diff --git a/mt5_multitf_pattern_scanner.py b/mt5_multitf_pattern_scanner.py
index da83771..179dd76 100644
--- a/mt5_multitf_pattern_scanner.py
+++ b/mt5_multitf_pattern_scanner.py
@@ -182,14 +182,16 @@ CFG = {
     # Example: Engulfing at 1.5:1, Morning Star at 2.5:1
     'rr_by_pattern': {},
     # Higher-timeframe ATR source per trading TF.
-    # M5/M15 ATR is tiny → use H1 ATR for SL/TP sizing on fast TFs.
+    # H1 ATR used for all TFs — backtest proven to give better R-relative results.
+    # Native H4/D1 ATR produces stops too wide for price to reach 1R (0.26-0.36R avg).
+    # H1 ATR gives tighter stops: H4 avg SL 38.8p (was 60p), D1 avg SL 72.3p (was 158.8p).
     # Set to None or same as trading TF to use native ATR.
     'atr_tf_by_tf': {
-        'M5':  'H1',   # Use H1 ATR for M5 signals (much wider, more realistic SL/TP)
-        'M15': 'H1',   # Use H1 ATR for M15 signals
+        'M5':  'H1',   # H1 ATR for M5 signals (avoids tiny native M5 ATR)
+        'M15': 'H1',   # H1 ATR for M15 signals
         'H1':  'H1',   # Native
-        'H4':  'H4',   # Native (H4 ATR is already meaningful)
-        'D1':  'D1',   # Native
+        'H4':  'H1',   # H1 ATR — tighter stops, +1.5% win rate, +39% avg max R vs native H4
+        'D1':  'H1',   # H1 ATR — -54% SL pips, +77% avg max R vs native D1
     },
 
     # ── Pattern Detection Thresholds ──────────────────────────────