polymarket sports bot

TypeScript/Node bot for sports prediction markets on Polymarket. Implements value betting and arbitrage strategies.

Contact me for more profitable bots

Sports types supported

The bot works with any binary sports market on Polymarket. You choose which sport (or mix) to trade by setting SPORTS_TAG_ID to a Gamma API tag, or leave it empty to scan all active events (all sports).

Examples of sports you can trade (tags are listed by Polymarket; IDs may change — check GET https://gamma-api.polymarket.com/sports):

Sport	Description
NBA	Basketball: game winners, playoffs, MVP, props
Soccer	Football / soccer: match result, league winner, cups (e.g. Champions League, World Cup)
NFL	American football: game spread, totals, Super Bowl
MLB	Baseball: game winner, World Series, awards
NHL	Hockey: game winner, Stanley Cup
UFC / MMA	Fight winner, method of victory
Tennis	Grand Slams, match winner
Other	Cricket, esports, Olympics, etc. — any category Polymarket exposes via tags

• Single sport: Set SPORTS_TAG_ID to that sport’s tag ID from /sports.
• All sports: Leave SPORTS_TAG_ID empty; the bot fetches all active events (any sport with binary Yes/No markets).

Strategies (profitable logic)

1. Value (edge betting)

• Idea: Buy when the market’s implied probability (best ask) is below your fair probability by at least a minimum edge.
• Logic: Place a limit BUY at fair - minEdge (or one tick below best ask). You only trade when ask < fair - edge, so in expectation you profit if your fair value is correct.
• Config: SPORTS_VALUE_MIN_EDGE (e.g. 0.02 = 2%), optional SPORTS_VALUE_FAIR_PROB (default 0.5). For real edge you’d plug in your own model or external odds.
• Best for: When you have a view (or data) that disagrees with the market; the bot executes when the market is “cheap” vs that view.

2. Arbitrage

• Idea: In binary markets, if best_ask_yes + best_ask_no < 1 - fee, buying both sides locks in profit.
• Logic: When askYes + askNo < 1 - SPORTS_ARB_MIN_PROFIT, place BUY orders on both outcomes with sized stakes so total cost is below 1 per share; payout is 1 per share.
• Best for: Rare mispricings; often short-lived. Set SPORTS_ARB_MIN_PROFIT (e.g. 0.01) to filter noise.

Setup

cd polymarket-sports-bot
npm install
cp .env.example .env
# Edit .env: POLYMARKET_PRIVATE_KEY, POLYMARKET_FUNDER_ADDRESS, SPORTS_STRATEGY, SPORTS_TAG_ID (optional)
npm run dev

Config

Variable	Description
SPORTS_STRATEGY	value or arbitrage
SPORTS_TAG_ID	Gamma sports tag (GET /sports). Empty = all active events
SPORTS_VALUE_MIN_EDGE	Min edge for value (e.g. 0.02)
SPORTS_VALUE_FAIR_PROB	Optional fair prob (0–1). Default 0.5
SPORTS_STAKE_USD	Stake per order (USDC)
SPORTS_MAX_ORDER_USD	Cap per order (0 = no cap)
SPORTS_ARB_MIN_PROFIT	Min arb profit (e.g. 0.01)
SPORTS_DRY_RUN	If true, no real orders

Wallet/API: same as copy-trading bot (POLYMARKET_PRIVATE_KEY, POLYMARKET_FUNDER_ADDRESS, optional API key or auto-derive).

Discovery

• Uses Gamma API GET /events?tag_id=...&active=true&closed=false (or no tag to scan all).
• List sports and tag IDs: GET https://gamma-api.polymarket.com/sports (e.g. NBA, soccer, NFL each have a tag ID).
• Bot parses binary markets (two tokens: Yes/No) and runs the chosen strategy on each. Works the same for NBA, soccer, or any other sport Polymarket lists.

Disclaimer

No strategy guarantees profit. Value betting depends on accurate fair values; arbitrage is rare and can vanish quickly. Use SPORTS_DRY_RUN=true first.