polymarket-copy-tracker

name	polymarket-copy-tracker
description	Analyze top Polymarket traders for copy trading. Use when asked to find best traders to copy, analyze trader performance/history, evaluate wallet PnL, discover high-ROI wallets, or research trader win rates on Polymarket. Supports Dune analytics, Polymarket Data API, and automated scoring.

Polymarket Copy Tracker

Analyze top Polymarket traders for copy trading signals.

Data Sources

1. Data API (data-api.polymarket.com) - Wallet history (public, no auth)
2. Polymarket.com/leaderboard - Top traders page (extract wallet addresses)

Quick Start

Step 1: Get wallet addresses

python3 scripts/find_top_traders.py --limit 20

Step 2: Analyze a wallet

python3 scripts/analyze_wallet.py 0xWALLET_ADDRESS
python3 scripts/analyze_wallet.py 0xWALLET_ADDRESS --json  # raw JSON output

Step 3: Score multiple wallets

python3 scripts/copy_score.py wallet1 wallet2 wallet3 ...

API Endpoints

Data API (Public)

# Open positions
GET https://data-api.polymarket.com/positions?user=WALLET&limit=50

# Closed/resolved positions
GET https://data-api.polymarket.com/closed-positions?user=WALLET&limit=50

# Trade history
GET https://data-api.polymarket.com/trades?user=WALLET&limit=100

Key fields in positions:

• title - Market name
• size - Number of shares
• avgPrice - Average entry price
• currentValue - Current market value
• cashPnl - Unrealized P&L (open positions)
• realizedPnl - Realized P&L (resolved positions)
• outcome - Selected outcome (Yes/No/Team name)
• redeemable - True if resolved and can be redeemed

Scoring Criteria

Metric	Weight	Good	Bad
Win Rate	40%	≥65%	<45%
Total PnL	30%	≥$1000	<$0
Risk Score	20%	<0.8	≥1.0
Sample Size	10%	≥20 trades	<5 trades

Risk Score = Avg Loss / Avg Win (lower = better, <1.0 means wins > losses)

Final Score: 0-100 → COPYABLE (≥70), CAUTION (45-69), NOT RECOMMENDED (<45)

Example Output

=== Analyzing Wallet: 0xf19572...8057 ===

CLOSED POSITIONS (Win Rate & PnL)
==================================================
Total Closed:    49
Win Rate:        67.3%  (33W / 16L)
Total PnL:       +$2,223,761
Avg Win:         $110,028
Avg Loss:        $87,948
Risk Score:      0.80  ✅ GOOD

OPEN POSITIONS: 8 | Value: $17,739 | Unrealized: -$445,325

COPY ASSESSMENT
==================================================
Score: 92/100  ✅ COPYABLE
  ✅ High win rate (67%)
  ✅ Excellent PnL ($2,223,761)
  ✅ Low risk (score 0.80)
  ✅ Good sample (49 trades)

Scripts

Script	Purpose
find_top_traders.py	Extract wallet addresses from polymarket.com/leaderboard
analyze_wallet.py	Deep-dive a single wallet (positions, trades, score)
copy_score.py	Score and rank multiple wallets in parallel

Curated Wallet Database

See references/top-wallets.md for a curated list of high-quality wallets organized by tier:

• Tier 2 wallets are best for copy trading (high ROI + human-scale activity)
• Includes filtering notes (skip protocol addresses, bots, etc.)

Discovery Methods

See references/discovery-methods.md for 5 methods to find wallets:

1. Dune Profitability Dashboard — PnL/ROI rankings (free, no login)
2. Dune Wallet Analyzooooor — Per-wallet win rate (free, needs login)
3. Polymarket Data API — Real-time positions (public API)
4. Polymarket Leaderboard — Quick browse top traders
5. Polymarket Profile Page — Visual current positions

Quick Workflow

# 1. Score a known wallet
python3 scripts/analyze_wallet.py 0x2785e7022dc20757108204b13c08cea8613b70ae

# 2. Compare multiple candidates
python3 scripts/copy_score.py 0x2785e702... 0xed107a85... 0x6a72f618...

# 3. Find new top traders
python3 scripts/find_top_traders.py --limit 20

Notes

• Data API is public (read-only, no auth required)
• Rate limit: ~100 req/min
• Not all wallets have public history (new or privacy)
• closed-positions has resolved trades, positions has open trades
• PnL is calculated from face value (shares × $1) minus cost (shares × avgPrice)
• Dune data is all-time historical; API shows recent activity
• Wallets with 0 trades but huge PnL are protocol addresses — skip them
• Wallets with >500K trades are likely market-making bots — not suitable for copy trading