Biggest model-vs-market gaps on Kalshi
Where the NOAA weather model disagrees most with the Kalshi implied probability. Sorted by absolute edge. This is research data — every market is live, every probability is real.
Biggest gaps between model and market
| Market | Market price | NOAA model | Edge | Lean |
|---|---|---|---|---|
| Will the **high temp in NYC** be <54° on Apr 19, 2026: 53° or below | 48¢ | 100.0% | +50.0pp | NO |
How the edge is calculated
Each row on this page is a live Kalshi weather market with a current NOAA ensemble probability from our tracking system. The "edge" column is the raw difference between the model's estimate and the market's implied probability.
A +30pp edge on a YES contract means the NOAA ensemble sees the event as 30 percentage points more likely than the Kalshi market does. That gap could close because: (a) the model is wrong, (b) the market is slow to incorporate the forecast, or (c) the event resolves and the wrong side was pricing it.
This page is not trading advice. Prediction markets price in information the model cannot see (breaking news, insider information, recent weather observations past our scan time). Treat the edge column as a research signal, not a recommendation.
Data sources
- Market prices: Kalshi public trade API — the implied probability is the midpoint between YES bid and YES ask.
- NOAA probabilities: Our internal weather model ensemble, blending GFS and ECMWF forecasts for the market's target location and strike.
- Refresh: Market prices refresh every several hours; NOAA signals refresh as new forecasts are published.