Hi @connor ! It’s there already defined a date when funds are going to be distributed to the projects ?
Hi everyone, thank you for your generous patience in waiting for revised matching calculations. We’re happy to finally have the updated results, which you can find here. We’re also extremely grateful to everyone who commented, provided feedback, criticism, questions, and more. It shows the community truly cares and wants to help make Gitcoin great.
I also want to specifically thank @epowell101 @PouPou @umarkhaneth and many others from Gitcoin and ODC who contributed their time and skills to this analysis.
In aggregate over 20 different tests were run on all transactions from the 5 rounds, many of which were tools built in the ODC hackathon. These tests considered factors like the seed funding of a wallet, number of transactions, interactions with other donors, interactions with various dapps (DEX’s, Tornado, Disperse, etc), likely airdrop farmers, clustered/consecutive transactions between different wallets, and more. Addresses were also cross-referenced with externally maintained Sybil/bot lists, as well as checked for interactions with addresses on those lists.
We then backtested the output of all these metrics against previously identified Sybils, suspect addresses discovered manually in this round, and checked the correlation with passport scores. Many of these tests did not appear to be significantly correlated with suspected Sybils, or they produced a large number of seemingly false positives. As a result, we decided on a select few tests to use for the final Sybil squelching, which appeared to be most effective at catching likely Sybil attackers while minimizing the number of honest users caught in the crosshairs. These tests included:
Suspicious seed - verify that the first incoming transaction (regular transaction and not internal transaction) is seeding the wallet = providing gas to the wallet. If the first transaction is not a seed, it is suspicious. This happens when the first normal transaction is an outgoing transaction. It is possible if it was funded from an internal transaction.
Clustered addresses - The LCS test runs on any address that has less than 10 transactions. It then compares the transaction history of that address with all other addresses looking for similar consecutive transactions. If the number of consecutive transactions is above 3 it raises the flag.
Less than 5 transactions - any fresh wallet with under 5 total transactions flagged.
Detected by stakeridoo (ODC) - Detected sybils using the fly trap method. He analyzed addresses that have received an airdrop and he managed to find clusters of addresses that sent the funds back to the main wallet demonstrating the connection between these wallets. More details here: CharmVerse - the all-in-one web3 space
Detected by doge (ODC) Analysis is here: https://github.com/DistributedDoge/sybilscope/blob/master/sus_on_chain.ipynb
He used flagged wallet by hop protocol to detect Sybils and as well as jacard similarity using these flags for similar donation patterns
These results were combined with the initial Sybil tests described in the first post of this thread, along with some additional manual reviews and flagging.
When looking at the results sheet, please note that the “Total Received USD” amount is what is reported by the platform before this additional analysis and revisions. However that USD amount does by default remove donations less than $1 and donations from failed passport scores.
Sybil/fraud detection and prevention on a decentralized protocol has certainly proved to be more difficult than on our previous (centralized) grant platform, where we had more and different data to leverage. That said as we build more robust on-chain analytical tools, this process will become easier and faster. I am (personally) excited to explore features like a “trust bonus” system (ie. a variable trust coefficient rather than a binary yes/no). This would mean the highest “verified” users have their donations matched the most, so building up a strong reputation over time is important. Future experimentation with pairwise bonding, plural QF, and other anti-collusion mechanisms may also prove to be effective at limiting bad actors.
Gitcoin will likely share a more detailed analysis of the data from the Beta Round and the calculated results soon, so keep an eye out for that. I believe @ evan will also be sharing more details and takeaways from ODC’d perspective. We look forward to collaborating with the ODC and others in the community more going forward.
From this point we still plan to wait 5 days for review and feedback on the revised results before moving to a Snapshot vote, which if passed will be followed by payouts shortly after. If you have any questions, please don’t hesitate to reply in this thread. Thanks all!
13 Likes
Hi Connor,
I am Andrei from WarmOFF project in the climate solutions round. I would like to express my gratitude for the efforts that you and the Gitcoin team have put into this work, along with other contributors such as the ODC members.
I have few comments on the final results. First of all I want to tell everyone that, as we stated in our project’s description, we were looking to raise $7000 - $10000, so 4.6k in matching + 1.6k in direct contributions is not that far from our goal, and so the money is not the reason I am writing this post. I am writing this post so that maybe you can take things into consideration.
Lets address the contributors count. I am 100% sure that the legos you used also flagged non-sybil donations because I managed to personally onboard 25-30 people. Also, my partners pitched our idea to multiple persons, and we sent around 50 e-mails per day with LinkedIn targeting. Our team is composed of me and two university professors, and so we have great reach.
The main thing that I think flagged lots of our donors as sybil is the fact that they are new to crypto, and so I want to take you all through our process of onboarding new people:
Few takeaway from this:
1. Many of them never had a Metamask wallet before or had one with no transactions;
2. The process of onboarding using passport is quite hard for a non Web3 person;
3. Because people we pitched or project to were new to Web3, we had to give them instructions on how to deposit funds in their wallet. They all used the same 2-3 ramps that are mentioned in the doc;
4. Gas fees as you know were so high that people didn’t want to pay $20-$30 in fees, and so they ended up contributing with larger sums;
Regarding your points:
-
“Less than 5 transactions - any fresh wallet with under 5 total transactions flagged.” : I believe lots of our contributors made their wallets for the single purpose of donating to our project so they don’t have 5 transactions;
-
“Clustered addresses - The LCS test runs on any address that has less than 10 transactions. It then compares the transaction history of that address with all other addresses looking for similar consecutive transactions. If the number of consecutive transactions is above 3 it raises the flag. :” They all have the same transactions because they all did the same thing: make a wallet get money in the wallet donate;
-
“Suspicious seed - verify that the first incoming transaction” : The first incoming transaction of many of our contributors were actually coming from the same 2-3 wallets. As you can see at the 3rd point in the doc I linked, we advised people on how to get ETH in their wallet. I know many of them used ramp.network because of their direct relationship with Revolut (Romania is the second biggest user of fintech app Revolut)
-
“Detected by stakeridoo (ODC)” : I don’t belive that any of our contributors were flagged for airdrop farming
How should project like ours supposed to get people involved?
• Our project is solely focused on scientific research and does not utilize cryptocurrencies or web3 technologies. Consequently, our supporters may not possess extensive knowledge in this domain.
• The methods employed for sybil detection seem to also flag newly onboarded individuals, thereby potentially affecting legitimate newcomers to our project.
Final thoughts: As I wrote in the beginning, the money is not the issue here. The issue is the contributors count and how that make communities feel about Gitcoin and other projects like this. If my project got from 43 contributors to 11 (and we are a legit project with deep ties in the academic space in Bucharest, Romania), what is going to make other grantees think that they will not have the same experience. As I stated at the beginning of this post, our goal was to raise 7-10k for our project and so we planned for this event ever since GR15 (you cand look up our messages with Yuki from Gitcoin support). From the beginning we thought that getting 40-50 contributors would get us in our range. Our judgment was that we wouldn’t have a chance against top projects (as they got hundreds of contributors in the past rounds), and so we were shocked to see we finished in top 5 projects sorted by contributors.
How do you think we should continue forward with our conversation as we are interested in more details about what kind of sybil detection was done in the case of our project?
Once again thank you for all your work,
Andrei Mihailescu, WarmOFF Team
3 Likes
Hey @mihailescuandrei thanks for your thoughtful and detailed reply. Perhaps we can find a time to connect to discuss further. Its often a fine line between coordination and collusion and sybil defense necessitates caution to ensure fairness. Your efforts to onboard people sound positive and it would be good to talk through how this can be done in ways that do not cross any lines in future rounds.
8 Likes
Thank you for your quick response, @M0nkeyFl0wer. I also think that there is a fine line between coordination and collusion and sybil defense necessitates caution to ensure fairness. But flagging the contributors that have < 5 transactions in their wallet seems kinda excessive to me because no freshly onboarded person will have 5 transactions already. Also, I think this “line” should have been stated clearly at the beginning of the round so grantees know who to target and contributors know what they need to do so their donation will get matched. I will be DM ing you on Twitter, and maybe we can have an in-depth convo. Cheers!
3 Likes
Don’t disagree in principle. The flip side is the more specific we are about those guidelines the easier it is for scammers to navigate. Its a balancing act. That being said I personally feel strongly about wanting to be a great experience for people new to the space and I’m happy to talk more with you about this.
4 Likes
Thank you for the collaboration and excellent TL;DR Connor.
It may be worth reiterating that many individuals including Connor traced individual transactions through repeatedly - painstaking work.
ODC is more of a research and development community today - and the needs of Gitcoin and other round operators are for not just R&D but the implementation of these algorithms and human screening informed by the understanding of the round and the implications. I believe we ended up working well together and I look forward to further improving the collaboration including aligning on where that handoff should and could be between developing tools for the community in the open - a current ODC remit - vs. actually doing analysis that protects a particular set of funding rounds - where ODC can help however the purview of Gitcoin and other round operators.
4 Likes
I hear you. But the best example is the law. Each law states clearly what individuals can or can’t do, some chose to obey it, some chose to go past it. I belive it’s Gitcoin’s job to state clearly what the laws in this community are and than make your best to keep bad actors away. Not to change the law after the fact is consumed. Do you imagine a world where the laws are kept secret?
Also there are two things I dont fully understand:
- Why is there a stamp in the passport that verifies you have 1 transaction on your wallet? Why not make it 5 transactions and mandatory from the get go if you anyway flag the accounts with less than 5?
- You say you want a great experience for people new to the space but it doesn’t seem like so given that 75% of my grant’s contributors were disqualified because they were new to the space
4 Likes
Hey @connor, sincerely appreciate all yall tirelessly fighting sybil attacks.
You mentioned “Total Received USD” mentioned in the excel is the one before the Sybil analysis. Will it be possible to get the revised USD figures that were actually considered for each project?
I ask because in the climate round, Mycelia is on top, receiving max matching with $1409 raised from 23 contributions while Water Dao with $1352 from 26 contributions is getting $7848 and Refi Barichara with $2565 from 22 contributors is getting $14,761.
Seems off
2 Likes
Hi Andrei, I wanted to give you the benefit of the doubt so spent a good amount of time digging into this further manually. But first, this onboarding doc:
Is seemingly encouraging users to game the Passport system in order to “fake” a high enough score:
If you don’t get to a score of 15 we recommend to make 5 github repositories and add one of them to favorites and find someone with github account that can fork one of your repository; Also, because I know many of you don’t really use twitter, the twitter account should have 10+ tweets so you get more points.
One could make a case that this alone is encouraging collusion and to some extent cheating. From the eligibility policy:
Additionally, the following are not permitted in Gitcoin-operated rounds:
- Falsification - Any type of hacking to falsify a contribution is prohibited. No encouraging or enabling Sybil attacks or other forms of malicious manipulation of the grants platform or the Gitcoin community
- Grantees can be eliminated from consideration in the round if they are found to be encouraging or enabling Sybil attacks or other forms of malicious manipulation of the grants platform or the Gitcoin community
But as for actually doing a deeper dive into the transactions - to me, looking closer only reinforced the automated detection outcome. I find it very hard to believe cold outreach with paid LinkedIn targeting led to 80-90% of your donors using the same exact CEX on-ramps, having extremely similar Sybil scores and stamps, and following the same exact on-chain patterns (CEX withdrawal, donate to Warmoff alone, or donate small amounts to 1 or 2 other grants and a larger amount to Warmoff alone, and then never making another on-chain tx again). In the Climate round as a whole, 60% of tx’s were ETH and 40% DAI. For Warmoff, 2 tx’s out of 46 (4%) were DAI, and the rest were in ETH with the majority having almost identical patterns, which seems statistically unlikely. Finally, we discovered some direct on-chain links between different donor wallets, although not many since most only withdrew from CEX, donated, and nothing else.
A persistent trend we’ve noticed over the years is that projects most likely to self-Sybil attack (or projects created mainly for the purpose of earning matching from Gitcoin) will have their entire Twitter presence focused around promoting their Gitcoin grant, and website lacking meaningful project details yet having a big CTA to donate on Gitcoin:
https://twitter.com/Warm_Off
https://warmoff.com/mission
https://warmoff.com/fundrasing
https://warmoff.com/progress
I apologize for being blunt here but it’s been a long day/week/month manually reviewing projects and on-chain data, to do our best to maximize funding going to users playing by the rules.
While I am sorry you will not get $35k, I think $5k should still be a significant amount to get your project off the ground. The alternative could be to take this to a formal appeal and dispute, where the outcome could be getting fully disqualified. The < 5 tx rule did likely catch some innocent new users as false positives, but it also helped us uncover many more cases of collusion and likely Sybils. We decided to use that metric as a result, without having it in mind before the round, because it was a net positive outcome for honest users. I am confident that if that was decided and known before the round, it would have only helped more Sybils get around it, while still hurting any new/less informed users.
5 Likes
Ooof, this sent me down a rabbit hole. Thank you for the question @solarpunkmaxi, I have good news and bad news (sort of).
First I’ll say you are absolutely right that having the data for the “revised total received” values would help a lot for anyone to sense check the results and get a better sense of how QF is weighting $ amounts vs. unique donors. Right now our revised matching calcs via the platform does not provide this value, but I will see if our product team can help add this in, or otherwise see if we can grab the data off-platform, either manually or automated.
The good(ish) news, the QF calcs are working correctly and those matches you see are right (based on the Sybil analysis inputs and subsequent changes), but without knowing the revised “total donated”, I can see why they seem off. The bad(ish) news, is that when investigating these 3 grants, I discovered more significant suspicious activity, and as I result I think we may need to reevaluate results and delay finalizing/payouts even further 
First - why you are seeing this:
WaterDAO
(some of this was caught via automated tests and some I actually just discovered):
WaterDAO grant’s address
1 of 2 gnosis safe
The two signers are aaronmandell.eth and ecotoken.eth :
both of which donated big amounts to their own grant (the multisig they control), $400 and $100
ecotoken clearly related / same org and the other ens is a real name of founder
https://twitter.com/AaronMandell
and then they got another $500 donation
from
who is also a co-founder / team member
https://twitter.com/ranchris11
So the “amount received” should be a lot lower for Water DAO given these didn’t/shouldn’t count, hence lower than expected match (and possibly needs additional tweaks)
Refi Barichara
Basically had all tiny donations ( > $20) except for one very large $2,400 tx
Grant address:
Big tx:
This transaction did count, but since it was one big outlier plus many much smaller transactions, the match is lower than what it would be with more “medium-sized” donations and the same total donations/sum amount.
However, while looking into this I noticed:
The donor address is hardly used but when it was first created the grantee address sent it some ETH right away:
Then the recipient sent it back right away as a donation in GR15:
And then besides that the donor address is hardly ever used except for a single donation to the grantee in the alpha round:
And then this big 1.3eth / $2.4k tx in the beta round. So - need to dig deeper, but something could be going on there…
Finally, diosdao.xyz by Mycelia:
Calculated to get the max match because they had a solid amount of donations that ranged from $40 to $200+ that were included for matching calcs. But since you mentioned and I looked into it more…
The grant’s address is:
It’s a 2 of 3 multisig:
With the creator:
And 2 other signers:
All 3 safe signers each donated about $100 to their own grant:
And then the multisig deployer (0x530) also directly funded and subsequently donated from tons of other wallets, basically every voter for this grant, only supporting their own grant and no others, and doing this in both the Climate and Metacrisis rounds.
Pattern below:
0x530 funding new wallet
New wallet donating to diosdao (0xE16)
_
_
_
_
_
_
And unfortunately, I found many more cases of this (feel free to go Etherscan digging 
). All that said - no matter the reason, this is absolutely against the rules, and will not be finalized for matching. Frankly, I am a bit frustrated at the amount of Sybil attacks we’ve seen this round and that were not prevented or discovered retroactively. I think so far we’ve already stopped more than $100k from going to attackers/colluders, and I want to make sure there aren’t more slipping through the cracks.
So TLDR - sorry everyone, I think we may need to revise the revised results again and further delay payouts, to ensure as much matching as possible goes to honest users
9 Likes
aah man really wouldnt want to be you right now. My comment was more to see if the Quadratic math was working fine as i assumed sybil filtering was done. Hats off to you on taking the time and having the patience to dig through this.
Would be interesting to get revised USD considered data from the product team if possible.
Do believe most grants are acting in good faith trying to maximise contributions to their grant, which is never an easy task esp with a lot of passport requirements being tech or Eth based. For the relatively newer ones the exact line between onboarding and colluding can be quite blurry, we are still learning after 4 rounds.
While its unfortunate the revised results need to be revised do hope it turns out for the best for all involved. Once again really appreciate the effort that goes into making these rounds as fair as possible. Looking forward to the 2nd and final revison 
2 Likes
Thank you for your detailed work and response, Connor. I am sorry I put you through more work, it was not my intention, as I really appreciate your efforts to keep sybils out of this ecosystem.
Few observations:
-
Is seemingly encouraging users to game the Passport system in order to “fake” a high enough score: As I stated earlier, the passport process was quite hard for a non-Web3 person. I did not make this doc from the get-go, but made it after receiving feedback from some close friends that ended up having different types of problems (ranging from topics like Metamask, buying crypto, passport, donating and fees). A couple of them actually had to wait a couple of days before they got their coins in the wallet because of the Metamask ramp provider. Also, some of the people that were new to the space actually spent up to 2 hours in the process of making a wallet, getting the money in the wallet, making a passport and contributing.
-
Grantees can be eliminated from consideration in the round if they are found to be encouraging or enabling Sybil attacks or other forms of malicious manipulation of the grants platform or the Gitcoin community : If you want to disqualify me for trying to make it easier for our community to support us, go ahead. @M0nkeyFl0wer said “my efforts to onboard people sound positive” and now you are talking about me “encouraging or enabling Sybil attacks or other forms of malicious manipulation”. I belive there is a “fine” line here too. As I stated before, I did not make this document from the get go, I made it after I received feedback and decided to help make thigs easier for my contributors.
-
“I find it very hard to believe cold outreach with paid LinkedIn targeting led to 80-90% of your donors using the same exact CEX on-ramps” - I never said that LinkedIn targeting was our way of getting most of our contributors, previously I just listed the methods we used. Also, I said I personally onboarded 25-30 people that were new to crypto & Web3, and so I think the majority of contributors were actually friends or acquaintances that were new to this ecosystem, and not LinkedIn targeted people.
-
The on-ramp situation: I have to take part of the blame here. I should have known that advising the contributors to purchase crypto from the same 2-3 addresses will not lead to good things, but in my defense, since the contributors were actually real people and there were no sybil activities going on, and on top of that, I heard from them that the passport process was so hard, I never thought that sybils will go by the passport gating. Also, take for example transak.com : tansak is going to offer you something like 3-4 options to buy crypto from, and they will be ranked by the final price including fees and so many of the contributors took the first option (the one with less fees) and went with it. And also, many of them used ramp.networks because of their direct link to fintech app Revoult which is widely used in Romania And so ye, I know many of them used the same on-ramps but again, I never thought to be a problem since they were acually human contributors to our project and I thought the reason of gitcoin passport is to verify authenticity of personhood and if they pass the passport they’ll be fine (as everywhere is stated, “Get your passport to a score of X so your donation will get matched.”).
-
“following the same exact on-chain patterns”: all they did was buy crypto - donating because they had never used crypto before and probably many of them will never use it again. You have to understand that we are not related to the Web3 ecosystem, we are purely a science project working on public goods ( maybe in the future we’ll find a connection between these two spaces but we are not focusing on this right now), and many of our contributors are working on the same space. Actually, many of them thought at first I try to pitch them some kind of pump-and-dump scheme. And so, they were happy to donate to my cause after I explained everything about Gitcoin, but there is a big chance that some of them will not actually use their wallets ever again.
-
"For Warmoff, 2 tx’s out of 46 (4%) were DAI ": I feel like I am repeating myself as I probably am but what is the point of telling a freshly onboarded person to buy ETH for the fees and DAI for the donation? Everything that we have done was for the sole purpose of making it a smooth experience for the people that wanted to contribute to our cause.
-
“Entire Twitter presence focused around promoting their Gitcoin grant, and website lacking meaningful project details yet having a big CTA to donate on Gitcoin” : Our project is a couple of months old, and again, we are a science project, it’s not like we are a community project. Our strategy was not to go after Gitcoin community donors because we are a very early project that does not have the progress that people want (have invested $2.000 of our own money and that’s it). We went after people who actually know us and our potential because we thought they are more likely to contribute. Twitter is not very used in Romania, so the official Twitter page of the project is not very active because our community is not there (yet anyways). Regarding the website, we have a page with its sole purpose of “Fundraising”, 2 mentions of that page on the home page, and that’s it as in website fundraising promotions. We do have a “Progress” page that you linked where we posted some of the pictures we took during our trials. As I said, we are a very early stage project and we don’t have the progress that people want mainly because of the large capital costs of some of our trials (a couple of grams of some type of material costs a couple hundred dollars). We do work 2-3 day/week in the lab and we have a 100+ page thesis that should be published next week.
Final thoughts: Getting a total of around $7k is what we thought from the beginning, and we got 6.4k ($4.7k matching + $1.7k donations). I would have been more than happy if this result had been obtained because the other projects had significantly more donors than us. But going down like this seems a bit frustrating to us since we know many of the donors you disqualified were actually non sybils. I am not taking my time to write this post because of the money, but because of the people that believed in our project and would, of course, want me to fight for their right to be counted as contributors.
Once again, I’m sorry that you have to put in more work given the recent discoveries that you posted, but I believe it’s for the best this way, so in the future you’ll have the experience of these past weeks. Cheers,
Andrei Mihailescu,
WarmOFF Team
EDIT: I asked some friends to send me their crypto purchase confimation e-mails so you can se why I started this onboarding doc. In the first case it took more than 10 h from when he initiated transfer till he got the coins in the wallet. In the second case, it took almost 24 hours for this process. You can’t tell me I am not in the right make this process more time efficient by telling them what to do step by step. Situations that made me do the onboarding doc - Google Docs
3 Likes
We have not discussed “Sum Sqrt” column yet which would be very helpful for fact-checking your QF math calculations, if it was to be reported truthfully.
Report, as it stands, is inconsistent and misleading in that “Sum Sqrt” column in the updated report does not reflect actual values used to calculate final matchings.
I am stating this after looking at Block 2030 (climate) grant which attracted 40$ in matching while having sum-of-sqrt of 184907254389 clear outlier when you compare this with sum-of-sqrt of surrounding grants. The project attracted much less money than it should if QF calculation would indeed use “Sum of square roots” value presented.
In previous report the same project was attributed ever-so-slightly smaller “sum-sqrt” yet it attracted 10x more matching (454$) which could hint that between reports manual intervention was done to penalize the project.
Since your post does confirm that, I think that what happened here is that someone forgot to update “Sum Sqrt” column to reflect final results after sybil-defense filtered out suspicious donations.
The problem with current approach is that “Sum Sqrt” column makes sense only in context of QF math calculations so reporting value of “Sum Sqrt” for anything other than final set of {all eligible donations remaining AFTER sybil defense} has absolutely no rationale and I cannot understand why it is being done that way?
Connor,
Tay and Mathilde from Mycelia here. Thanks for this in-depth analysis, and to open the debate in public.
We learn yesterday morning the accusation of Sybil attack on Mycelia. In the meantime, we understood the importance and value of explaining the situation in public, especially in the light of the context and we hope this will help to make the difference between a Sybil attack and a real public good with hard work and commitment to the Gitcoin community.
On-chain analysis is objective, but the context around is always subjective, especially in the context of a Beta round with high gas fees and no-tech savvy public like indigenous peoples.
TL;DR
-
Indigenous Peoples are among the poorest communities in the world (in terms of financial capital), with enormous technological discrimination and gap. How do we connect them to Gitcoin and quadratic funding ? Mycelia is working on that, and we have proof of our commitment to the Gitcoin community.
-
The 0x530 is the main address of Mycelia in relation with indigenous peoples, from where we spent money to pay back gas fees to indigenous peoples and allies. For them it was important to participate in their own campaign and be involved in the first steps. (See our defense for more info).
-
The report mentions that the campaign fund gets from our multi-sig to the 0x530 at the end of the campaign. Indeed, it is the case, and we took the time to ask Gitcoin first if we could use it to pay our work with indigenous peoples (see in our defense for more info).
-
Three irrelevant points were also mentioned in the report we read, the first one was the accusation of a large donor supposedly connected to Mycelia, not the case at all (more info in the defense), then a point on a wallet named myceliumvalley.eth contribute to the campaign and the report mentioned as “clearly related”, but it’s in not related at all to our project, on-chain analysis can prove it (and if anyone hold this ENS domain please reach out), and finally we got mentioned to have not given to any other grants of the campaign, also not the case, we gave to 36 different grantees in total (with tx proof see our defense for more info).
-
Mycelia got accepted on the Gitcoin Citizens Round, showing that the Gitcoin Team recognised the retroactive work we did for the community. Gitcoin | Explorer
Here is the full version of our defense : Defense Mycelia to Gitcoin - Google Docs
The primary mission of Gitcoin is to fund real public good, with real impact on the world. Sybil attack is a real threat that must be taken seriously, but as much as honest actors who are working with other realities on the edge like with indigenous communities where we need to adapt strategies to onboard them.
Fundamentally the core of this debate : Is Mycelia really taking the money for personal interest, with manipulation and a fraudulent mindset, OR is Mycelia is in a unique context of a Beta campaign, with unexpected high gas fees, and a specific reality with honest work to on-board indigenous peoples in web3?
This is the only question we should ask to grantees in this context of a unique Beta Round. On-chain analysis has its limitations and can devastate a whole project, especially as early as Mycelia. We don’t want to fight for the first place if Gitcoin decide so, this our first campaign we are also here to learn with the all community, but we ask to not being qualified as Sybil attack and take in consideration our defense, in respect and honor of all indigenous participants who are part of this project and believe in what we are doing.
We ask the Gitcoin Team and the Community to really treat this question in its fundamentals. We are here to stay, we love this community, we love on-board indigenous communities on Gitcoin, and we want to keep going for a long time.
Happy to open discussion in public.
From the Mycelia Team
5 Likes
Sadly there is no algorithmic way to distinguish legitimate behavior from malicious behavior. Any automated process that lets all legitimate behavior through, will necessarily let some illegitimate behavior through. Conversely, any process that catches all illegitimate behavior will also flag some legitimate behavior.
This is why human intelligence and analysis are critical,
regardless of what automation and signal intelligence are available
In terms of pure math: for grant matching amounts, the pairwise coordination penalty also needs to be considered. If most donors to a project donate to just that one project, their donations are given a penalty. The “number of donors” and “donation amounts” are too crude a data point, especially for small rounds.
8 Likes
I really appreciate the tremendous work that the Gitcoin team do in ensuring that money goes to fund the causes that matter.
Special thanks to @connor @ale.k @koday @umarkhaneth @M0nkeyFl0wer @jon-spark-eco @nategosselin @gravityblast
I don’t personally know the Mycelia team, however, I personally know the struggle of onboarding people into web3 when they are completely new to it, and even more with Indigeneous ppl, as I’ve worked with Mayan populations in Chetumal, Quintana Roo, México.
I’ve spent some chat with the Mycelia team and I’ve spoken to others that worked with them directly in Brazil. Reason why I think it is critical for the Gitcoin flagging system to add other manual mechanisms that include contextual factors, such as the realities of projects coming and working within the Global South and Global Indigeneous communities.
For the time sake, I think that going through interviews would make sense in getting the off-chain context that will enrich the insights between fraudsters and real ppl trying their best to improve the world.
For future rounds, a potential way forward could be delivering the QF through escrows, like the one of RaidGuild. That way, if the project is real and really doing what they promise, they would be getting the funds as they achieve progress on what they promised.
If projects participating in Gitcoin work with budgets and milestones, it won’t matter if they are in an early stage or in acceleration mode, or if they are from Global South or North, we all will have accountability of what it’s happening.
And those projects that have gamed the QF system won’t get away without delivering value or at least it will make the value delivery a thing harder to fake.
I’m just trying to contribute with some ideas. I respect and value both what the Gitcoin team puts lots of thought on and what the Gitcoin Grantees do to help improve the world.
Looking forward to this very interesting and highly valuable convo.
8 Likes
Big thanks to the Gitcoin team putting in such arduous work to ensure that funds are fairly distributed between grantees. It’s truly inspiring to see the team really go in on this and improve the process.
I’d like to add my two cents from the grantee perspective because I have onboarded thousands of people through our “Impact Onboarding” or “onboarDOING” events. I know how difficult it is to onboard indigenous communities, even university educated travelers from developed countries.
My first piece of advice to any grantees doing this work is to document it. It really benefits you to be able to show the actual work that goes into what you’re doing and taking a few photos and videos is very minimal effort compared to actually walking people through use of web3 tools.
Some examples of this are: These Garifuna teenagers were onboarded to wallets, layer2, and even Gitcoin passport. It legit took 5 hours to just get the wallets installed and teach them how to make basic transactions.
Day 1: created wallets, dejargonized blockchain, L2s, DAOs, and practiced receiving/ making payments with DAI on @arbitrum. @MakerDAO_Lat @MakerDAO. Between slow wifi and old phones, this took 5 hours of hands on practice. pic.twitter.com/O7i0YO1xPk
— web3beach.eth (@web3beach) December 16, 2022
That was just wallet creation. We then moved onto creating Gitcoin Passports. That took another 5 hours!
Day 2: created @gitcoinpassport on mobile devices. It's way more complicated than doing it on a laptop and only 8 participants were able to verify at least 1 stamp. Google and Facebook are the only 2 they are eligible for. This took another 5 hours of fiddling around. pic.twitter.com/Gh2OtuXv5P
— web3beach.eth (@web3beach) December 16, 2022
An important question to ask here is - Why would they need a Gitcoin Passport? At first glance, it might look like I would use this as an opportunity to send them funds, then have the new web3 users contribute to my grant. But, no. I do this with two goals in mind: Teach Decentralized ID because some of them don’t have a national ID, which is common in rural areas. The other goal is to onboard these communities and give them the skills necessary to use web3 to regenerate their communities and apply for their own grant funding and not live under a web3beach umbrella. This would be more feasible instead of trying to get them to attain a high enough passport score to enable matching.
Only eight of these participants were able to activate stamps and the only two they were eligible for were facebook and google. These two stamps alone were a headache because they sometimes had to reset their password because they don’t use email for more than activating social media accounts. I would have had to spend at least another week in this village and a decent amount of eth trying to enable web3 stamps like NFT, ENS, spend/ hold enough eth, create a linkedin, discord or twitter. I have failed at enabling these communities to participate and vote for their own grants because it’s incredibly difficult and expensive.
This is a good time to point out that I explained the importance of Gitcoin Passport and made it clear that even if they were to have enabled matching, they were not supposed to contribute to my grant because that would be a sybil attack - part of the DID conversation.
We start the day creating web3 wallets and teaching new regens how to protect them, enable layer2s, and connect to Dapps with the browser or @WalletConnect. This is where they will receive a #proofofimpact NFT pic.twitter.com/AVe8BPTiGj
— web3beach.eth (@web3beach) November 29, 2022
Let’s move on to college educated and tech savvy noobs from N. America and Europe. They were able to install the wallets much faster because they have modern devices and are accustomed to constantly toying with current apps. Onboarding almost 50 people at once took a little over an hour. I would not have spent another few hours trying to get them to enable Gitcoin Passport because I don’t think Decentralized ID matters to them, they don’t have projects that could apply for grant funding - the only reasons I go through the trouble of having Gitcoin Passports created. Even if we did try, I highly doubt they would have reached a passport score that enables matching.
A middle of the road example would be onboarding NGOs in Colombia and subsidizing all the expenses necessary for them to enable Gitcoin Passport and apply for grant funding. This took three days to accomplish. As difficult and expensive as it was, we were able to accomplish the goal of empowering these communities. Teaching them to fish, not just giving them a fish.
Day1 (March 30th)
Yesterday we onboarded 4 environmental NGOs in Cartagena. They'll be onboarding all of their staff and volunteers to web3. We created wallets, enabled Optimism, enabled the DAI contract, discussed public goods, impact certificates, L2s, p2p, made and received payments + more. pic.twitter.com/ebca5JR7Zz
— web3beach.eth (@web3beach) March 31, 2023
Day 2 (April 4th)
We've onboarded 4 #Colombia NGOs with a proven track record of social and environmental impact. They were onboarded to @optimismFND learned self custody, decentralization, L2s, sending + receiving funds, ReFi and more. Today we onboard 2 more NGOs. pic.twitter.com/Yhs50rFX93
— web3beach.eth (@web3beach) April 4, 2023
Day 3 (April 5th)
Continuing #ImpactOnboarding in #Colombia. NGOs now have an @ensdomains, created projects on @gitcoin + @Giveth , and a created @gitcoinpassport. Next tools? @poapxyz and @hypercerts intro. Now they go and onboard their staff, volunteers, friends, and families with our support. pic.twitter.com/bmz3nCbh6k
— web3beach.eth (@web3beach) April 4, 2023
My reason for sharing this isn’t to discredit anyone else’s claims. If you’re working to empower the global south and indigenous communities, I respect that. One big thing any of these projects could have done better is document and share their work, it’s much easier than the actual onboarding process.
I’ve been onboarding people from all over the world for almost 2 years and think I have done a good job at making the process more efficient, but I would be unable to accomplish what some of these grants have been able to. I’m open to learning from their hands-on experience in hopes of making my processes better.
Long story short. Documenting your onboarding efforts and sharing it is much easier than actually onboarding.
3 Likes
Hi all.
Not sure my input will hold any weight, but I do know Mycelia co-founders (Tay & Mathilde). In fact they’ve welcomed me onto there team and an indigenous gathering recently.
We’ve just finished attending this gathering where Mycelia was invited to act as videographers and digital bridger since Mathilde has extensive experience filming indigenous cultures, while living amoung the Indigenous Peoples of Brazil for 7 years.
I’ve found the skills of Tay & Mathilde to be complimentary and have no doubt they have the tech skills to lead this project for the long haul. But it’s the heart wisdom that allows them to work with Indigenous Peoples. That requires a different value system. One that puts everything else ahead of money accumulation for personal gain.
I’m honored to be with them, and will continue to be, as long as I remain useful. I’d stake my reputation on them, and can’t see any evidence of tendancies to game the system for some advantage. It’s not in their DNA.
You can hear them sharing about Mycelia on MAX IMPACT EP. 306 just recorded this morning ; https://twitter.com/i/spaces/1djGXlovqoyGZ
1 Like