in transition — five conceded in their last three home league games, plus goals shipped at Elversberg and a 0-0 blank at Schalke. The case against: Union's only away fixture against a top side this season produced 0.46 xG created and 3.47 conceded at Leverkusen, and with Burke, Ilic, Imeri and Markgraf out and Friedrich and Nsoki doubtful, they lack even the counter-attacking outlets needed to punish Bayern's rest defence. On balance, the away-evidence view wins: Union's attacking numbers came from chaos-favouring home games, not trips to Munich, so 3-0 edges out 3-1 as the single most likely scoreline — but the margin between them is genuinely close.

The decisive structural factor

The non-obvious driver, flagged independently by Claude and GLM, is that Union are no longer the deep-block side that historically frustrated Bayern. Under Carles Martínez they press higher and play out from the back — and the by-product is a 3-3 and two heavy defeats in three games. A patched-together back line (potentially three centre-back options missing) facing Kane between the lines and Olise and Díaz wide is a structural mismatch, not just a quality one. Meanwhile Bayern have a full week's rest, no rotation incentive with the next European fixture a week away, and a near first-choice XI available — GPT-5.6's projected front six of Kimmich, Pavlovic, Olise, Musiala, Díaz and Kane is a materially different creative level from anything Union have faced.