Weather Edge Paper Trader
A paper-trading bot that bets the NO side on overpriced temperature markets using a Gaussian probability model.
This bot systematically targets Polymarket temperature markets, placing bets on the NO side when the implied probability is higher than the model estimates. It combines live METAR weather observations with forecasts from Open-Meteo to calculate a fair probability.
The strategy uses a Gaussian probability model to estimate the likelihood of a given temperature range. The difference between this calculated probability and the market's implied probability creates the edge, and the bot takes the NO side when the market overprices the YES.
The project is explicitly designed for paper trading (simulation), which means it's been tested for signal generation but not for live execution risks like slippage or latency. The bot is implemented in Python and has a substantial codebase.
Its historical performance is tracked, but since it's paper trading, there is no verified real-money profit. The main risks are model error in predicting weather, which is a meteorological challenge, and potential liquidity issues in these specific markets. The bot's edge comes from a data-driven approach to a relatively niche market.