How It Works
This page shows transit patterns correlated with wins and losses from historical match data stored in the research SQLite database.
- Loss patterns (red border): transits where the player's win rate drops significantly below their baseline
- Protective patterns (green border): transits where win rate exceeds baseline
- Significance: HIGH = delta ≥15% with ≥10 appearances, MEDIUM = delta ≥10% with ≥7 appearances
- Phase 3: Deep loss analysis including angle transits (requires birth time tier A)
To add a new player, run: python research/backtest.py --player player_id
Phase 3 — Loss Signatures
Select a player to view backtesting results.