Backtest
Backtest
Performance Results
Summarizes key SPD and percentile results from the latest model backtest snapshot.
The values below are rendered from the current docs snapshot.
Metric terms on this page use canonical definitions from Glossary.
As of: 2026-03-08
Source artifact: backtest-metrics.snapshot.json
Generation command: python backtest.py
SPD Statistics
| Metric | Value | Interpretation |
|---|---|---|
| SPD min | 1,084.79 sats/$ | Worst window result |
| SPD max | 22,722.45 sats/$ | Best window result |
| SPD mean | 7,085.31 sats/$ | Average across windows |
| SPD median | 3,300.38 sats/$ | Typical window result |
Percentile Statistics
| Metric | Value | Interpretation |
|---|---|---|
| Percentile min | 4.40% | Worst relative performance |
| Percentile max | 85.66% | Best relative performance |
| Percentile mean | 42.77% | Average percentile |
| Percentile median | 45.26% | Typical percentile |
| Exp-decay avg | 56.21% | Recent windows weighted more |
Strategy Validation
| Metric | Value | Requirement |
|---|---|---|
| Windows tested | 2,554 | All rolling 1-year windows |
| Losses | 908 | Windows where dynamic < uniform |
| Win rate | 64.45% | >= 50% required |
| Status | Passed | Strategy ready |