cristal@0xCristal
"My Son is a Polymarket Genius"
That's exactly what my dad said
Why? Answer👇
I told him: my bot scored these addresses with a custom algorithm and turned $500 → $13,782 in 21 days, without me manually placing a single trade
Most people blindly copy the leaderboard's 90%+ win-rate accounts. Win rate is a total trap
Buying a YES share at 82¢ means you are risking 82¢ just to win 18¢. One single loss wipes out nearly five wins. In that bracket, an 80% win rate is essentially a break-even point
So, the bot ignores the win-rate vanity metric and scores wallets based on:
S(w) = α·PnL_realized + β·Consistency(t) + γ·Specialization(c) − δ·MaxDD
The Strict Filtering Engine:
- 80+ resolved trades
- No single trade accounts for more than 30% of total PnL (to filter out one-hit wonders)
- Entries strictly between 25¢ and 65¢
- Standardized position sizing (no erratic revenge bets)
- Active within the last 14 days
- 5+ months of a consistent upward equity curve
Only 16.8% of wallets are net profitable. After filtering: ~40 worth copying out of 1.73M
Only 11.2% of all wallets are net profitable over the long term. After running my filters, only ~62 wallets out of 2.4M were actually worth following
Performance Comparison:
0x4f12...a8b3 - 62% win rate, $942,500 PnL
0x9c33...e2f9 - 74% win rate, $185,000 PnL
The second wallet wins more often, but the first wallet makes 5x the money because they understand value. The bot picks the first one every time.
It also uses Category-Locking: if a wallet consistently prints money in Global Politics, the bot only copies its politics trades and ignores its attempts at betting on Pop Culture or Sports
Sizing Logic:
To protect the bankroll, the bot uses a conservative Fractional Kelly Criterion:
f = min(bankroll × kelly_quarter, max_position)
kelly_quarter = ((p·b − q)/b)/4
Don't forget to drop a like and follow me
The Result: 21 days. 18 wallets tracked. 485 trades. $500 → $13,782
My dad still doesn't get the math, but the PnL doesn't need him to
