REJECTED — Judges did not all approve

Iran-Oman Hormuz Agreement by December 31?

View market on Polymarket →

Why it looked attractive

New leg among the top events by volume (24h volume $5,324); picked by code for a first check.

Why it was rejected

Council round2: 1/3 approve (ruleset v4). Objections: [openai/gpt-6-astra] REJECT (confidence 8): Objection 1 remains unresolved: withdrawing the unsupported base rate and market validation is appropriate, but the replacement 35% estimate has no checkable calculation or calibration linking the cited reports to that probability, and “above a coin-flip baseline” does not explain a 35% estimate. The rebuttal also upgrades an announcement “expected soon” into a scheduled event without providing supporting evidence. Objection 2 is resolved by explicitly retracting the conflict claim and correctly distinguishing the corridor and fee tracks, although the rules permit a corridor-only agreement and do not require fee agreement. [anthropic/claude-opus-5.5] REJECT (confidence 6): Objection 1 (no sourced base rate, improper market validation) is conceded but not fixed. Dropping the anchor leaves 0.35 as a bare assertion with no derivation. The rebuttal is also internally inconsistent: it says the evidence supports "a probability meaningfully above a coin-flip baseline," then sets 0.35, which is below a coin flip. Objection 2 (no real conflict between Items 1 and 2) is conceded and resolved by retracting the "conflicting reports" framing. However, the replacement mechanism misreads the rules: a corridor-only agreement qualifies on its own, so the rule does not require both tracks. With the edge now only about 5 points over the 0.30 ask and the estimate's foundation still uncheckable, I no longer have grounds to approve.

Considered betting
YES (it happens)
Our estimate
38% chance YES
Would have paid
30¢ per YES share
Market
29% chance YES
Date
today
Key figures
  • Iran-Oman Hormuz Agreement by Dec 31: yes 0.29 / ask 0.30 on Polymarket as of this scan
Evidence sources
Council verdicts
VerdictModelConfidenceCost
REJECT GPT-6 Astra 7/10 $0.0532
APPROVE Claude opus 5.5 6/10 $0.0343
APPROVE Kimi K3 7/10 $0.0165
REJECT GPT-6 Astra 8/10 $0.0431
REJECT Claude opus 5.5 6/10 $0.0356
APPROVE Kimi K3 7/10 $0.0333