docs: takeable tagging shipped, efficiency challenger blocked (board + state)

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc
This commit is contained in:
Kev
2026-07-31 00:20:18 -04:00
parent 2bfaeff572
commit 896e6e1a00
2 changed files with 75 additions and 0 deletions
+27
View File
@@ -489,3 +489,30 @@ CLV edge (which measured `p_win`, never the letter), so it is not a third shadow
**not independently bounded**: with no threshold step to multiply, efficiency scaling presupposes
probability grading. It is rule **R4 of `specs/full-output-grade-mapping.md`** and belongs to that
MLB-first challenger.
---
# TAKEABLE TAGGING BUILT · EFFICIENCY CHALLENGER BLOCKED — 2026-07-31 → `specs/takeable-tagging.md`
**Champion grade byte-identical (verified by diff).** Additive tags only.
**BLOCKED — the efficiency challenger.** Review Zero found all three inputs ABSENT: efficiency scores
(zero occurrences anywhere), base edge thresholds (`engine1.js` has zero `edge` references — the grade
is not edge-vs-threshold), and the claimed ±0.05 additive nudge (the only 0.05s on the grade path are a
teammate-absence feature, a bvp cutoff, and the shrink-toward-0.5 term). With no additive scaling to
swap, no threshold to multiply, and **zero `sport` references in `engine1.js`**, a
differs-in-exactly-one-thing challenger cannot be constructed. It needs **R1 probability grading**
first (`specs/full-output-grade-mapping.md`); shipping R1+R4 together would attribute an R1-driven
re-letter to the efficiency fix. Coverage compounds it: **9 of 11 live markets have no specced score.**
**BUILT — ledger takeable tagging (deferred C2).** `src/config/takeableStandard.js`: floor on the minus
side, **uncapped plus** — deliberately NOT `valueEngine.isTakeable` (the 160..+200 promotion band). A
+400 prop is not promotable but IS takeable; a test locks the divergence. Absent price → `null`, never
`false`. The floor is **policy, not derived**, and each row records `takeable_floor` so a re-derivation
can re-tag. Migration 034 applied + tracked. **Backfill: 1,254 rows → 1,246 tagged (781 takeable / 465
below floor), 8 NULL with `null_despite_price = 0`; settled 1,163 and graded 1,254 unchanged.**
**NOT applied — the model-version boundary tag:** no scaling change shipped, so no boundary exists;
stamping one would record a model transition that never happened.
312 suites / 3,890 tests green, web build exit 0.