Weather-Poly Arbitrage Bot (3-category bucket)
A production-ready Python bot exploiting systematic edge in Polymarket's daily weather bucket markets across three categories (high, low, precipitation) using weather forecast vs market price discrepancies.
This bot analyzes Polymarket's daily weather bucket markets (high temperature, low temperature, and precipitation) and compares the implied probability from market prices to objective weather forecast data from the National Weather Service. When the market price implies a probability that is statistically different from the forecast-based probability, the bot places a trade.
Entry: The bot triggers when the absolute difference between implied probability and forecast probability exceeds a dynamic threshold (e.g., 5%). It bets YES if the market underprices the event, NO if overpriced. Exit: Positions are held until market resolution (next day) or until the mispricing corrects.
Verified results: No verified PnL; the repo is marked as production-ready but lacks backtest data. It does have a well-structured codebase with 427 KB of Python.
Implementation notes: Uses Python, CLOB API, and weather APIs. Risks include forecast error, market manipulation, and low liquidity on weather markets.