AWP | Mortis (Field-Tested)
7-day forecast validation for the Rifle category, 96 items (walk-forward, 48 folds)
| Model | Mean abs. error | vs. no change | Folds won | Range coverage (calibrated) | Interval score | Direction accuracy |
|---|---|---|---|---|---|---|
| No change (baseline) | 0.0577 | baseline | — | 80.9% | 0.3105 | — |
| Average drift | 0.0581 | -0.7% | 18/48 | 80.4% | 0.3093 | 48.8% |
| AR(1) momentum/reversion | 0.0581 | -0.8% | 21/48 | 80.4% | 0.3089 | 50.0% |
No model beats “no change” significantly and consistently (best: Average drift, t=-1.35, 18/48 folds won). Errors are on log-price changes; significance is tested on per-date averages with Newey-West standard errors. Range widths are calibrated on earlier folds only; the closer coverage is to 80%, the more trustworthy the range. The interval score rewards ranges that are both narrow and right (lower is better).
30-day forecast validation for the Rifle category, 96 items (walk-forward, 48 folds)
| Model | Mean abs. error | vs. no change | Folds won | Range coverage (calibrated) | Interval score | Direction accuracy |
|---|---|---|---|---|---|---|
| No change (baseline) | 0.1217 | baseline | — | 80.0% | 0.6752 | — |
| Average drift | 0.1246 | -2.4% | 23/48 | 80.4% | 0.6309 | 49.1% |
| AR(1) momentum/reversion | 0.1257 | -3.3% | 21/48 | 80.1% | 0.6366 | 48.1% |
No model beats “no change” significantly and consistently (best: Average drift, t=-0.85, 23/48 folds won). Errors are on log-price changes; significance is tested on per-date averages with Newey-West standard errors. Range widths are calibrated on earlier folds only; the closer coverage is to 80%, the more trustworthy the range. The interval score rewards ranges that are both narrow and right (lower is better).