rename timing "insider" verdict -> "last-minute" (honest label)

The lead-time gate in validate_timing.py is a COPYABILITY heuristic, not
proof of inside information: a short entry->resolution lead can be a
genuine insider OR just someone who trades fast-resolving markets (live
sports, hourly) well — indistinguishable, and irrelevant for copy
purposes since either way the window is too tight to mirror. Rename the
verdict accordingly. Gate unchanged (median lead >=24h to count as a
copyable sharp; 50 sharps). Docs updated to current p80 numbers
(218 profile matches, 62/83 forward-profitable, 50 sharps).

The genuine insider-detection scanner (insider.py, z-score / Bubblemaps
fingerprint) keeps its name — that's a different, correctly-named thing.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

FINDINGS.md

@@ -179,17 +179,21 @@ validate June) under p80 finds **218 wallets** matching the profile; forward,
 (The original flat-$200 run found 69 wallets, 25/37 forward, +11.7%.)
 
 Then the decisive copyability filter, `live/validate_timing.py`: a near-100% win
-rate is either foresight (copyable) or last-second information (uncopyable). The
-tell is **entry->resolution lead time on winning conviction bets**. Of the 69,
-**21 were insiders** (median lead < 6h — you can't mirror them), leaving **23
-validated copyable sharps** (`watch_sharps.json`) with multi-day leads. The
-standout `0x60ec1744…` held 80% win over **1,017 forward conviction bets**; even
-the suspiciously-perfect `0x72e1…` (99/100% win) enters ~7 days early — a real
-forecaster, not an insider. These 23 are surfaced live on jaxperro.com/trading.
+rate is only useful if we can mirror it. The tell is **entry->resolution lead
+time on winning conviction bets** — this is a copyability heuristic, NOT proof of
+inside information (a short lead can be a genuine insider or just someone good at
+fast-resolving markets; we can't tell, and for copy purposes it doesn't matter).
+Of the 218, the gate drops the **"last-minute" wallets** (median lead <24h — you
+can't get the trade in before resolution), leaving **50 validated copyable sharps**
+(`watch_sharps.json`) with multi-day leads. The standout `0x60ec1744…` held 80%
+win over **1,017 forward conviction bets**; even the suspiciously-perfect `0x72e1…`
+(99/100% win) enters ~7 days early — a real forecaster, clearly not last-minute.
+These 50 are surfaced live on jaxperro.com/trading.
 
 *Lesson: score conviction bets, not all bets; require avg entry ~0.4-0.6 (edge,
-not favorites); and gate on lead time to drop insiders. That funnel produces a
-copyable, forward-validated set — the strongest evidence in this project that
+not favorites); and gate on lead time to drop last-minute (un-mirrorable) wallets.
+That funnel produces a copyable, forward-validated set — the strongest evidence in
+this project that
 followable skill exists.*
 
 ## Repo layout

README.md

@@ -194,9 +194,10 @@ win rate ≠ edge, again. (The `value`/longshot archetype — wallets that beat
 **What does work (the repeatable find).** Scoring wallets on their **high-
 conviction bets — the top 20% by stake size (per-wallet p80, not a flat $200)** —
 which win 70–80% on genuinely-uncertain (~0.4–0.6)
-markets — trained pre-June and validated June: **25/37 stayed profitable forward
-(p=0.024)**. A lead-time gate (`validate_timing.py`) then drops the ~30% that are
-uncopyable insiders (entry <6h before resolution), leaving **23 validated
+markets — trained pre-June and validated June: **62/83 stayed profitable forward
+(p≈0)**. A lead-time gate (`validate_timing.py`) then drops wallets whose wins come
+too close to resolution to mirror — "last-minute" entries (median lead <24h), which
+may be genuine insiders or just fast-market specialists — leaving **50 validated
 copyable sharps** (`watch_sharps.json`), surfaced live on
 [jaxperro.com/trading](https://jaxperro.com/trading). One held **80% win over
 1,017 forward bets**. This is the strongest evidence in the project that

live/README.md

@@ -83,11 +83,13 @@ The per-wallet cutoff reproduces flat-$200's win-rate lift while adapting to eac
 wallet's scale (a whale's $200 bet isn't conviction; a minnow's is).
 
 - `conviction_scan.py` — train pre-June / validate June on conviction bets →
-  69 matches, **25/37 profitable forward (p=0.024)**. → `conviction_wallets.json`.
+  218 matches, **62/83 profitable forward (p≈0)**. → `conviction_wallets.json`.
 - `validate_timing.py` — the copyability gate: entry→resolution **lead time** on
   winning conviction bets separates copyable sharps (multi-day lead) from
-  uncopyable insiders (<6h). Drops 21 insiders → **23 validated copyable sharps**.
-  → `watch_sharps.json` (shown live on jaxperro.com/trading).
+  "last-minute" wallets (median lead <24h) we can't mirror in time — could be
+  genuine insiders or just fast-resolving-market specialists, can't tell which.
+  → **50 validated copyable sharps** in `watch_sharps.json` (shown live on
+  jaxperro.com/trading).
 
 Identifiers for a follow-worthy wallet: on its **≥$200 bets** — win ≥65%, avg
 entry 0.35–0.70 (edge, not favorites), +copy-ROI, FDR-significant, **median lead

live/validate_timing.py

@@ -1,11 +1,14 @@
 #!/usr/bin/env python3
-"""Insider-vs-sharp check on the standout conviction wallets.
+"""Last-minute-vs-sharp check on the standout conviction wallets.
 
-A near-100% win rate is either genuine foresight (copyable) or information /
-last-second entry (uncopyable insider). The tell is entry->resolution lead time
-on their WINNING conviction bets:
-  * mostly < 1h before resolution  -> insider/news-reaction, you can't follow it
+This is a COPYABILITY heuristic, not proof of inside information: a near-100%
+win rate is only useful to us if we can actually mirror it. The tell is entry->
+resolution lead time on their WINNING conviction bets:
+  * mostly < 1h before resolution  -> last-minute, you can't follow it in time
   * hours-to-days of lead          -> a sharp you could actually mirror
+A short lead can mean a genuine insider OR just someone who trades fast-resolving
+markets (live sports, hourly) well — we can't tell which, and for copy purposes
+it doesn't matter: either way the window is too tight to mirror.
 """
 
 import json
@@ -30,7 +33,7 @@ def lead_profile(w):
         return None
     med = st.median(leads)
     u6 = sum(1 for l in leads if l < 6) / len(leads)
-    verdict = ("insider" if (med < 6 or sum(1 for l in leads if l < 1) / len(leads) > 0.5)
+    verdict = ("last-minute" if (med < 6 or sum(1 for l in leads if l < 1) / len(leads) > 0.5)
                else "borderline" if med < COPYABLE_MED_LEAD else "sharp")
     return dict(n=len(leads), med=med, u6=u6, verdict=verdict)
 
@@ -66,7 +69,7 @@ def main():
               f"{fw:>7}{fr:>7}{c['fwd_n']:>5}  {c['wallet']}")
 
     json.dump(sharps, open(os.path.join(HERE, "watch_sharps.json"), "w"), indent=2)
-    print(f"\n-> watch_sharps.json ({len(sharps)} copyable sharps, insiders filtered out)")
+    print(f"\n-> watch_sharps.json ({len(sharps)} copyable sharps, last-minute wallets filtered out)")
 
 
 if __name__ == "__main__":