What the market asked
The Polymarket event asked a narrow weather question: what would be the highest temperature in Paris on 26 August 2026? Resolution depended on a single official source—the highest whole-degree Celsius value in the Temp column of the NOAA WRH timeseries for station LFPB (Paris-Le Bourget) on the local calendar day. Markets of this type typically split into discrete whole-degree outcomes, so the council had to rank which exact reading was most probable rather than simply call “hot” or “cool.”
What the council concluded
On report date 2026-08-26 the council (Gemini, Grok, and GPT, with Grok as chairman) issued a BUY_NO recommendation at a 21% probability and 72% confidence. Its stated reasoning was:
“Council recommends BUY_NO (NO). Probability: 21%, Confidence: 72%. Resolution is strictly the highest whole-degree Celsius value appearing in the Temp column of the NOAA WRH timeseries for station LFPB (Paris-Le Bourget) on the local calendar day 26 August 2026.”
Separately, and decisive for grading, the panel’s most-likely outcome was the specific contract “Will the highest temperature in Paris be 29°C on August 26?” at 48%. That single-degree call became the graded pick under the most_likely basis.
How the probabilities sat
The council’s internal distribution put nearly half its mass on 29°C while still treating the broader “NO” side of other framed contracts as the higher-conviction trade (21% implied probability on the side it bought, 72% confidence). In other words, the models were not claiming high certainty that 29°C would win outright; they were claiming it was the mode of the distribution and that certain adjacent or alternative claims were overpriced enough to fade. No external market mid-prices appear in the graded record, so the comparison rests on the council’s own 48% mode versus the binary BUY_NO stance.
What actually happened
The winning outcome was exactly the council’s most-likely pick: the highest temperature in Paris on 26 August 2026 resolved to 29°C under the LFPB NOAA rule. The graded pick was therefore correct. The Brier score on that call was 0.2704—respectable for a multi-outcome weather market where a 48% favorite still leaves substantial residual uncertainty across neighboring degrees. Because grading used the most_likely basis, the successful identification of 29°C, not the BUY_NO ticket alone, determined the win.
Takeaway
This was a clean hit on a source-constrained temperature market. The council correctly elevated 29°C as the mode, stayed disciplined about the official LFPB series, and accepted that even a well-supported single-degree favorite is far from a lock. Weather markets reward precise station rules and calibrated distributions more than narrative; here the models delivered both, and the scoreboard reflected it.
AI-generated analysis for informational purposes only. Not financial advice.