Météo-France has filed a police complaint after two abrupt temperature spikes were recorded this month at a station near Paris Charles de Gaulle Airport. The first came on April 6 and the second on April 15. In both cases, readings increased by about 4 to 6 degrees in a very short period before falling again.
According to French media reports, these instances lined up with Polymarket contracts on Paris temperatures. Those contracts reportedly produced gains of roughly $14,000 in one case and about $20,000 in another.
What’s Beyond Two Winning Trades
The sums were not large by broader market standards. The more significant aspect is that the outcome appears to have depended on a physical data source that might have been tampered with. Météo-France said that it filed the complaint based on physical findings on one of its instruments and sensor data analysis. This takes the discussion away from merely strange market activity and towards the settlement process itself.
Prediction market platforms often present themselves as information tools as much as trading venues. Their credibility depends not only on liquid pricing. It is also shaped by confidence that the underlying event cannot be nudged by someone standing close enough to touch the source.
How the Suspicion Took Shape
After the first spike, a sensor problem or direct sunlight still looked possible. After the second, that became harder to defend. The jump on April 15 happened at about 9:30 p.m., after dark. Le Monde reported that weather observers discussing the case considered deliberate human intervention the most plausible explanation. In particular, they pointed to the possibility that someone briefly heated the area around the sensor with a battery-powered hair dryer. The station itself was described as easy to access from the roadside.
Polymarket’s Paris temperature market has since moved to a different reference station, at Le Bourget Airport. That does not resolve the underlying issue. However, that change suggests the original setup came under immediate scrutiny once the anomaly surfaced.
A French Market Problem Had Already Existed
The episode happened in a regulatory market that was already showing hostility towards prediction markets. France’s gambling regulator said these sites are not authorized in France and are considered illegal gambling sites. ANJ also said that major operators have been implementing geo-blocking from France upon its request, excluding the use of VPNs.
A prediction market doesn’t necessarily have to go through a major fraud scandal to lose its credibility. In some cases, an oddly timed data peak is enough to expose the most vulnerable part of this model. Once a reference source becomes accessible, manipulable, or subject to influences of any kind, the market built on top of it loses much of its neutrality.
