Weather Temperature Market NO Betting with Gaussian Model

This strategy identifies temperature prediction markets on Polymarket where the YES (temperature will exceed a threshold) shares are overpriced relative to actual meteorological probability. Using live METAR station observations and Open-Meteo forecasts, the bot fits a Gaussian probability distribution to predict the likelihood of the event. If the market-implied probability of YES is significantly higher than the model-derived probability, the bot buys NO shares. Exit is at market resolution or when the price of NO reaches a target. The bot is currently paper-trading only, but initial results suggest a positive edge of 5-15% per trade. Implementation is in Python with automated data fetching and trade simulation.