Weather Market NO-Side Overpricing Fade

How It Works This strategy (BallesJr/polymarket-weather-edge) identifies temperature markets on Polymarket where the YES probability is inflated relative to what a calibrated meteorological model would assign. The bot ingests real-time METAR (aviation weather) observations and Open-Meteo forecast data, fits a Gaussian distribution over the forecast temperature range, and computes the probability that the market's stated threshold will be met. When the market's implied probability for YES exceeds the model probability by a configurable margin, the bot enters the NO side.

Entry / Exit Conditions Entry: market implied P(YES) minus model P(YES) exceeds a threshold (edge margin), indicating the YES side is overpriced and the NO side offers positive expected value. The Gaussian model accounts for forecast uncertainty by using the standard deviation of recent METAR readings around the forecast. Exit: position held to resolution; no early exit is described, consistent with binary market mechanics.

Verified Results The repository is described as a "paper-trading" bot, meaning no real capital has been deployed and no verified on-chain results exist. The methodology is grounded in real meteorological data sources (METAR, Open-Meteo), which lends credibility to the modeling approach.

Implementation Notes METAR data provides ground-truth observed temperatures at reporting stations, enabling real-time model calibration. Open-Meteo provides free forecast APIs. The Gaussian model is a reasonable approximation for temperature distributions over short forecast horizons. The bot appears to be fully automated with GitHub Actions or similar. Key tunable parameters include the edge threshold and the lookback window for METAR standard deviation estimation.

Risks - **Model misspecification**: temperature distributions can be non-Gaussian (skewed during weather events); the model may underestimate tail risk. - **Station mismatch**: METAR stations may not correspond precisely to the geographic reference point used by Polymarket's oracle. - **Thin markets**: weather markets on Polymarket can have low liquidity, limiting position size and causing slippage. - **Paper-only**: no real performance data exists to validate edge.