MVRVPlus Strategy¶
Summary¶
strategy_id:mvrv-plus- intent mode:
profile - support tier:
experimental - promotion stage:
candidate - owner:
StackSats Maintainers
Why this model exists¶
mvrv-plus is an experimental candidate that layers BRK-aware regime logic on top of the stable MVRV baseline. It exists to test whether overlays can improve comparison results without breaking the maintainability of the core model family.
Feature dependencies¶
- required feature sets:
core_model_features_v1 - required transformed columns: overlay-specific transformed columns from the candidate implementation
Benchmarks¶
- benchmark strategy IDs:
uniform,mvrv
Expected comparison behavior¶
Candidate-stage runs should show distinct, interpretable behavior against mvrv and uniform on the same window. Comparisons should be repeated, not judged from a single favorable run.
Known failure modes and caveats¶
This model is not yet stable. If it only improves on narrow date ranges or becomes too sensitive to overlay availability, it should remain experimental.