WTA Model: Week 22 Analysis
Our WTA model identified 63 tradeable edges this week on Polymarket, going 30W-33L for -$6.00 net profit.
Results
- Win rate: 47.6%
- Average edge: -2.8c per trade
- Net P&L: -$6.00
Example Trades
- Barbora Krejcikova: 50c entry, Lost (-$1.06)
- Magda Linette: 32c entry, Won (+$2.23)
- Yuliia Starodubtseva: 90c entry, Won (+$0.20)
How the WTA Model Works
The model predicts win probability from live game state — score differential, time remaining, period, and Elo ratings. It never sees market prices during training, so when it disagrees with the market by 8c+, that divergence is a real signal.