fix: purge ML validation label leakage

config/llm_settings.json

@@ -1,9 +1,9 @@
 {
-  "provider": "openrouter",
+  "provider": "ollama",
-  "model": "minimax/minimax-m2.5",
+  "model": "gemma4:12b-mlx",
   "providers": {
     "ollama": {
-      "base_url": "http://100.100.242.21:11434"
+      "base_url": "http://100.79.255.5:11434"
     },
     "lmstudio": {
       "base_url": "http://100.100.242.21:1234"

data/score_history.jsonl

@@ -110,3 +110,40 @@
 {"timestamp": "2026-03-21T22:54:22.144542+00:00", "composite_score": 70.0, "scored_count": 9, "metrics": {"fear_greed": {"score": 10, "value": 12}, "puell_multiple": {"score": 8, "value": 0.6602699608966011}, "mvrv_zscore": {"score": 8, "value": 0.5211180167687892}, "drawdown": {"score": null, "value": null}, "price_vs_200w_sma": {"score": 7, "value": 58895.78086828114}, "reserve_risk": {"score": 10, "value": 0.0012985709697654493}, "rhodl_ratio": {"score": 4, "value": 1230.6243545314708}, "nupl": {"score": 8, "value": 0.22243290955405431}, "lth_realized_price": {"score": 5, "value": 43346.58756410873}, "hash_ribbons": {"score": 3, "value": null}}}
 {"timestamp": "2026-03-21T22:55:08.385540+00:00", "composite_score": 71.0, "scored_count": 10, "metrics": {"fear_greed": {"score": 10, "value": 12}, "puell_multiple": {"score": 8, "value": 0.6602699608966011}, "mvrv_zscore": {"score": 8, "value": 0.5211180167687892}, "drawdown": {"score": 8, "value": 44.26554568527919}, "price_vs_200w_sma": {"score": 7, "value": 58895.78086828114}, "reserve_risk": {"score": 10, "value": 0.0012985709697654493}, "rhodl_ratio": {"score": 4, "value": 1230.6243545314708}, "nupl": {"score": 8, "value": 0.22243290955405431}, "lth_realized_price": {"score": 5, "value": 43346.58756410873}, "hash_ribbons": {"score": 3, "value": null}}}
 {"timestamp": "2026-03-21T22:55:33.933753+00:00", "composite_score": 71.0, "scored_count": 10, "metrics": {"fear_greed": {"score": 10, "value": 12}, "puell_multiple": {"score": 8, "value": 0.6602699608966011}, "mvrv_zscore": {"score": 8, "value": 0.5211180167687892}, "drawdown": {"score": 8, "value": 44.26316624365482}, "price_vs_200w_sma": {"score": 7, "value": 58895.78086828114}, "reserve_risk": {"score": 10, "value": 0.0012985709697654493}, "rhodl_ratio": {"score": 4, "value": 1230.6243545314708}, "nupl": {"score": 8, "value": 0.22243290955405431}, "lth_realized_price": {"score": 5, "value": 43346.58756410873}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-27T18:10:22.517545+00:00", "composite_score": 63.3, "scored_count": 3, "metrics": {"fear_greed": {"score": 8, "value": 15}, "puell_multiple": {"score": null, "value": null}, "mvrv_zscore": {"score": null, "value": null}, "drawdown": {"score": 8, "value": 52.07407994923858}, "price_vs_200w_sma": {"score": null, "value": null}, "reserve_risk": {"score": null, "value": null}, "rhodl_ratio": {"score": null, "value": null}, "nupl": {"score": null, "value": null}, "lth_realized_price": {"score": null, "value": null}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-27T18:25:30.189079+00:00", "composite_score": 63.3, "scored_count": 3, "metrics": {"fear_greed": {"score": 8, "value": 15}, "puell_multiple": {"score": null, "value": null}, "mvrv_zscore": {"score": null, "value": null}, "drawdown": {"score": 8, "value": 52.098667512690355}, "price_vs_200w_sma": {"score": null, "value": null}, "reserve_risk": {"score": null, "value": null}, "rhodl_ratio": {"score": null, "value": null}, "nupl": {"score": null, "value": null}, "lth_realized_price": {"score": null, "value": null}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-27T18:26:00.556392+00:00", "composite_score": 63.3, "scored_count": 3, "metrics": {"fear_greed": {"score": 8, "value": 15}, "puell_multiple": {"score": null, "value": null}, "mvrv_zscore": {"score": null, "value": null}, "drawdown": {"score": 8, "value": 52.098667512690355}, "price_vs_200w_sma": {"score": null, "value": null}, "reserve_risk": {"score": null, "value": null}, "rhodl_ratio": {"score": null, "value": null}, "nupl": {"score": null, "value": null}, "lth_realized_price": {"score": null, "value": null}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-27T18:40:30.805472+00:00", "composite_score": 63.3, "scored_count": 3, "metrics": {"fear_greed": {"score": 8, "value": 15}, "puell_multiple": {"score": null, "value": null}, "mvrv_zscore": {"score": null, "value": null}, "drawdown": {"score": 8, "value": 52.04473350253808}, "price_vs_200w_sma": {"score": null, "value": null}, "reserve_risk": {"score": null, "value": null}, "rhodl_ratio": {"score": null, "value": null}, "nupl": {"score": null, "value": null}, "lth_realized_price": {"score": null, "value": null}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-27T18:55:31.544772+00:00", "composite_score": 63.3, "scored_count": 3, "metrics": {"fear_greed": {"score": 8, "value": 15}, "puell_multiple": {"score": null, "value": null}, "mvrv_zscore": {"score": null, "value": null}, "drawdown": {"score": 8, "value": 52.066148477157356}, "price_vs_200w_sma": {"score": null, "value": null}, "reserve_risk": {"score": null, "value": null}, "rhodl_ratio": {"score": null, "value": null}, "nupl": {"score": null, "value": null}, "lth_realized_price": {"score": null, "value": null}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-27T18:57:42.639217+00:00", "composite_score": 74.0, "scored_count": 10, "metrics": {"fear_greed": {"score": 8, "value": 15}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.04790609137056}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-27T19:12:44.307793+00:00", "composite_score": 74.0, "scored_count": 10, "metrics": {"fear_greed": {"score": 8, "value": 15}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.02014593908629}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-27T19:27:45.019177+00:00", "composite_score": 74.0, "scored_count": 10, "metrics": {"fear_greed": {"score": 8, "value": 15}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.22874365482234}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-27T19:42:45.679921+00:00", "composite_score": 74.0, "scored_count": 10, "metrics": {"fear_greed": {"score": 8, "value": 15}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.35009517766498}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-27T19:57:46.474768+00:00", "composite_score": 74.0, "scored_count": 10, "metrics": {"fear_greed": {"score": 8, "value": 15}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.34850888324873}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-27T20:12:47.228456+00:00", "composite_score": 74.0, "scored_count": 10, "metrics": {"fear_greed": {"score": 8, "value": 15}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.2604695431472}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-27T20:27:47.904578+00:00", "composite_score": 74.0, "scored_count": 10, "metrics": {"fear_greed": {"score": 8, "value": 15}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.27633248730964}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-27T20:42:48.631663+00:00", "composite_score": 74.0, "scored_count": 10, "metrics": {"fear_greed": {"score": 8, "value": 15}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.26919416243655}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-27T20:57:49.365132+00:00", "composite_score": 74.0, "scored_count": 10, "metrics": {"fear_greed": {"score": 8, "value": 15}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.264435279187815}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-27T21:12:50.093194+00:00", "composite_score": 74.0, "scored_count": 10, "metrics": {"fear_greed": {"score": 8, "value": 15}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.202569796954315}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-27T21:27:50.802146+00:00", "composite_score": 74.0, "scored_count": 10, "metrics": {"fear_greed": {"score": 8, "value": 15}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.16132614213198}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-27T21:42:51.460898+00:00", "composite_score": 74.0, "scored_count": 10, "metrics": {"fear_greed": {"score": 8, "value": 15}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.24857233502538}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-27T21:57:52.223283+00:00", "composite_score": 74.0, "scored_count": 10, "metrics": {"fear_greed": {"score": 8, "value": 15}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.318369289340104}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-27T22:12:52.898551+00:00", "composite_score": 74.0, "scored_count": 10, "metrics": {"fear_greed": {"score": 8, "value": 15}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.37071700507614}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-27T22:27:54.711182+00:00", "composite_score": 74.0, "scored_count": 10, "metrics": {"fear_greed": {"score": 8, "value": 15}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.429409898477154}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-27T22:42:55.510257+00:00", "composite_score": 74.0, "scored_count": 10, "metrics": {"fear_greed": {"score": 8, "value": 15}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.31043781725888}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-28T20:31:14.232026+00:00", "composite_score": 74.0, "scored_count": 10, "metrics": {"fear_greed": {"score": 8, "value": 18}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.70066624365482}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-28T20:35:38.591732+00:00", "composite_score": 74.0, "scored_count": 10, "metrics": {"fear_greed": {"score": 8, "value": 18}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.73080583756345}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-28T20:42:21.081121+00:00", "composite_score": 74.0, "scored_count": 10, "metrics": {"fear_greed": {"score": 8, "value": 18}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.721288071065985}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-28T20:57:21.749415+00:00", "composite_score": 74.0, "scored_count": 10, "metrics": {"fear_greed": {"score": 8, "value": 18}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.75301395939086}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}}}
+{"timestamp": "2026-06-28T21:09:38.418891+00:00", "composite_score": 68.1, "scored_count": 16, "metrics": {"fear_greed": {"score": 8, "value": 18}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.73159898477158}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}, "sopr": {"score": 8, "value": 0.990007085019496}, "sellside_risk": {"score": 10, "value": 0.000734882488382521}, "active_address_momentum": {"score": 4, "value": -0.09386733861382784}, "txcount_momentum": {"score": 4, "value": 0.04454611494295241}, "nvt_price": {"score": 5, "value": 54673.815447216126}, "vdd_multiple": {"score": 4, "value": -0.030495759573562122}}}
+{"timestamp": "2026-06-28T21:11:19.117377+00:00", "composite_score": 69.4, "scored_count": 16, "metrics": {"fear_greed": {"score": 8, "value": 18}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.723667512690355}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}, "sopr": {"score": 8, "value": 0.990007085019496}, "sellside_risk": {"score": 10, "value": 0.000734882488382521}, "active_address_momentum": {"score": 4, "value": -0.09386733861382784}, "txcount_momentum": {"score": 6, "value": 0.04454611494295241}, "nvt_price": {"score": 5, "value": 54673.815447216126}, "vdd_multiple": {"score": 4, "value": -0.030495759573562122}}}
+{"timestamp": "2026-06-28T21:26:19.843356+00:00", "composite_score": 69.4, "scored_count": 16, "metrics": {"fear_greed": {"score": 8, "value": 18}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.723667512690355}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}, "sopr": {"score": 8, "value": 0.990007085019496}, "sellside_risk": {"score": 10, "value": 0.000734882488382521}, "active_address_momentum": {"score": 4, "value": -0.09386733861382784}, "txcount_momentum": {"score": 6, "value": 0.04454611494295241}, "nvt_price": {"score": 5, "value": 54673.815447216126}, "vdd_multiple": {"score": 4, "value": -0.030495759573562122}}}
+{"timestamp": "2026-06-28T21:39:03.697500+00:00", "composite_score": 69.4, "scored_count": 16, "metrics": {"fear_greed": {"score": 8, "value": 18}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.68797588832488}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}, "sopr": {"score": 8, "value": 0.990007085019496}, "sellside_risk": {"score": 10, "value": 0.000734882488382521}, "active_address_momentum": {"score": 4, "value": -0.09386733861382784}, "txcount_momentum": {"score": 6, "value": 0.04454611494295241}, "nvt_price": {"score": 5, "value": 54673.815447216126}, "vdd_multiple": {"score": 4, "value": -0.030495759573562122}}}
+{"timestamp": "2026-06-28T21:42:37.143098+00:00", "composite_score": 69.4, "scored_count": 16, "metrics": {"fear_greed": {"score": 8, "value": 18}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.71890862944163}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}, "sopr": {"score": 8, "value": 0.990007085019496}, "sellside_risk": {"score": 10, "value": 0.000734882488382521}, "active_address_momentum": {"score": 4, "value": -0.09386733861382784}, "txcount_momentum": {"score": 6, "value": 0.04454611494295241}, "nvt_price": {"score": 5, "value": 54673.815447216126}, "vdd_multiple": {"score": 4, "value": -0.030495759573562122}}}
+{"timestamp": "2026-06-28T21:57:37.803751+00:00", "composite_score": 69.4, "scored_count": 16, "metrics": {"fear_greed": {"score": 8, "value": 18}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.96002538071066}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}, "sopr": {"score": 8, "value": 0.990007085019496}, "sellside_risk": {"score": 10, "value": 0.000734882488382521}, "active_address_momentum": {"score": 4, "value": -0.09386733861382784}, "txcount_momentum": {"score": 6, "value": 0.04454611494295241}, "nvt_price": {"score": 5, "value": 54673.815447216126}, "vdd_multiple": {"score": 4, "value": -0.030495759573562122}}}
+{"timestamp": "2026-06-28T22:12:38.557670+00:00", "composite_score": 69.4, "scored_count": 16, "metrics": {"fear_greed": {"score": 8, "value": 18}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.72049492385786}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}, "sopr": {"score": 8, "value": 0.990007085019496}, "sellside_risk": {"score": 10, "value": 0.000734882488382521}, "active_address_momentum": {"score": 4, "value": -0.09386733861382784}, "txcount_momentum": {"score": 6, "value": 0.04454611494295241}, "nvt_price": {"score": 5, "value": 54673.815447216126}, "vdd_multiple": {"score": 4, "value": -0.030495759573562122}}}
+{"timestamp": "2026-06-28T22:27:39.293792+00:00", "composite_score": 69.4, "scored_count": 16, "metrics": {"fear_greed": {"score": 8, "value": 18}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.94733502538072}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}, "sopr": {"score": 8, "value": 0.990007085019496}, "sellside_risk": {"score": 10, "value": 0.000734882488382521}, "active_address_momentum": {"score": 4, "value": -0.09386733861382784}, "txcount_momentum": {"score": 6, "value": 0.04454611494295241}, "nvt_price": {"score": 5, "value": 54673.815447216126}, "vdd_multiple": {"score": 4, "value": -0.030495759573562122}}}
+{"timestamp": "2026-06-28T22:42:40.175733+00:00", "composite_score": 69.4, "scored_count": 16, "metrics": {"fear_greed": {"score": 8, "value": 18}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 53.05678934010152}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}, "sopr": {"score": 8, "value": 0.990007085019496}, "sellside_risk": {"score": 10, "value": 0.000734882488382521}, "active_address_momentum": {"score": 4, "value": -0.09386733861382784}, "txcount_momentum": {"score": 6, "value": 0.04454611494295241}, "nvt_price": {"score": 5, "value": 54673.815447216126}, "vdd_multiple": {"score": 4, "value": -0.030495759573562122}}}
+{"timestamp": "2026-06-28T22:57:40.874889+00:00", "composite_score": 69.4, "scored_count": 16, "metrics": {"fear_greed": {"score": 8, "value": 18}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 53.22255710659899}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}, "sopr": {"score": 8, "value": 0.990007085019496}, "sellside_risk": {"score": 10, "value": 0.000734882488382521}, "active_address_momentum": {"score": 4, "value": -0.09386733861382784}, "txcount_momentum": {"score": 6, "value": 0.04454611494295241}, "nvt_price": {"score": 5, "value": 54673.815447216126}, "vdd_multiple": {"score": 4, "value": -0.030495759573562122}}}
+{"timestamp": "2026-06-28T23:12:41.665346+00:00", "composite_score": 69.4, "scored_count": 16, "metrics": {"fear_greed": {"score": 8, "value": 18}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 53.02744289340101}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}, "sopr": {"score": 8, "value": 0.990007085019496}, "sellside_risk": {"score": 10, "value": 0.000734882488382521}, "active_address_momentum": {"score": 4, "value": -0.09386733861382784}, "txcount_momentum": {"score": 6, "value": 0.04454611494295241}, "nvt_price": {"score": 5, "value": 54673.815447216126}, "vdd_multiple": {"score": 4, "value": -0.030495759573562122}}}
+{"timestamp": "2026-06-28T23:27:42.349311+00:00", "composite_score": 69.4, "scored_count": 16, "metrics": {"fear_greed": {"score": 8, "value": 18}, "puell_multiple": {"score": 5, "value": 0.7044739567707577}, "mvrv_zscore": {"score": 8, "value": 0.22409759021503936}, "drawdown": {"score": 8, "value": 52.95923223350254}, "price_vs_200w_sma": {"score": 10, "value": 62284.65298428873}, "reserve_risk": {"score": 10, "value": 0.0010258831016337609}, "rhodl_ratio": {"score": 7, "value": 882.5868942025234}, "nupl": {"score": 8, "value": 0.11309045542914359}, "lth_realized_price": {"score": 7, "value": 49767.33015910989}, "hash_ribbons": {"score": 3, "value": null}, "sopr": {"score": 8, "value": 0.990007085019496}, "sellside_risk": {"score": 10, "value": 0.000734882488382521}, "active_address_momentum": {"score": 4, "value": -0.09386733861382784}, "txcount_momentum": {"score": 6, "value": 0.04454611494295241}, "nvt_price": {"score": 5, "value": 54673.815447216126}, "vdd_multiple": {"score": 4, "value": -0.030495759573562122}}}

ml/optimizer.py

@@ -43,6 +43,9 @@ START_DATE = "2018-02-01"
 TRAIN_CUTOFF_DAYS = 365
 # Target: forward 365d return > 30% = "good time to buy"
 GOOD_BUY_THRESHOLD = 30.0
+# Validation embargo/purge horizon: labels use 365-day forward returns.
+LABEL_HORIZON_DAYS = 365
+VALIDATION_SPLITS = 5
 
 # The 8 core metrics we score
 METRIC_KEYS = [
@@ -95,6 +98,112 @@ def score_range(value, ranges):
     return 0
 
 
+SCORE_KEYS = [
+    "puell_multiple", "mvrv_zscore", "reserve_risk", "rhodl_ratio",
+    "nupl", "fear_greed", "drawdown", "pct_above_200w_sma", "pct_above_lth_rp",
+]
+
+SCORE_FEATURES = [f"score_{k}" for k in SCORE_KEYS]
+RAW_FEATURES = [
+    "raw_puell_multiple", "raw_mvrv_zscore", "raw_reserve_risk",
+    "raw_rhodl_ratio", "raw_nupl", "raw_fear_greed",
+    "raw_pct_above_200w_sma", "raw_pct_above_lth_rp", "raw_drawdown",
+]
+DELTA_FEATURES = [
+    "delta_30d_mvrv_zscore", "delta_30d_nupl",
+    "delta_30d_puell_multiple", "delta_30d_reserve_risk",
+]
+INTERACTION_FEATURES = ["mvrv_x_nupl", "puell_x_reserve"]
+CYCLE_FEATURES = ["days_since_ath"]
+FEATURE_COLS = SCORE_FEATURES + RAW_FEATURES + DELTA_FEATURES + INTERACTION_FEATURES + CYCLE_FEATURES
+
+BRACKETS = [
+    (0, 20, "Extreme Caution"),
+    (21, 40, "Caution"),
+    (41, 55, "Neutral"),
+    (56, 70, "Moderate Opportunity"),
+    (71, 85, "Strong Accumulation"),
+    (86, 100, "Extreme Accumulation"),
+]
+
+
+def _row_date(row):
+    return datetime.strptime(row["date"], "%Y-%m-%d")
+
+
+def purged_time_series_splits(rows, n_splits=VALIDATION_SPLITS,
+                              label_horizon_days=LABEL_HORIZON_DAYS,
+                              embargo_days=0):
+    """Yield expanding-window splits with overlapping forward-label windows removed.
+
+    A row dated T with a 365-day forward-return label consumes information up to
+    T+365. For validation beginning at V, any training row whose label window
+    reaches V is removed. This keeps validation metrics out-of-sample for the
+    forward-return label, not just for features.
+    """
+    base_splitter = TimeSeriesSplit(n_splits=n_splits)
+    row_dates = [_row_date(r) for r in rows]
+    horizon = timedelta(days=label_horizon_days)
+    embargo = timedelta(days=embargo_days)
+
+    for train_idx, val_idx in base_splitter.split(np.arange(len(rows))):
+        val_start = row_dates[val_idx[0]]
+        val_end = row_dates[val_idx[-1]]
+        purged_train = []
+        for idx in train_idx:
+            label_end = row_dates[idx] + horizon
+            before_validation_label_window = label_end <= val_start - embargo
+            after_validation_embargo = row_dates[idx] > val_end + embargo
+            if before_validation_label_window or after_validation_embargo:
+                purged_train.append(idx)
+        if purged_train:
+            yield np.array(purged_train, dtype=int), np.array(val_idx, dtype=int)
+
+
+def _build_model():
+    return GradientBoostingClassifier(
+        n_estimators=300,
+        learning_rate=0.05,
+        max_depth=4,
+        subsample=0.8,
+        min_samples_leaf=20,
+        random_state=42,
+    )
+
+
+def derive_metric_weights(feature_cols, importances):
+    """Aggregate feature importances back to transparent score metric weights."""
+    metric_names = list(SCORE_KEYS)
+    feature_to_metric = {}
+    for m in metric_names:
+        feature_to_metric[f"score_{m}"] = m
+        feature_to_metric[f"raw_{m}"] = m
+    feature_to_metric["delta_30d_mvrv_zscore"] = "mvrv_zscore"
+    feature_to_metric["delta_30d_nupl"] = "nupl"
+    feature_to_metric["delta_30d_puell_multiple"] = "puell_multiple"
+    feature_to_metric["delta_30d_reserve_risk"] = "reserve_risk"
+
+    metric_importances = {m: 0.0 for m in metric_names}
+    for name, imp in zip(feature_cols, importances):
+        if name in feature_to_metric:
+            metric_importances[feature_to_metric[name]] += float(imp)
+        elif name == "mvrv_x_nupl":
+            metric_importances["mvrv_zscore"] += float(imp) / 2
+            metric_importances["nupl"] += float(imp) / 2
+        elif name == "puell_x_reserve":
+            metric_importances["puell_multiple"] += float(imp) / 2
+            metric_importances["reserve_risk"] += float(imp) / 2
+        elif name == "days_since_ath":
+            metric_importances["drawdown"] += float(imp)
+
+    total_imp = sum(metric_importances.values())
+    if total_imp > 0:
+        weights = {k: round(v / total_imp, 4) for k, v in metric_importances.items()}
+    else:
+        weights = {k: round(1 / len(metric_importances), 4) for k in metric_importances}
+    return dict(sorted(weights.items(), key=lambda x: x[1], reverse=True))
+
+
 def build_dataset(index, thresholds):
     """Build aligned training dataset: metric scores + forward returns."""
     # Get all dates from 2018-02-01 onward
@@ -257,39 +366,33 @@ def train_model(rows):
     log.info("Target distribution: %d positive (%.1f%%), %d negative",
              positive, positive / len(labeled) * 100, len(labeled) - positive)
 
-    # Feature columns: scores + raw values + deltas + interactions + cycle position
+    feature_cols = FEATURE_COLS
-    score_features = [
-        "score_puell_multiple", "score_mvrv_zscore", "score_reserve_risk",
-        "score_rhodl_ratio", "score_nupl", "score_fear_greed",
-        "score_drawdown", "score_pct_above_200w_sma", "score_pct_above_lth_rp",
-    ]
-    raw_features = [
-        "raw_puell_multiple", "raw_mvrv_zscore", "raw_reserve_risk",
-        "raw_rhodl_ratio", "raw_nupl", "raw_fear_greed",
-        "raw_pct_above_200w_sma", "raw_pct_above_lth_rp", "raw_drawdown",
-    ]
-    delta_features = [
-        "delta_30d_mvrv_zscore", "delta_30d_nupl",
-        "delta_30d_puell_multiple", "delta_30d_reserve_risk",
-    ]
-    interaction_features = ["mvrv_x_nupl", "puell_x_reserve"]
-    cycle_features = ["days_since_ath"]
-
-    feature_cols = score_features + raw_features + delta_features + interaction_features + cycle_features
 
     X = np.array([[r[f] for f in feature_cols] for r in labeled])
     y = np.array([r["target"] for r in labeled])
 
     log.info("Feature matrix: %d samples x %d features", X.shape[0], X.shape[1])
 
-    # Time-series cross-validation (expanding window, 5 splits)
+    # Purged time-series cross-validation. Standard TimeSeriesSplit is not
-    tscv = TimeSeriesSplit(n_splits=5)
+    # enough here because each label consumes the next 365 days of returns.
     cv_scores = []
     cv_f1 = []
     cv_precision = []
     cv_recall = []
+    fold_results = []
 
-    for fold, (train_idx, val_idx) in enumerate(tscv.split(X)):
+    splits = list(purged_time_series_splits(
+        labeled,
+        n_splits=VALIDATION_SPLITS,
+        label_horizon_days=LABEL_HORIZON_DAYS,
+        embargo_days=0,
+    ))
+    if not splits:
+        log.error("No viable purged validation splits. Need more history for %dd label horizon.",
+                  LABEL_HORIZON_DAYS)
+        return None
+
+    for fold, (train_idx, val_idx) in enumerate(splits):
         X_train, X_val = X[train_idx], X[val_idx]
         y_train, y_val = y[train_idx], y[val_idx]
 
@@ -297,14 +400,7 @@ def train_model(rows):
         X_train_s = scaler.fit_transform(X_train)
         X_val_s = scaler.transform(X_val)
 
-        model = GradientBoostingClassifier(
+        model = _build_model()
-            n_estimators=300,
-            learning_rate=0.05,
-            max_depth=4,
-            subsample=0.8,
-            min_samples_leaf=20,
-            random_state=42,
-        )
         model.fit(X_train_s, y_train)
 
         y_pred = model.predict(X_val_s)
@@ -320,27 +416,40 @@ def train_model(rows):
         cv_precision.append(prec)
         cv_recall.append(rec)
 
-        train_dates = f"{labeled[train_idx[0]]['date']} to {labeled[train_idx[-1]]['date']}"
+        fold_weights = derive_metric_weights(feature_cols, model.feature_importances_)
-        val_dates = f"{labeled[val_idx[0]]['date']} to {labeled[val_idx[-1]]['date']}"
+        fold_results.append({
+            "fold": fold + 1,
+            "train_idx": train_idx.tolist(),
+            "val_idx": val_idx.tolist(),
+            "weights": fold_weights,
+            "metrics": {
+                "auc": round(float(auc), 4),
+                "f1": round(float(f1), 4),
+                "precision": round(float(prec), 4),
+                "recall": round(float(rec), 4),
+            },
+            "date_ranges": {
+                "train": f"{labeled[train_idx[0]]['date']} to {labeled[train_idx[-1]]['date']}",
+                "validation": f"{labeled[val_idx[0]]['date']} to {labeled[val_idx[-1]]['date']}",
+            },
+            "n_train": len(train_idx),
+            "n_validation": len(val_idx),
+        })
+
+        train_dates = fold_results[-1]["date_ranges"]["train"]
+        val_dates = fold_results[-1]["date_ranges"]["validation"]
         log.info("Fold %d: Train %s | Val %s | AUC=%.3f F1=%.3f P=%.3f R=%.3f",
                  fold + 1, train_dates, val_dates, auc, f1, prec, rec)
 
-    log.info("CV Mean AUC: %.3f (+/- %.3f)", np.mean(cv_scores), np.std(cv_scores))
+    log.info("Purged CV Mean AUC: %.3f (+/- %.3f)", np.mean(cv_scores), np.std(cv_scores))
-    log.info("CV Mean F1:  %.3f (+/- %.3f)", np.mean(cv_f1), np.std(cv_f1))
+    log.info("Purged CV Mean F1:  %.3f (+/- %.3f)", np.mean(cv_f1), np.std(cv_f1))
 
     # Train final model on all labeled data
     log.info("Training final model on all %d labeled samples...", len(labeled))
     scaler = StandardScaler()
     X_scaled = scaler.fit_transform(X)
 
-    final_model = GradientBoostingClassifier(
+    final_model = _build_model()
-        n_estimators=300,
-        learning_rate=0.05,
-        max_depth=4,
-        subsample=0.8,
-        min_samples_leaf=20,
-        random_state=42,
-    )
     final_model.fit(X_scaled, y)
 
     # Feature importances
@@ -357,48 +466,7 @@ def train_model(rows):
         bar = "#" * int(imp * 200)
         log.info("  %-30s %.4f %s", name, imp, bar)
 
-    # Extract optimal weights by aggregating importance per metric
+    weights = derive_metric_weights(feature_cols, importances)
-    # Map each feature back to its parent metric
-    metric_names = [
-        "puell_multiple", "mvrv_zscore", "reserve_risk", "rhodl_ratio",
-        "nupl", "fear_greed", "drawdown", "pct_above_200w_sma", "pct_above_lth_rp",
-    ]
-    feature_to_metric = {}
-    for m in metric_names:
-        feature_to_metric[f"score_{m}"] = m
-        feature_to_metric[f"raw_{m}"] = m
-    # Delta features map to their base metric
-    feature_to_metric["delta_30d_mvrv_zscore"] = "mvrv_zscore"
-    feature_to_metric["delta_30d_nupl"] = "nupl"
-    feature_to_metric["delta_30d_puell_multiple"] = "puell_multiple"
-    feature_to_metric["delta_30d_reserve_risk"] = "reserve_risk"
-    # Interaction terms split evenly between constituent metrics
-    # mvrv_x_nupl -> mvrv_zscore + nupl
-    # puell_x_reserve -> puell_multiple + reserve_risk
-
-    metric_importances = {m: 0.0 for m in metric_names}
-    for name, imp in feat_imp:
-        if name in feature_to_metric:
-            metric_importances[feature_to_metric[name]] += imp
-        elif name == "mvrv_x_nupl":
-            metric_importances["mvrv_zscore"] += imp / 2
-            metric_importances["nupl"] += imp / 2
-        elif name == "puell_x_reserve":
-            metric_importances["puell_multiple"] += imp / 2
-            metric_importances["reserve_risk"] += imp / 2
-        # days_since_ath maps to drawdown conceptually
-        elif name == "days_since_ath":
-            metric_importances["drawdown"] += imp
-
-    # Normalize weights to sum to 1
-    total_imp = sum(metric_importances.values())
-    if total_imp > 0:
-        weights = {k: round(v / total_imp, 4) for k, v in metric_importances.items()}
-    else:
-        weights = {k: round(1 / len(metric_importances), 4) for k in metric_importances}
-
-    # Sort by weight descending
-    weights = dict(sorted(weights.items(), key=lambda x: x[1], reverse=True))
 
     log.info("\nOptimal Metric Weights:")
     log.info("-" * 50)
@@ -413,6 +481,7 @@ def train_model(rows):
     log.info("COMPARISON BACKTEST: ML-Weighted vs Equal-Weight")
     log.info("=" * 60)
     comparison = run_comparison(rows, weights)
+    out_of_sample_comparison = run_out_of_sample_comparison(labeled, fold_results)
 
     # Build output
     result = {
@@ -424,6 +493,9 @@ def train_model(rows):
             "mean_f1": round(float(np.mean(cv_f1)), 4),
             "mean_precision": round(float(np.mean(cv_precision)), 4),
             "mean_recall": round(float(np.mean(cv_recall)), 4),
+            "validation_method": "purged_expanding_window",
+            "label_horizon_days": LABEL_HORIZON_DAYS,
+            "folds": fold_results,
         },
         "training_info": {
             "n_samples": len(labeled),
@@ -435,66 +507,47 @@ def train_model(rows):
             "model": "GradientBoostingClassifier",
         },
         "comparison": comparison,
+        "out_of_sample_comparison": out_of_sample_comparison,
         "trained_at": datetime.now(tz=__import__('datetime').timezone.utc).isoformat(),
     }
 
     return result
 
 
-def run_comparison(rows, ml_weights):
+def _composite_score(row, mode, ml_weights=None):
-    """Compare ML-weighted scoring vs equal-weight scoring across score brackets."""
+    scores = [row[f"score_{k}"] for k in SCORE_KEYS]
-    # Metrics used in scoring (maps to score_* columns)
+    if mode == "equal_weight" or not ml_weights:
-    score_keys = [
+        return sum(scores) / len(SCORE_KEYS) * 10
-        "puell_multiple", "mvrv_zscore", "reserve_risk", "rhodl_ratio",
+    equal_weight = 1.0 / len(SCORE_KEYS)
-        "nupl", "fear_greed", "drawdown", "pct_above_200w_sma", "pct_above_lth_rp",
+    weighted_sum = sum(row[f"score_{k}"] * ml_weights.get(k, equal_weight) for k in SCORE_KEYS)
-    ]
+    return weighted_sum * 10
-    n_metrics = len(score_keys)
-    equal_weight = 1.0 / n_metrics
 
-    brackets = [
-        (0, 20, "Extreme Caution"),
-        (21, 40, "Caution"),
-        (41, 55, "Neutral"),
-        (56, 70, "Moderate Opportunity"),
-        (71, 85, "Strong Accumulation"),
-        (86, 100, "Extreme Accumulation"),
-    ]
 
-    # Only use rows with forward returns
+def _summarize_brackets(scored_rows, score_key):
-    scored_rows = [r for r in rows if "fwd_365d" in r]
+    results = []
-
+    for low, high, label in BRACKETS:
-    results = {"equal_weight": [], "ml_weighted": []}
+        days_in = [r for r in scored_rows if low <= r[score_key] <= high]
-
+        if not days_in:
-    for mode in ["equal_weight", "ml_weighted"]:
+            results.append({
-        for r in scored_rows:
+                "range": f"{low}-{high}", "label": label,
-            scores = [r[f"score_{k}"] for k in score_keys]
+                "days": 0, "avg_365d": None,
-            if mode == "equal_weight":
-                composite = sum(scores) / n_metrics * 10
-            else:
-                weighted_sum = sum(r[f"score_{k}"] * ml_weights.get(k, equal_weight) for k in score_keys)
-                composite = weighted_sum * 10
-            r[f"composite_{mode}"] = composite
-
-        for low, high, label in brackets:
-            days_in = [r for r in scored_rows if low <= r[f"composite_{mode}"] <= high]
-            if not days_in:
-                results[mode].append({
-                    "range": f"{low}-{high}", "label": label,
-                    "days": 0, "avg_365d": None,
-                })
-                continue
-            returns_365 = [r["fwd_365d"] for r in days_in]
-            win_rate = len([r for r in returns_365 if r > 0]) / len(returns_365) * 100
-            results[mode].append({
-                "range": f"{low}-{high}",
-                "label": label,
-                "days": len(days_in),
-                "avg_365d": round(sum(returns_365) / len(returns_365), 2),
-                "median_365d": round(sorted(returns_365)[len(returns_365) // 2], 2),
-                "win_rate_365d": round(win_rate, 1),
             })
+            continue
+        returns_365 = [r["fwd_365d"] for r in days_in]
+        returns_sorted = sorted(returns_365)
+        win_rate = len([r for r in returns_365 if r > 0]) / len(returns_365) * 100
+        results.append({
+            "range": f"{low}-{high}",
+            "label": label,
+            "days": len(days_in),
+            "avg_365d": round(sum(returns_365) / len(returns_365), 2),
+            "median_365d": round(returns_sorted[len(returns_sorted) // 2], 2),
+            "win_rate_365d": round(win_rate, 1),
+        })
+    return results
 
-    # Print comparison
+
+def _log_comparison_table(results):
     log.info("\n%-18s | %-8s %-8s %-8s | %-8s %-8s %-8s",
              "Bracket", "EQ Avg", "EQ Med", "EQ Win%", "ML Avg", "ML Med", "ML Win%")
     log.info("-" * 80)
@@ -508,6 +561,47 @@ def run_comparison(rows, ml_weights):
         log.info("%-18s | %-8s %-8s %-8s | %-8s %-8s %-8s",
                  eq["label"], eq_avg, eq_med, eq_win, ml_avg, ml_med, ml_win)
 
+
+def run_comparison(rows, ml_weights):
+    """Compare final ML-weighted scoring vs equal-weight scoring across all labeled rows.
+
+    This is retained for backwards compatibility with existing output. It is an
+    in-sample/full-history comparison; prefer out_of_sample_comparison for model
+    selection decisions.
+    """
+    scored_rows = [dict(r) for r in rows if "fwd_365d" in r]
+    for r in scored_rows:
+        r["composite_equal_weight"] = _composite_score(r, "equal_weight")
+        r["composite_ml_weighted"] = _composite_score(r, "ml_weighted", ml_weights)
+
+    results = {
+        "equal_weight": _summarize_brackets(scored_rows, "composite_equal_weight"),
+        "ml_weighted": _summarize_brackets(scored_rows, "composite_ml_weighted"),
+    }
+    _log_comparison_table(results)
+    return results
+
+
+def run_out_of_sample_comparison(rows, fold_results):
+    """Compare fold-specific ML weights on validation rows only."""
+    validation_rows = []
+    for fold in fold_results:
+        weights = fold.get("weights", {})
+        for idx in fold.get("val_idx", []):
+            if idx >= len(rows) or "fwd_365d" not in rows[idx]:
+                continue
+            r = dict(rows[idx])
+            r["fold"] = fold.get("fold")
+            r["composite_equal_weight"] = _composite_score(r, "equal_weight")
+            r["composite_ml_weighted"] = _composite_score(r, "ml_weighted", weights)
+            validation_rows.append(r)
+
+    results = {
+        "folds": len(fold_results),
+        "validation_days": len(validation_rows),
+        "equal_weight": _summarize_brackets(validation_rows, "composite_equal_weight"),
+        "ml_weighted": _summarize_brackets(validation_rows, "composite_ml_weighted"),
+    }
     return results
 
 

tests/test_ml_optimizer_validation.py

@@ -0,0 +1,74 @@
+from datetime import datetime, timedelta
+
+from ml import optimizer
+
+
+def _row(date, returns=10.0, score_200w=10, score_drawdown=0):
+    row = {
+        "date": date,
+        "price": 100.0,
+        "fwd_365d": returns,
+        "score_puell_multiple": 0,
+        "score_mvrv_zscore": 0,
+        "score_reserve_risk": 0,
+        "score_rhodl_ratio": 0,
+        "score_nupl": 0,
+        "score_fear_greed": 0,
+        "score_drawdown": score_drawdown,
+        "score_pct_above_200w_sma": score_200w,
+        "score_pct_above_lth_rp": 0,
+    }
+    return row
+
+
+def test_purged_time_series_splits_remove_overlapping_forward_label_windows():
+    start = datetime(2020, 1, 1)
+    rows = [_row((start + timedelta(days=i)).strftime("%Y-%m-%d")) for i in range(900)]
+
+    splits = list(
+        optimizer.purged_time_series_splits(
+            rows,
+            n_splits=3,
+            label_horizon_days=365,
+            embargo_days=0,
+        )
+    )
+
+    assert splits, "expected at least one viable split"
+    for train_idx, val_idx in splits:
+        val_start = datetime.strptime(rows[val_idx[0]]["date"], "%Y-%m-%d")
+        latest_allowed_train_date = val_start - timedelta(days=365)
+        assert len(train_idx) > 0, "purging should keep non-overlapping expanding-window training rows"
+        for idx in train_idx:
+            train_date = datetime.strptime(rows[idx]["date"], "%Y-%m-%d")
+            assert train_date <= latest_allowed_train_date
+
+
+def test_run_out_of_sample_comparison_scores_only_validation_rows_with_fold_weights():
+    rows = [
+        _row("2020-01-01", returns=-10, score_200w=0, score_drawdown=10),
+        _row("2020-01-02", returns=-5, score_200w=0, score_drawdown=10),
+        _row("2020-01-03", returns=100, score_200w=10, score_drawdown=10),
+        _row("2020-01-04", returns=120, score_200w=10, score_drawdown=10),
+    ]
+    fold_results = [
+        {
+            "fold": 1,
+            "val_idx": [2, 3],
+            "weights": {"pct_above_200w_sma": 1.0, "drawdown": 0.0},
+        }
+    ]
+
+    comparison = optimizer.run_out_of_sample_comparison(rows, fold_results)
+
+    assert comparison["validation_days"] == 2
+    assert comparison["folds"] == 1
+    assert sum(bucket["days"] for bucket in comparison["ml_weighted"]) == 2
+    assert sum(bucket["days"] for bucket in comparison["equal_weight"]) == 2
+
+    extreme_ml = next(bucket for bucket in comparison["ml_weighted"] if bucket["label"] == "Extreme Accumulation")
+    assert extreme_ml["days"] == 2
+    assert extreme_ml["avg_365d"] == 110.0
+
+    caution_equal = next(bucket for bucket in comparison["equal_weight"] if bucket["label"] == "Caution")
+    assert caution_equal["days"] == 2

tests/test_scoring_engine_ml.py

@@ -0,0 +1,63 @@
+import math
+
+from scoring import engine
+
+
+def _complete_metrics():
+    return {
+        "fear_greed": {"value": 10, "classification": "Extreme Fear"},
+        "puell_multiple": {"value": 0.3},
+        "mvrv_zscore": {"value": -0.1},
+        "drawdown": {"value": 60.0, "ath": 250.0},
+        "price": {"price": 100.0},
+        "200w_sma": {"value": 120.0},
+        "reserve_risk": {"value": 0.001},
+        "rhodl_ratio": {"value": 50.0},
+        "nupl": {"value": -0.1},
+        "lth_realized_price": {"value": 120.0},
+        "hash_ribbons": {"buy_signal": False},
+    }
+
+
+def test_score_all_ml_normalizes_displayed_weights_and_contributions(monkeypatch):
+    monkeypatch.setattr(
+        engine,
+        "load_ml_weights",
+        lambda: {
+            "fear_greed": 0.40,
+            "puell_multiple": 0.20,
+            "mvrv_zscore": 0.15,
+            "drawdown": 0.10,
+            "pct_above_200w_sma": 0.05,
+            "reserve_risk": 0.04,
+            "rhodl_ratio": 0.03,
+            "nupl": 0.02,
+            "pct_above_lth_rp": 0.01,
+        },
+    )
+
+    scored = engine.score_all_ml(_complete_metrics())
+
+    assert scored["ml_mode"] is True
+    valid_metrics = [m for m in scored["metrics"] if m["score"] is not None]
+    assert scored["ml_weight_total"] == 1.01  # trained weights + small hash-ribbons fallback
+    assert math.isclose(sum(m["ml_weight"] for m in valid_metrics), 1.0, abs_tol=0.001)
+    assert math.isclose(
+        sum(m["ml_contribution"] for m in valid_metrics),
+        scored["composite_score"],
+        abs_tol=0.05,
+    )
+
+    hash_ribbons = next(m for m in valid_metrics if m["key"] == "hash_ribbons")
+    assert hash_ribbons["ml_raw_weight"] == 0.01
+    assert hash_ribbons["ml_weight"] == round(0.01 / 1.01, 4)
+
+
+def test_score_all_ml_preserves_classic_fallback_when_weights_missing(monkeypatch):
+    monkeypatch.setattr(engine, "load_ml_weights", lambda: {})
+
+    scored = engine.score_all_ml(_complete_metrics())
+
+    assert scored["ml_mode"] is False
+    assert scored["ml_error"] == "ML weights not found — run ml/optimizer.py"
+    assert "classic_score" not in scored