👻Sybil Analysis
Last updated
Last updated
In this context, a sybil refers to an individual masquerading as multiple people to illicitly claim matching funds. The term originates from a woman named Sybil, known for her dissociative identity disorder, symbolizing the act of adopting multiple identities. This poses a challenge in Quadratic Funding, where donations from a diverse group of people lead to higher matching funds. Sybils, by drawing more funds under false pretenses, reduce the amount available for legitimate participants. Therefore, conducting sybil analysis is crucial to maintain the integrity of the funding process.
Any time there’s money on the table, there’s potential for some fraud and fudging of numbers. We have found that there are two main attack vectors that occur during grants rounds:
Donors “spray” donations throughout the round. They might be hoping for a POAP, an airdrop, or simply looking for ways to make a bot-controlled wallet look as though it were controlled by a human…
Grantees try to vote more than once for their own project, to increase their chances of receiving matching funds.
💡 In addition to this preparation, we highly encourage you to gateway donations with Gitcoin Passport — this is the best way to guarantee the uniqueness of donors and protect the fairness of the round. Refer to the Passport doc to know how to use it in your round.
Below are the ways in which you can analyse results and detect potential sybils.
When the round ends, focus on two specific tabs for detailed information (refer to the provided screenshot for guidance):
"Round Stats" Tab
"Round Results" Tab
The "Round Stats" are updated in real-time throughout the round, allowing you to monitor:
The total amount of donations received
The remaining matching funds at your disposal
The count of distinct contributors
The total number of contributions they have made
Additionally, at the bottom, there's a display showing the number of contributions each project has garnered, along with an estimated percentage of matching funds.
Round Results
Vote Coefficients File (CSV) - This file provides detailed data for every donation made during the round, enabling you to understand:
The identity of each voter
The project they voted for
The amount they contributed
Their Passport score, if it's relevant to your round
Their coefficient value, which indicates their eligibility for matching funds. This value is either 0 (no matching funds) or 1 (full matching). If the coefficient is a fractional number, like 0.63, it signifies that they received 63% of the potential matching funds. The extent of matching is determined by the voter's Passport score.
The significance of this data lies in its utility for conducting sybil analysis. Occasionally, sybils might achieve the necessary Gitcoin passport score and contribute. As a round operator, your goal is to ensure these sybils don't receive matching funds. To achieve this:
Access this file.
After identifying sybils, you have the option to edit this file by changing their coefficient from 1 to 0.
Following these modifications, you can then upload the updated file.
If your round is large than $50k, we should have a chat to better understand the requirements and we can provide an expert.
Indicators to be cautious of include:
The timing of the donations.
The donation amounts.
Passport Scores of the donors.
Projects attracting an unusually high number of voters.
Donors exclusively supporting a single project and no others.
The possibility of two projects receiving support from a common group of donors.
If these patterns appear overly consistent, it's likely indicative of a bot attack rather than human activity. Such uniformity suggests a deliberate attempt to manipulate the system.
