Prediction Market Alpha Playbook

Prediction Market Alpha Playbook

A comprehensive reference for building trading bots on prediction markets — **Polymarket**, **Kalshi**, and cross-venue. This repository distills hard-won lessons from production bot operation: what edges work, what APIs are useful, what architectural patterns scale, and what antipatterns burn real capital.

Who this is for

• AI agents and engineers bootstrapping prediction-market bots from scratch
• Quantitative researchers looking for validated edge ideas with honest evidence levels
• Anyone operating prediction-market systems who wants to avoid well-documented pitfalls

Everything here is written as **generic instructional content**. Treat it as a starting map, not a recipe — markets evolve, rules change, and what was alpha yesterday may be arbed out tomorrow.

Contents

File	Purpose
`PLAYBOOK.md`	Step-by-step quickstart: from zero to first paper fire in a day
`EDGES.md`	~20 categorised edge types with mechanisms, evidence levels, and implementation hints
`APIS.md`	~25 relevant APIs with auth, endpoints, rate limits, gotchas, and minimal working examples
`ARCHITECTURE.md`	Battle-tested system design patterns — event loop, graduation pipeline, safety guards, observability
`ANTIPATTERNS.md`	~20 failure modes that will burn you if you don't know to look for them
`METHODOLOGY.md`	How to systematically find, validate, and graduate strategies to live trading
`NEG_RISK_NO_CARRY.md`	Deep-dive on one specific edge from EDGES.md §6b: structural NEG_RISK long-tail NO carry, hold-to-resolution

Quick orientation for AI agents

If you're an AI agent being asked to build or improve a prediction-market bot, ingest the files in this order:

1. ARCHITECTURE.md — understand the target shape of the system (event loop, pap