diff --git a/research/experiments/replay.py b/research/experiments/replay.py
index 9816f11..2ce66b0 100644
--- a/research/experiments/replay.py
+++ b/research/experiments/replay.py
@@ -157,7 +157,8 @@ def reconstruct_fires(store, since: str | None = None) -> list[Fire]:
 
 
 def reconstruct_fires_from_setups(store, since: str | None = None,
-                                  limit: int | None = None) -> list[Fire]:
+                                  limit: int | None = None,
+                                  setup_version: str | None = None) -> list[Fire]:
     """Simulated fires from the labeler's setup_events — the 18k-strong backtestable
     ORB universe, vs the handful of LIVE round-trips. Each setup carries the ORB
     entry (entry_reference_price) and stop (invalidation_price); the exit
@@ -173,6 +174,12 @@ def reconstruct_fires_from_setups(store, since: str | None = None,
     if since:
         q += " AND session_date >= ?"
         args.append(since)
+    if setup_version:
+        # scope to one track — the /backtest replay judges the LIVE strategy
+        # (v1) only; shadow research tracks (liq1/vr1/run1/pm1) share this
+        # table and would dominate 8:1 with tight-stop artifacts.
+        q += " AND setup_version = ?"
+        args.append(setup_version)
     q += " ORDER BY session_date, setup_time"
     if limit:
         q += f" LIMIT {int(limit)}"
@@ -654,7 +661,7 @@ def replay_window(store, since, until, *, population_cfg=None,
     harness) — assembly only, no new simulation logic."""
     fidelity = fidelity or {"atr_frac": 0.25, "spread_mult": 3, "adverse_ticks": 1}
     until_d = _to_date(until)
-    fires = [f for f in reconstruct_fires_from_setups(store, since=str(since))
+    fires = [f for f in reconstruct_fires_from_setups(store, since=str(since), setup_version="v1")
              if f.session_date <= until_d]
 
     screened_population = 0
diff --git a/tests/unit/api/test_backtest_endpoint.py b/tests/unit/api/test_backtest_endpoint.py
index 5b4f04a..e4a23f6 100644
--- a/tests/unit/api/test_backtest_endpoint.py
+++ b/tests/unit/api/test_backtest_endpoint.py
@@ -225,3 +225,22 @@ def test_backtest_route_serves_html():
 
 if __name__ == "__main__":  # pragma: no cover
     pytest.main([__file__, "-q"])
+
+
+def test_replay_window_replays_live_track_only():
+    """/backtest must judge the strategy that actually trades: setup_version v1.
+    Shadow research tracks (liq1/vr1/run1/pm1 -- refuted, never traded live)
+    share setup_events and would otherwise dominate the replay 8:1 with
+    tight-stop artifacts (the 2026-07-17 -1.8R/trade garbage-scale bug)."""
+    store = _seed_store()
+    # a liq1 shadow fire on the same session/bars -- must NOT be replayed
+    store.con.execute(
+        "INSERT INTO setup_events (setup_id, symbol, session_date, setup_time, "
+        "setup_name, setup_version, entry_reference_price, invalidation_price) "
+        "VALUES (?,?,?,?,?,?,?,?)",
+        [f"liq-AAA-{SESS1}", "AAA", SESS1,
+         datetime(SESS1.year, SESS1.month, SESS1.day, 13, 35),
+         "liquid_breakout", "liq1", 10.0, 9.0])
+    from research.experiments.replay import replay_window
+    out = replay_window(store, str(SESS1), str(SESS2))
+    assert len(out["trades"]) == 4  # the 4 v1 fires only; liq1 excluded