# Virality Expert V9: X before the gate (side task)

## READ THIS FIRST
- **Verdict: X before the gate is NOT a filter. Do not add it.** Most v3 coins are silent on X before we buy, winners and losers alike.
- **69 wide trades (artifact coin removed), X search by full contract address, 24 h before each gate (twitterapi.io, 82 calls):**
  - 57 of 69 had **zero** posts before the gate. Only 12 had any.
  - Only 2 had a 10k+ account before the gate: 0x99f2e3420c1071bff35b8aef98a1339ea823ce24 (StrongHedge, 170k, 0.87x) and 0xcb142bca82c9c852ef33c5100a6f26577ad40f3c (a 10.5k "vote" spam account, 1.38x). Neither hit 5x.
  - Index check: the same search finds posts in the 24 h AFTER the gate for 53 of 69 coins (897 posts), so the zeros are real silence, not a broken search.
- **Loud vs quiet ($500, ladder (a) all out at 5x, 0.3x stop, my swap-by-swap pricer):**

| Group | First half | Second half |
|---|---|---|
| All | 36 trades, 2.25x, $1,612/day | 33 trades, 1.88x, $1,620/day |
| Any X post before gate | 6, 2.05x, 2 hit 5x | 6, 0.60x, **0 hit 5x** |
| Quiet before gate | 30, 2.29x, 13 hit 5x | 27, 2.17x, 11 hit 5x |
| Skip the loud ones (rule) | 30, 2.29x, **$1,387/day** | 27, 2.17x, $1,753/day |

- Skipping loud coins lifts the multiple a little but **loses $ a day in the first half** (fewer trades). Ladder (b) is the same story: $1,352 vs $1,509 a day first half, $3,260 vs $3,127 second half.
- The second half result (0 of 6 loud coins hit 5x) has about a 7% chance by luck alone (11 of 33 coins hit 5x). Not enough to act on.

## What the pre-gate posts actually are
- Almost none are people. They are bots: "DexScreener Paid", "Trending DEX Alpha Alert", "YOUR vote matters", "SCAM ALERT" (RaOnChain), "migration detected".
- In the second half all 6 loud coins had paid or vote bot posts. Those bots only show up from 1 Oct, so the half split is partly just "bots started posting".
- Real early calls by people with reach: only StrongHedge on 0x99f2e3420c1071bff35b8aef98a1339ea823ce24. It stopped at 0.87x.
- This matches virality map v1: winners stay quiet on X early. X comes after the move, not before.

## Two extra checks
- **DexScreener paid profile before the gate** (DexScreener orders API, real payment time, works live, no X needed): 28 of 69 paid before the gate. Paid early: first half 2.13x vs 2.39x, second half 1.52x vs 2.02x. Slightly worse both halves, but skipping them cuts $ a day in both halves ($846 vs $1,612; $1,362 vs $1,620). Not a filter.
- **X posts in the first hour AFTER the gate** (a possible hold or exit signal): flips sign. First half: posts 1.75x vs none 2.37x. Second half: posts 2.42x vs none 1.38x. Not a signal.
- Descriptive only (lookahead, not tradable): coins with more posts in the 24 h after the gate did much better (second half 2.44x vs 0.14x). X follows price.

## Data
- Queries: `<contract> since_time:<gate−24h> until_time:<gate>`, Latest, up to 5 pages. Same for gate to gate+24 h and gate to gate+1 h.
- Gate times and outcomes: wide trade list (sw-a2-16-wide-trades.csv) priced in V6 wide.

## First principles score
- **8 of 10.** Every coin searched by its full contract address with a time window; silence checked against the after-gate search; outcomes priced swap by swap.
- Missing: ticker searches (copycat and no-CA posts), and Telegram, where early calls on this chain likely live.

Version: v1, 6 Oct 2026, 16:24 Bangkok. Latest version: http://159.223.88.168/hx-c9ois/sw/sw-virality-v9.md