# H07_E_EXPLORATORY_RUN_RESULT_v0_1

**Freeze:** `H07_E_EXPLORATORY_FREEZE_v0_1` sha `175d4c1a414e3dae9e6df355e18c6f3687511c3a275ea2f1baadaa8adfc22ec5`  
**Result JSON sha256:** `16ceaac4057dd8bcc457dc7f996f1d3ed9251c7c8b5b9130f60c557c26bfd398`  
**Hub job:** `77356229-193e-4693-a90c-cae8dcb0bc4a`  
**Created:** 2026-09-13T23:25:14Z

## STATUS
EXPLORATORY · VINTAGE UNVERIFIED/INCOMPLETE · CLAIM_STRENGTH LOW · CAPITAL_USE PROHIBITED · MODEL_CHANGE NO

## Sample
| Split | N |
|-------|--:|
| Development | 59 |
| Holdout | 81 |
| Exclusions | 0 |

Development destination years: 2015–2019  
Holdout destination years: 2020–2026

## Metrics (holdout RMSE primary)
| Model | RMSE | MAE | Directional accuracy |
|-------|------|-----|----------------------|
| A baseline | 0.397669 | 0.131833 | 0.6173 |
| B baseline+WPS gas+distillate PS | 0.400919 | 0.140396 | 0.4568 |

- Absolute improvement (RMSE): **-0.003249**
- % improvement: **-0.8171%**
- Success threshold: **5.0%**
- Model B WPS coefficient: **-0.000028** (expected POSITIVE; match=NO)

## Disposition
# **KILL**

H07-E exploratory: EIA weekly gas+distillate product-supplied surprises did not clear the frozen 5% holdout RMSE bar. Holdout RMSE baseline=0.3977, +WPS=0.4009, relative improvement=-0.82%. Disposition KILL. Vintage unverified; capital prohibited.

## Mechanical risks (upfront — especially if PROMOTE)
1. STEO near-history may absorb WPS/PSM product-supplied actuals (same concept family)
2. Live hist xls revises — not first-print WPS
3. Release timing SCHEDULE_PROXY (Fri week-end + 5d)
4. COVID 2020 demand shock in holdout

## Vintage limitations
- EIA WPS from live hist xls — not first-print
- STEO forecast dates mix PARSED + SCHEDULE_PROXY
- No capital / no validated-edge claim