The Future of Reputation Score Will Be Community Specific. What Trade-Offs Will Communities Face To Quantify Reputation Better?
Eliot Couvat
0xC024
March 15th, 2022

Hey there! This essay is a collaboration between Coinvise, a web3 platform that helps creators & communities build and manage their Social Tokens, and Kerman Kohli, founder at ARCx, a decentralized scoring protocol that powers on-chain identity and credit-score based borrowing. We collaborated to bring our expertise together and write an easy-to-read essay with the trade-offs of the current technology to quantify social reputation. Enjoy! 🔥


In the physical world, we can use our driver's license as a way to prove our identity, but it doesn't really tell us much about a person or its reputation. To get to know a person in the physical world and see what they have done in the past, you might ask shared acquaintances or have a conversation about their favorite things. But the process is time-consuming and inefficient to know someone deeply.

Knowing someone and seeing clearly their social reputation is easier on the Web. Looking at Facebook and searching on google takes only a few minutes, and you can get a good sense of someone's reputation by looking at their photos, ideas, articles they shared, etc. The problems with the current Web is that first, everyone struggles to keep control over their online reputation, and second, it's hard to prove what's true and what's not.

And that's when the decentralized Web (aka Web3) comes in. With (most) of our data now on-chain, meaning open, easily accessible, interoperable, and can be used across many different types of decentralized applications, getting to know a person and seeing Social reputation has never been that easy.

Indeed, our online identity and social reputation are more than just the core vital stats (name, serial number, rank), they are constantly evolving, and there are many sides to our identity, which we employ depending on the context (at work, on Twitter, with friends, etc.) The more work we do to put these pieces of information on-chain, the closer we are getting to quantifying a human's worth. Leveraging Web3 technologies to well-assessed Social reputation will enable people to see each other's profiles (competency, worth, status, etc.) from the lens of verification instead of the lens of trust & unlock new levels of interactions.

However, while we're slowly getting more and more of our data on-chain, we still lack an efficient system to assess them well and efficiently quantify our Social reputation. Different methods exist today, but no solution is perfect, and they all come with trade-offs.

Solution 1 - Quantifying Reputation via Social Tokens

Social tokens are virtual currencies that live on the blockchain. They are earned by accomplishing specific missions and can be redeemed in exchange for other cryptocurrencies or special perks within the community the token is associated with, such as access to token-gated content or the right to vote on future strategic decisions.

These virtual currencies are essential in making Decentralized Autonomous Organizations (DAOs) run efficiently, as they serve as salaries and allow to create trust better, align incentives and retain contributors.

In the context of work, by serving as a virtual currency in DAO (that we could compare to Web3 startups), social tokens represent a great way to measure skin-in-the-game and involvement in a community and are a great way to quantify social reputation. They offer a lot of granularity in how we quantify and measure experience, reputation, and achievements online.

The more tokens someone has, the more they have participated and helped a DAO grow. Some platforms, such as Coinvise, have already implemented such features focusing on social tokens to better reflect the quality of work and implications in communities.

Social tokens could serve as verifiable credentials when applying to DAOs. These decentralized communities would, in exchange, have the capabilities to assess the candidate value upfront and make them work in projects accordingly.

Social Tokens offer a great system to discover, analyze and use data that a user has created on another application, platform, network, or DAO.

However, there are a few issues with social tokens that still need to be addressed:

  • Anyone can buy tokens - Today, one can buy tokens and increase their reputation and voting powers, which is unfair for those who worked hard to earn tokens. We would need a system to differentiate the 'Purchased Tokens' and the 'Earned Tokens.'
  • Web3 Wallets don't optimize for Social Tokens - With current wallets, users cannot easily show their tokens to others. Web3 wallets should optimize for social tokens by creating a dedicated space for them and allowing users to manage them easily (hide, showcase, etc.)
  • Social Tokens are still in the pilot phase of development - As of today, there is no real network effect, and we'll need more DAO to adopt this way of quantifying social reputation to see social tokens thrive in this context.
  • On-Chain makes things expensive - With network members having to pay Gas fees to send social tokens, it's today expensive to get a social reputation only based on social tokens (note that Layer 2 blockchains are coming quickly to solve this problem).

In short, Social Tokens have a great potential to measure social reputation and are already used today to get a sense of community involvement. However, they are still too early to be widely used, and major problems still need to be fixed.

Solution 2 - Quantifying Reputation via NFTs

Non-fungible tokens (NFTs) are tokens that (often) provide a way to access a community. They come with status, scarcity, and belonging within this community. Collections of NFTs represent membership in different communities and provide special perks in these communities.

They're a great way to quantify reputation as, thanks to the hype around NFTs in 2020, they already have a massive network effect. It's today one of the Web3 techs the most widely adopted, with people already creating their online identity around their NFTs (e.g. @punk6529 or @BAYC2745), making NFTs easy to adopt, use and integrate. Furthermore, dozens of projects build Web 3 social/identity tools for NFTs. This tech's network effect is the strongest factor in favor of using NFTs as a basis for identity.

NFTs are also a great way to showcase which communities one belongs to, with Web3 wallets' interfaces optimized for showcasing them. Buy a rare NFT and increase your Social Reputation right away. It's that easy.

Finally, as NFTs don't need liquidity to exist and have value, there's no cost for upkeep liquidity, making NFTs a promising and cost-efficient way to quantify social reputation.

But several problems persist for NFTs to become efficient at well quantifying reputation:

  • Can only be issued once at a particular point in time. Serve more as a "badge" than a live-view of someone's reputation.
  • Transferrable and represent wealth and status more than reputation (e.g. Crypto Punks, BAYC etc..)
  • Good for communities but not really focused on an individual (FWB, WHALE, etc)
  • They are, by default, public by virtue of being on public blockchains. However, since they are chain specific, the reputation is siloed due to to the lack of cross-chain support for NFTs

Solution 3 - Quantifying Reputation via Scores

As we all know, the famous saying, "it takes years to build a reputation and a day to ruin it." The nature of reputation implies that it exists as a spectrum rather than a binary yes or no answer. Not only is it a spectrum, but it also changes based on the person who perceives that reputation. For example, how you are with your friends could be different from how you are with your family or co-workers.

Scores are a new perspective of quantifying on-chain reputation. They can be viewed as context-specific numbers that continually update and assess an address based on their on-chain and off-chain behavior (provided they're tied to the same address). This mimics how reputation works in the real world by enabling a wider spectrum of expression about how someone is "trusted" and the sources. Through combining on-chain data, snapshot voting data, and/or DAO contribution metrics, a score can be as narrow or as broad as the score creator chooses to be. Examples of scores could include:

  • Token specific HODL score that measures how long you've held a token
  • Airdrop specific Airdrop score that measures how much of your airdrop you've retained relative to other participants
  • Protocol specific Credit score that measures your creditworthiness to offer dynamic load conditions

Relative to other forms of reputation, scores are still very new and have lots to be discovered about them. One of the biggest concerns of scores is that they become dystopic in nature, similar to China's social scoring system. The crucial difference here is that behavior is only judged at a wallet level and the user can completely start all over again from fresh if they don't want their past actions to follow them.

The other major component of scores is that they're purely computed off-chain and uploaded on-chain – given the large amount of data sets to process and traverse through. This means they can allow reputation on one chain to be available on all other chains. The downside is that there is implicit trust in the off-chain computing engine and ensuring the scores are published with integrity back on-chain. Platforms such as ARCx are working on creating scores that can power reputation-enabled web3 apps.

Closing Thoughts

The world of on-chain reputation and identity is the next major frontier of crypto that enables a whole new class of applications. We're currently at a point where we're all trying to understand how the various architectural approaches can accomplish the end goal of creating real end-user value. As we stand today, the current outstanding trade-offs come down to the following dimensions:

  1. Does the reputation element need to be dynamic or static?
  2. Should it be all on-chain, or can certain elements be off-chain?
  3. Is non-transferability a hard requirement, or should it be optional?
  4. Do they need to be tied to a real-world identity (social tokens), or is a pseudonymous identity okay?
  5. How on-chain should the reputation element be? Should it be uploaded as an NFT, an oracle, or even a Merkle root that keeps updating?

As you can see from the above, these are all fundamental questions to answering how we quantify reputation, but we're revealing the answers to all these questions as more products get launched in the market.


Like this essay?

Make sure to follow Coinvise and Kerman Kohli on Twitter. You can also join Coinvise’s Discord to continue the discussion about Reputation Systems.

Thanks for being here 🔥

Arweave TX
aNocDVtVWOJ56XptSUJSHS3zm9I1E0lFz9bIqMAtjFw
Ethereum Address
0xC0241e384dB48C35E7b3FEded3BfEcd29A3E73e3
Content Digest
ibaLVi2tW4B9iAlPmwUBbCZYu2CvKl7X0NOg29eEMg4