Default online backtest to production GTL

README.md

@@ -459,6 +459,11 @@ metrics. Keep the SSH tunnel open only while creating the snapshot; the backtest
 run itself is local/offline. `backtest_snapshots/` and generated backtest reports
 are git-ignored.
 
+Without a research override, both this runner and the scheduled/Admin online
+backtest use `explicit_target_ladder`, matching the live scanner's gate-target
+path. Clean Structural S/R detector arms must be selected explicitly with
+`--sr-variant`; they are not production backtest defaults.
+
 ### Archived GTL tuning decision
 
 The completed replacement, cohort-composition, and strength-sensitivity

app/services/backtest_service.py

@@ -176,8 +176,17 @@ SR_RESEARCH_VARIANTS = {
 
 
 def _sr_research_variant() -> str:
-    """S/R policy arm for local research; never read by the live scanner."""
-    value = os.getenv("BACKTEST_SR_VARIANT", "rewrite").strip().lower()
+    """Backtest target policy; research arms require an explicit override.
+
+    The unconfigured online/scheduled backtest must replay the same transient
+    Gate Target Ladder as the live scanner. Clean Structural S/R remains an
+    opt-in research arm here because it is intentionally chart/alert structure,
+    not the production gate's target source.
+    """
+    value = os.getenv(
+        "BACKTEST_SR_VARIANT",
+        EXPLICIT_TARGET_LADDER_VARIANT,
+    ).strip().lower()
     if value not in SR_RESEARCH_VARIANTS:
         allowed = ", ".join(sorted(SR_RESEARCH_VARIANTS))
         raise ValueError(f"Unknown BACKTEST_SR_VARIANT={value!r}; expected one of {allowed}")

docs/research/sr-levels-and-exits.md

@@ -912,6 +912,12 @@ model. This snapshot is now exhausted for GTL fitting; any future challenger
 must be pre-registered and evaluated on genuinely new forward data rather than
 another iteration over the same history.
 
+The scheduled/Admin online backtest and the unconfigured local snapshot
+backtest therefore default to `explicit_target_ladder`. Structural detector
+variants remain explicit research overrides. After deployment, rerun the Admin
+backtest once to replace any cached report produced by the former `rewrite`
+default.
+
 The temporary GTL matrix scripts, configurable detector branches, and
 confirmation hooks were removed after this decision. The normal snapshot
 backtester and frozen `explicit_target_ladder` parity arm remain.

tests/unit/test_backtest_service.py

@@ -793,6 +793,9 @@ def test_window_setups_too_short_returns_empty():
 
 
 def test_sr_research_variant_is_explicit_and_validated(monkeypatch):
+    monkeypatch.delenv("BACKTEST_SR_VARIANT", raising=False)
+    assert bt._sr_research_variant() == bt.EXPLICIT_TARGET_LADDER_VARIANT
+
     for variant in (
         "production_control",
         "legacy_range_grid_touch",