Strategy Insights
Honest base rates from every strategy ever graded on Tessen — out-of-sample, net of fees, on 6 years of data. These are facts about the past corpus, never predictions about your strategy. Get your own number free at /grade.
Grade distribution
The win-rate trap, in our own data
Of the 31 graded strategies that win more than 60% of their trades, 9.7% still graded F. A high win rate is not an edge — expectancy after fees is. That's why the grade never uses win rate to pass a strategy.
What actually kills strategies
Share of graded strategies failing each hard gate (172 with full gate verdicts).
Pattern base rates
Grouped by timeframe + primary entry indicator. Thin samples are labeled — small n means noise, not signal.
| Timeframe | Entry indicator | Graded | Fail rate | Median expectancy |
|---|---|---|---|---|
| 1h | rsi | 57 | 47.4% | +14.41bp |
| 1h | roc | 21(thin sample) | 38.1% | +10.69bp |
| 15m | ema | 13(thin sample) | 30.8% | +3.88bp |
| 4h | rsi | 9(thin sample) | 22.2% | +10.22bp |
| 1h | bb_pband | 8(thin sample) | 50% | +3.84bp |
| 15m | rsi | 7(thin sample) | 71.4% | +5bp |
| 1h | stoch | 5(thin sample) | 80% | -7.02bp |
Aggregates over all 186 strategies ever graded on Tessen (public grades + saved agents), out-of-sample, net of fees. Base rates about the past — not predictions.
Want your strategy in this dataset — and the honest answer about whether it works?