EtherScore: A Decentralized Reputation System

Oracles and Data
1

TLDR

  • Reputation systems in blockchains are missing (applications have difficulties identifying and rewarding their users)
  • Projects are tackling these issues mostly by building their own centralized methods which lack transparency, composability, and interoperability
  • EtherScore provides a way for Dapps and users to build on-chain reputations based on web3 achievements represented through badges / Soulbound Tokens (SBTs)
  • EtherScore badges can later be used by any Dapp to distribute any kind of incentives such as airdrops, discounts, DAO voting power, etc.
  • Many new use cases are yet to be explored, such as reputation-based Dapps, DAO paradigms based on contributions instead of wealth, etc.

Background

  • SBTs / Soulbound tokens: Non-transferable NFTs, NFTs tied to a specific person/address. Can be used to represent identity, diplomas or achievements (paper from Ohlhaver, Buterin and Weyl, Discussion post on SCRF from James Mc Girk)
  • Oracle: Oracles are smart contract structures that deliver off-chain information to on-chain systems. (definition from SCRF Glossary)
  • Data indexer: In the context of blockchains, a data indexer is collecting and aggregating data from the blockchain to enable querying this processed data easily instead of having to analyze and parse the blockchain (can be used for example to get the list of transactions performed by an address, or the list of interactions with a specific Dapp) (more details in The Graph protocol documentation)
  • Sybil resistance / Sybil protection: Protection against an attack where a node in a network illegitimately claims multiple identities.(definition from SCRF Glossary, other definition from Gitcoin here)

Problem

Despite blockchain’s focus on decentralization and equality, we have seen numerous projects trying to acknowledge and reward their users’ contributions in recent years (or new projects trying to target specific audiences of users to avoid bad actors). In the absence of a native reputation system in web3, projects are developing their own siloed methods by analyzing user behavior using off-chain analytics. There is a lack of transparency in these algorithms due to the fact that most projects don’t share and reuse each other’s work, which results in a very low level of composability and interoperability for these reputation models. Users will have to build reputation over and over and over again inside each silo. This is anathema to web3.

In addition, projects are spending extensively for user acquisition/fidelity programs but those investments often lead to capital inefficiencies:

  • Airdrops cheated by bots and insider trading (or distributed to users dumping their tokens directly) (Ribbon finance, analysis on UNI airdrop results)
  • DeFi overcollateralization (most loans in DeFi must have a collateral loan-to-value ratio > 100%)
  • DeFi protocols competing with each other by distributing token incentives to attract liquidity

We can also identify other opportunities:

  • DAOs are based on traditional finance paradigms and equity/shareholders with most protocols using the 1-token-1-vote model (the richest have more voting power; which is not in accord with the original ethos of blockchains). Indeed in some models, a core project contributor having small capital will often see his votes being insignificant (compared to large capital contributors) even if he was the single most involved user since the early days of the project.
  • No cross-Dapp user profiles (‘login with Ethereum’ used to connect to a lot of applications but very little data is extracted from it due to to the need of data processing through off-chain methods)

Solution

A reputation is traditionally defined and built based on an individual’s actions. By aggregating actions performed on-chain by addresses, the EtherScore protocol works to replicate this concept in web3.

Certifications and badges were always used to refer to each other and bring confidence between humans (from academics to soldiers, artists and video games). In this regard, the format of on-chain badges easily makes sense for a decentralized reputation system.

EtherScore provides a way for Dapps to easily identify and reward their reputable users over time. By deploying NFT badges, applications can develop fidelity programs and provide benefits for contributors without relying on off-chain analytics platforms. A fully on-chain fidelity program can be built by integrating user analytics checks directly inside smart contracts.

EtherScore badge factory is the system used by Dapps to customize and deploy NFT badges. This factory enables developers to customize badge esthetics but also parameters such as eligibility conditions, transferability, burnability etc. Eligibility conditions are verified by a Chainlink oracle using data sources such as The Graph protocol, Snapshot API etc. Such mechanisms are used to try to stay as decentralized as possible (but it also limits the complexity of conditions in terms of data processing). We are not building our own users’ analytics or processing AI algorithms on data, as we do not want to position EtherScore as a third-party (we are just a tool provider). EtherScore badges are only based on objective data and conditions and results are written on-chain, which makes them verifiable even if the project should become no longer maintained.

Existing alternatives on the market are usually not as decentralized, which makes Dapps dependent on them as they become a required trusted party in the process.

We can consider EtherScore as an evolution of the POAP badges used to collect proof of attendance at events/conferences (POAPs are transferable, tradable, and can also be obtained without participating in the event as the only condition to collect them is to scan a QR code).

Discussions

This post briefly presents the concepts of EtherScore, a decentralized reputation system. While any questions are of course welcome, we would like to engage the discussion on the topic of which kind of new ‘reputation-based services’ you can think of. Numerous use cases are present in DeFi and DAOs, and even GameFi, but we are also thinking about other areas such as DeSci and talents/invoicing (areas in which SCRF users’ feedback would be useful based on experiences and the SourceCred experiment).

Another set of questions evolve around what kind of DAO governance could be developed through hybrid solutions mixing SBTs in addition to ERC20 tokens (EtherScore is a funded project aiming to issue its own tokens which would be used to pay for badges and to be staked to receive a share of protocol fees). Our alpha version is currently experimenting with a first step for a contribution-based DAO governance in which one badge is equal to one vote, meaning that the vote weighting of an address is proportional to the number of badges held (more details, EtherScore Snapshot proposal using 1 badge = 1 vote). Later versions will include customizable ponderations called ‘score’ for each badge to highlight specific achievements. For the moment, we are thinking about vote weights by multiplying the number of tokens and badges ‘score’.

A third question would be what kind of behavior can be considered as ‘positive’ in the ecosystem. We have proposed a selection of badges focused on the aspect of “contributions to the ecosystem” with badges claimable for Gitcoin donors, content creators on Mirror or Lens, people active in DAOs on Snapshot, but also we tried to work on the aspect of Sybil resistance by providing badges for people verified on Proof of Humanity (we may integrate BrightId or directly the recent Gitcoin’s passport).

Further Reading

10 Likes
2

Glad to see your project is making such progress! And thank you for writing about it on the forum.

I’m increasingly interested in identity and reputation on Web3 and it’s great to see a project looking to manage this in a decentralized manner.

Some of my questions would be similar to those that are coming up in the Spectral Finance post from a few months ago regarding the management/verification of one’s reputation. If a reputation is connected to a wallet, let’s say, it would be possible to create multiple reputations. I believe that this question was also raised in a chat somewhere and part of the discussion was about whether or not that is an actual problem that needs to be solved.

I’m also interested in what would constitute good behavior, but maybe the discussion is best served exploring one of your questions at a time!

5 Likes
3

I have similar questions to Paul. I have so many organizations and data bases my life spans across having one wallet that spans them all or service to support them all seems untenable. Perhaps etherscore would work best on one particular ecosystem or service space of my life and then connect with others (I know you mentioned avoiding oracles and I posted additional asks in your discord!)

4

It is essential to implement a strong and centralized reputation system to simplify and streamline the user experience within the Cryptosphere. Such a system would increase transparency by allowing users to build a single reputation across multiple projects and platforms. It would also promote composability, interoperability, and trust between users and projects. Establishing an inclusive global standard for user trustworthiness would benefit the entire Cryptosphere by providing enhanced security and friction-free interaction.

I developed the following reputation-based services in response to your query, which I deem to be of paramount importance:

  1. Reputation Trading Marketplace
  2. Trust-Based Networking Platforms
  3. Reputation Insurance Services
  4. Verified Online Currencies
  5. Pre-sale Rating System
  6. Reputation Hedging Platforms

I developed a comprehensive list of ten concepts for DAO’s governance for an episode I reviewed of “SCRF Interviews” some time ago. The following are two of these concepts:

  1. Dual Issuance of Token Systems: For instance, Security-Backed Tokens (SBTs) for long-term incentives and Ethereum Request for Comment (ERC20) for short-term liquidity.
  2. Split Issue Governance: This allows token holders to jointly exercise control over a particular decision with the inclusion of certain stipulations.

There are numerous beneficial activities that can be undertaken, including:

  1. Contributing code to open-source projects
  2. Engaging the community through sponsored discussions on forums such as SCRF
  3. Writing informative blog posts and articles on related topics
  4. Creating stories, bug reports, and feature requests for internal or external dapps

I trust you find this information useful.

1 Like