What the market asked
The Polymarket event covered the winner of the 2026 Oklahoma Governor Republican primary. Under Oklahoma law, a candidate needs a majority to take the nomination outright; otherwise the top two advance to a mandatory runoff. The graded leg of the report focused on one named candidate market: whether Mike Mazzei would be the eventual primary winner.
The council published its view on 2026-08-23, two days before the scheduled August 25 runoff.
What the council concluded
All four models—Gemini, Grok, Claude, and GPT—participated, with Grok acting as chairman. The council’s action was BUY_NO on Mazzei winning. It put the probability of that NO side at 62%, with 70% confidence in the recommendation.
The published reasoning stated:
“Council recommends BUY_NO (NO). Probability: 62%, Confidence: 70%. Resolution criteria are straightforward: the market pays the overall winner of the Oklahoma Republican primary for governor, which under state law requires a majority and therefore includes the mandatory August 25 runoff between the top two finishers when no…”
In short, the panel treated the market as a clean majority-plus-runoff contract and judged Mazzei more likely than not to fall short of becoming the nominee. That made the Mazzei contract their graded pick: 62% on NO (implied ~38% on YES).
How the council framed the price
The report does not include a full external order-book snapshot, so the cleanest comparison is internal: the council’s own 62% NO / 38% YES split on Mazzei. The action was explicitly directional—BUY_NO—rather than a passive probability dump. Confidence was logged at 70%, high enough to take a side but not presented as near-certainty. The emphasis on runoff mechanics suggests the models were discounting a first-round knockout and weighing Mazzei’s odds across a two-stage process.
What actually happened
Mike Mazzei did not win the 2026 Oklahoma Governor Republican primary. The graded outcome therefore matched the council’s BUY_NO. The call was marked correct on the most-likely basis. The Brier score was 0.1444, which lines up with a 38% YES forecast against a binary outcome of 0 (Mazzei did not win): (0.38 − 0)² = 0.1444.
Takeaway
This was a process-aware call more than a personality call. By anchoring on Oklahoma’s majority-and-runoff rule and publishing two days before the runoff, the council avoided treating the primary as a simple plurality race. The 62% NO was not extreme; the Brier score shows the forecast was usefully calibrated rather than lucky at the extremes. When resolution criteria and calendar constraints are clear, a moderate lean against a named candidate can still be the right side of the market—provided the models actually respect how the race is decided. The public grade stands: BUY_NO on Mazzei was correct.
AI-generated analysis for informational purposes only. Not financial advice.