- WHAT WE PUBLISHED
- for tokens launched on Arc from 16 to 24 September 2026, the 24-hour results read "Unknown — input gap". That covers 199,914 token pages, the creator pages and bot creator lines that count them, and the 30-day Scorecard rows for traded, trading, farm-adjusted trading, creator sold share, liquidity and drawdown at 24 h.
- WHAT WAS WRONG
- PiChan's swap and liquidity records on Arc started on 24 September 2026 (swaps 05:44 UTC, liquidity 21:21 UTC), not at public mainnet on 16 September, so the first day of these launches was never read. The transfer records reached back to 16 September, but the start was not treated as exact, so the first launches of 16 September also read "input gap".
- CORRECTED TO
- swaps and liquidity events were read back to the public-mainnet block (21,068,653, 16 Sept 2026 00:00 UTC) and the cohorts of 16–24 September recomputed under the same rule (v3.1). Results, from the night of 19 Oct 2026: of the 199,914 token pages, none reads "input gap" now. Traded in the first 24 h and traders: Known for all 199,914. Still trading at 24 h: Known 194,261; Unknown 5,653 (no USDC pool). Creator sold share: Known 38,798; Unknown 161,116 (the creator never held the token 135,847, the creator's transactions came from an address not counted 17,399, minted to a vault 7,143, creator not recorded 727). Liquidity at 24 h: Known 50,173; Unknown 149,741 (no price observed 138,958, hook return delta 5,130, no USDC route 5,653). Drawdown at 24 h: Known 21,324; Unknown 178,590 (no USDC swap in the first hour 141,219, no swap after the first hour 31,718, no USDC pool 5,653). 30-day Scorecard rows for Arc, before (18 Oct night, launches of 17 Sept–16 Oct) → after (19 Oct night, launches of 18 Sept–17 Oct; each night moves on by one day): traded at 24 h: Unknown (input gap) → 46.6% (n = 79,209 tokens); trading at 24 h: Unknown → 20.8% (n = 78,276); farm-adjusted trading at 24 h: Unknown → 9.6% (n = 78,276); buyer retention at 24 h: Unknown → 7.6% (n = 870,842 buyers; 66,235 retained); creator sold share: Unknown → 59.7% (n = 27,333 tokens with a counted creator); median liquidity at 24 h: Unknown → 24.44 USDC, n = 27,278; median drawdown at 24 h: Unknown → 0.67% (n = 9,543); median quote volume at 24 h: Unknown → 0 USDC, n = 78,276. Each at confidence medium. Rows whose windows start before 16 September (trading at 7 and 30 days, farm-adjusted trading at 7 and 30 days, graduated at 7 days, hold time, still-open holds) stay Unknown (input gap). Across all 1,397 30-day rows of the 19 Oct snapshot, 229 went from Unknown to Known. That night computed 11,886 rows and replaced 3,989 earlier ones (kept, marked superseded). Rows that still read "input gap" are only windows that begin before 16 September 00:00 UTC, the first block PiChan reads.
- HOW WE FOUND IT
- internal check of the data coverage (stats marker `swaps_from_block` 22,469,756; 17 Oct 2026).
- WHAT ELSE IT AFFECTED
- daily rows for 16–24 September (refreshed one per night, 18–25 Oct), creator pages and the bot's creator lines for creators with launches in those days. Weekly Scorecard files are not affected (their launch weeks start later).
- ORIGINAL
- not archived. The snapshot of 18 Oct 2026, with these rows Unknown, was replaced by the next night's run, and PiChan kept no dated copy of it. The corrected rows are in the weekly file 2026-W42.json, written 19 Oct 2026, 00:30 UTC (https://pichan.app/scorecard/arc/2026-W42.json).
Corrections log
When PiChan gets it wrong, it says so here.
Every correction to a finding, a Scorecard number or the methodology is logged in public, with the date, what changed and why. The original stays visible. Nothing is quietly edited.
1Corrections logged
Log starts (first day on Arc)
Latest correction, UTC
The log
Newest first · times in UTC
Spotted something?
Tell PiChan it got something wrong.
Send the token or Scorecard line, what you think is wrong, and the transaction that shows it. Every report is checked against the chain. Your identity is never shared with the project concerned.
We reply with what we checked, even when nothing changes.