Weather Temperature Market Betting (NO Side)

The bot scrapes real-time METAR weather observations and Open-Meteo forecasts to compute an implied probability distribution for daily high/low temperatures. It compares this modeled probability to the market-implied price of YES tokens on Polymarket temperature contracts. If the market price is significantly overpriced relative to the Gaussian model (i.e., the probability of a YES outcome is lower than 40% but token price is above 50 cents), it places a NO bet. Entry is on the hourly market refresh, exit at market resolution or if the price moves to fair value. Paper-trading results indicate 70%+ win rates on overpriced temperature markets. Implementation uses Python with statistical libraries (scipy for Gaussian fitting) and websocket feeds for live data.