Gitcoin Grants Round 11 Governance Brief

Dear Funders League and Community stewards, 

Grants Round 11 (GR11) closed on September 23rd, 2021. As is our duty in Quadratic Lands, we are honored to perform our sacred duty to support digital public goods. We seek to honor this commitment by ratifying the round, and transferring matching funds from the Grants MultiSig to the non-custodial deployment contract. This unlocks payments from the round’s matching pool to all eligible projects.

Ask of the Stewards

We ask our Community Stewards to ratify the GR11 payout amounts as being correct, fair, and abiding to community norms, including the judgements and sanctions made by the Fraud Detection & Defense workstream of GitcoinDAO.

Summary of Results

  • Grants Round 11 (GR11) closed on September 23, 2021. It was the largest round in the history of Grants with $2.6 million raised for digital public goods including $965k from the matching pool and $1.59 million from the community across 425k contributions, comprising over 56x growth since Grants Round 1 in February 2019.

Grants Round 11

Our mission is to empower communities to build and fund the open web. We do that by giving builders opportunities to earn in a number of ways, including sponsored hackathons, ongoing bounties, and quarterly grants rounds. In particular, grants rounds are among our most significant economic vehicles. And because of the Quadratic Funding mechanism, they offer the community extraordinary involvement in deciding the allocation of public funds.

Thanks to the generosity of the Funders League and the Ethereum community, the multisig has over $7.5mm in total assets (excluding AKITA tokens) destined to bolster digital public goods. Approximately $965k will be used to pay out this round, leaving approximately $6.5 million in funding for future rounds. As was agreed when the multisig was formed, the social contract is that funds are to be used for public goods distributed through Gitcoin Grants. The first grants round (GR1) concluded in Feb 2019 with $44k funded.

In this eleventh grants round (GR11) we raised approximately $2.6 million, achieving over 56x growth in just 2.5 years. There were over 486 thousand crowdfunding contributions (+189k from GR10). These contributions were made by 16k unique users (+1.5k from GR10), both all-time highs.

Total funds raised by the community was $1.59 million, which when combined with the $965k matching pool funds, equaled almost $2.6 million raised to support digital public goods, also an all-time high. We continue to be astonished by the outpouring of community support for digital public goods.

Grants Round 11 marked significant growth in side rounds, which allow more ecosystems to participate in Grants by creating their own matching pools. Participating projects will continue to support the main pool, thereby supporting both their individual ecosystem and the larger ecosystem simultaneously.

Side rounds in GR11 included a $50k Gitcoin building Gitcoin (repeated from GR11) and a $50k round sponsored by Uniswap (alongside their $50k allocation to the primary matching pool). We are seeing significant interest in side round participation for GR12 and expect to see at least a half dozen side rounds for GR12.

GR11 saw the introduction of the Africa regional category, which was proposed and ratified by the community. We look closely at the impact of opening new categories, and were pleased to see that this new category unlocked dozens of new grants from the region. We are confident that this is only the beginning of growth in this region over the coming rounds. 

We are immensely thankful to Aragon for helping us seed the decentralized Governance (dGov) category again in GR11, building on the foundation created when the partnership was established in GR10. The GitcoinDAO uses many of the projects in this category for its governance and operations  and we are hopeful that this category will continue to support many projects that are  innovating in the area of decentralized governance.

Finally, we’d like to take this opportunity to thank the projects that are continuing to support digital public goods through matching grants:

Ethereum Foundation, @optimismPBC, @Synthetix, @iearnfinance, @0xPolygon, @DefianceCapital & Three Arrows Capital, #FerretPatrol, @chainlink, @YAMFinance, @realmaskbook, 1337 working group, $MEME, @1kxnetwork, 1inch, @binance, @harvest_finance, BadgerDAO, @BalancerLabs, @Krakenfx, @graphprotocol, @uniswap, @sushiswap, @knotmegan, @econar, @nanexcool, @bantg, @future_fund_, @rleshner, @Andrewdarmacap, @Jordanlyall, @tenQkp, @stakefish, @10b57e6da0, @fcmartinelli, @BasedProtocol, @bc_workshop, @ideamarkets_, SNX, Splunk, Auryn,, @ENSdomains, @CMSHoldings, @Pasta_DAO, @DoraHacks, @AragonProject, @UnstoppableWeb, @UseTeller, @integralhq, @harvest_finance, @epnsproject, daosquare

Sub-Round (Category) & Individual Grant Leaderboard

Here is the leaderboard for the top 10 grants awarded in each category for Grants Round 11. These are the final ratified results, updated on October 18, 2021.


GR11 Fraud Detection & Defense (FDD)

As we zoom back out and evaluate Grants Round 11 as a whole, we would be remiss to not also talk about the challenges of sybil attacks, collusion and the desires for bad actors to game the system. With our Fraud Detection & Defense workstream and partnership with BlockScience we have been able to keep ahead of the problems, but the red team versus blue team game certainly continues. BlockScience has posted their thoughts on the round, and we offer some details on the process and actions taken below.

Sybil Attacks, a form of fraud, and collusion constitute the existential threats quadratic funding faces. This workstream is tasked with minimizing the effect of these liabilities on the community. 

FDD Anti-Sybil ML L2 Stream

The Anti-Sybil ML L2 stream of FDD is responsible for detecting potential sybil accounts. They currently do this through a semi-supervised reinforcement machine learning algorithm run by BlockScience. 

During Q3, the stream found community contributors to stand up DAO cloud services and a DAO repo for hosting this algorithm. These contributors are now working on feature engineering, model training, model validation, and model supervision. 

In addition, the Evaluation Squads L2 stream sources, trains, and incentivizes participation in the human oversight & training of the algorithm in order to train the algorithm to think like “the Gitcoin user” rather than any individual. 

For GR12, this stream aims to create an on-chain forensics flagging system, a new community run ML algorithm, and statistical validation processes to ensure accuracy and fairness by decentralizing the results. 

Fraud Tax

The fraud tax represents the difference in allocations if the users suspected as sybil attackers were to have their matching eligibility shut off. The fraud tax is one metric which may be used to measure the effectiveness of the workstream. How is the Fraud Tax calculated? Check the math here


  • Fraud Tax Total = $5,787
  • Change RoR = (-$8,653)
  • % of pool = 0.61%
  • % of round = 0.24%
  • Est. Unbiased Sybil = 6.4%
  • Est. Sybil Users = 1,127
  • Flagged Sybil = 853
  • Effectiveness = 83%

A fraud tax of $5,787 is down over 50% from GR10 with both the percentage of the pool and the percentage of the total round having seen improvements. 

Sybil Detection Detail 

Using the calculations for before and after shutting off matching for sybil accounts allows us to compare how the round ended on the website to the final payout amounts adjusted for sybil-defense. 

These calculations have to happen after the round is over, after the ML algorithm is run and verified against statistical analysis, and after the grant disputes have all been settled. 

The sybil detection algorithm aggressiveness tunes the trade off between sensitivity and specificity of the flagging model. When it is set to 50%, the algorithm will be optimized to maximize sybil detection accuracy — even if at the cost of wrongly flagging non-sybil users — while values closer to 0% means that it will be optimized for minimizing false detection rate, but could miss out on detecting some sybil users. 

An aggressiveness score of 30% was used during GR11. During GR10, a lower aggressiveness model flagged 1,377 sybil accounts. This increase was possible because of the improved processes in the anti-sybil ml pipeline including a 12 hour cycle time for running the algorithm, feature improvements, and most importantly, a much higher number of human evaluations training the model to more accurately detect likely sybil accounts. 

Sanctioning of these accounts will affect the final amounts that grant creators receive pending steward ratification of the GR11 results. This will be explicity decided by a steward vote on snapshot.

FDD Policy L2 Stream

The Policy L2 stream exists to create and maintain policies affecting platform use and round participation. The stewards and contributors in this stream set definitions based on reviewer feedback and advise on judgements for flags, disputes, appeals, and sanctions during the round. 

The policy is held on a “living document” which can be found on our grants policy wiki

The items below are approved/denied/judged based on black/white interpretation of current policy, however, there are “gray areas” when novel grant types or situations occur. In these cases, the FDD will make a decision based on the norms and values of the community to then be ratified into policy via the steward ratification of the round payout. 

Full transparency to the community is available for Grant Approvals, and Grant Disputes. User Actions & Reviews is currently in “open review” allowing for select participation to stewards due to sensitive Personal Identifiable Information (PII) data and potential vulnerability to counter-attacks. 

The decisions are communicated via the round’s governance brief. Ratification of the round by stewards accepting the final payout amounts based on the sanctions adjudicated by the FDD workstream is an implicit endorsement of these decisions. 

Grant Approvals

New grants are all reviewed prior to being activated on the platform. Grants are evaluated for both their platform eligibility and for participation in the rounds they select.

There are multiple levels for approval:

Platform | Ecosystem | Round | Sub-Round | Side-Round

  • Platform – All grants allowed on
  • Ecosystem – A subset of grants supported by a continuous matching pool. The ecosystem establishes their own norms & values. (Ethereum, GitcoinDAO, Matic)
  • Round – Time based instance of a matching event. Grant participation rules are set by the ecosystem. (GR10, GR9, GR8, ZCash 1, Sia 1)
  • Sub-Round – A subset of grants allowed in the ecosystem and round which are eligible for a specific category of matching. (May be inclusive or exclusive – Set by ecosystem)
  • Side-Round – A separate ecosystem using the hype of a larger round to promote their ecosystem. Uniswap side round during GR11 is an example. Funds for matching provided by Uniswap, therefore they set the Ecosystem and Round participation policy. 

During GR11, all of the approvals were conducted by the FDD Grant Approvals L3 squad which had one Gitcoin Holdings team member facilitating their needs. This is a big step in decentralizing the process as previous rounds were judged by the Gitcoin Holdings team. Review the decisions made during GR11 here: Public Oversight of Grant Applications

The Gitcoin Holdings team currently has final say on grant approvals via their administrative control of access to the system. During GR11 they did not dispute any of the findings from the FDD and all actions were processed directly by the FDD. 

Grant Disputes

Grant disputes happen when a user flags a grant on the system or reports it in the gitcoin-disputes channel in Discord. Disputes are made public at the Gitcoin Disputes Twitter. During GR11, disputes were determined by the FDD workstream lead with advice by the FDD Policy L2 stream. Review the decisions made during GR11 here: Grants Disputes Public Oversight

User Review

Users found to be in violation of the Gitcoin Terms of Service and/or grants policy wiki may be blocked from accessing the site or have their contributions matching bonus turned off for grants rounds. This is most common for users found to be sybil. 

GR11 user reviews were determined by the FDD workstream with Gitcoin Holdings as oversight as the administrative hosts of user data. During GR11 they did not dispute any of the findings from the Anti-Fraud workstream.

Official Statements from Grants Round 11


The progress from GR10 to GR11 was impressive, but less than what we are capable of delivering. The FDD workstream was in its infancy during GR10, then had its first budget pass at the end of August. With only a few weeks prior to the round to onboard and train contributors, we are proud of the GR11 results and have excitement looking forward to what this community led group will accomplish to make GR12 the best round yet.

Disruption Joe, FDD Workstream Lead

Gitcoin Holdings Inc.

The Gitcoin Team is committed to working with the Community Stewards and greater DAO to progressively decentralized governance over Gitcoin Grants. Our goal is to make significant headway each round, and we believe we were able to do so in GR11. Final round payouts will be ratified by the Community Stewards through a Snapshot vote. 

We thank the Stewards and the greater Gitcoin Community for participating in this process, and look forward to our continued partnership. Our request as stated above:

We ask our Community Stewards to ratify the GR11 payout amounts as being correct, fair, and abiding to community norms, including the judgements and sanctions made by the Fraud Detection & Defense workstream of GitcoinDAO. 

– The Gitcoin Team


BlockScience, along with the Token Engineering and Gitcoin communities, have been working for more than a year to improve fraud detection on the Grants platform. It has simultaneously been the front lines of a Sybil battleground, and an enlightening research initiative in gaming and attacks in digital identity-dependent systems like Quadratic Funding. 

The process is ongoing and the technical work, along with countless hours from contributors and volunteers, has supported huge improvements in Sybil detection and provided valuable research for the ecosystem. We are excited to see the machine learning algorithm we’ve built in use, and look forward to a continuing partnership with the Gitcoin community.

-The BlockScience Team