OpenFIFA's multi-agent decision hands a single World Cup market to six specialist frontier models that debate it in the open and return an explainable, directional call — not just a cold probability.

Five specialist agents each own one lens: the Odds agent reads market consensus, the Quant agent runs probability models, the Team-News agent tracks injuries and lineups, the Tactical agent weighs style matchups, and the Skeptic agent hunts for risk and contrarian evidence.

The debate runs in three rounds: each agent first states an independent view, then cross-examines the others' reasoning, and finally converges on the evidence. The Skeptic casts no directional vote and acts purely as a risk gate, so the panel never blindly follows the favorite.

The Judge agent synthesizes every argument into a final call, a confidence level, and the key reasons — and states plainly whether there is real edge versus the current price.

Every model sees the same inputs: order-book snapshot, bookmaker consensus, Elo, injury intel, and tactical data — and every reasoning step is public, so you can judge who to trust.

“A good prediction isn't picking the winner — it's finding the truth the price has underpriced.”