Minnesota United FC vs Orlando City SC prediction: 3 of 4 AI models back Minnesota United FC

All four experts converge on one fundamental insight: Bet365's pricing of Minnesota at 1.73 is built on a 5-1 away demolition of San Jose that does not reflect who the Loons are at home. The agreement is striking, and the underlying data is robust across a 10-game home sample — Minnesota average just 1.6 total goals per game at Allianz Field, have won only twice in ten home league matches, and have drawn four of their last five. Their home conversion rate sits at a miserable 6% despite respectable xG generation. This is not a small-sample blip; it is a persistent structural pattern under Cameron Knowles.

Resolving the Home Win vs Draw Tension

Three of four models predict a home win (2-1, 2-1, 1-0), while GLM-5.2 — the only model to break from the pack — calls a 1-1 draw. Weighting by historical accuracy complicates the consensus: Grok 4.6, the strongest performer across both 7-day (58% result accuracy, 52% pick win rate) and 30-day windows, predicts the lowest-scoring outcome (1-0) and emphasizes Minnesota's 1.6 goals-per-game home average most forcefully. Meanwhile, Claude Opus 5 (57% / 53%) and GPT-5.6 (53% / 50%) both predict 2-1 yet curiously recommend prediction against their own predicted result by backing X2 at 2.05 — an internal inconsistency that reveals low genuine conviction in the home win.

> The models clustering on a home win reflects a favourite-leaning tendency that the data does not support. When three of four predict the same result but two of them pick against it, the consensus is weaker than it appears.