Passport’s Model Based Detection

  • This evolution has two parts: connection-oriented cluster matching (COCM) and Passport’s new Model based detection.
    • COCM analyzes the “uniqueness” of donation patterns to decrease the impact of users who donate towards the same sets of projects, thereby decreasing the impact of airdrop farmers and sybil attackers who tend to follow a set of donation patterns.
    • Passport’s Model Based Detection analyzes the on-chain history of donors to increase the impact of users who have proven transactions and history on Ethereum.
  • Together, we estimate that this provides a 20% higher efficacy in distributing matching funds for GG20.


  • Does this mean that people with new wallets (e.g. new to web3) will not qualify for matching?
    • Wallets with no history will likely not contribute to matching funds in GG20, but every donation will still go directly to the project they’re supporting!
    • We’ve focused on bettering the web3 experience for this round and look forward to providing support for non-web3 natives in future rounds, likely through fiat on-ramping.
  • Where can I see my Passport’s Model Based Detection score?