The vision for Ethos is to provide a single, objective, and transparent credibility score for anyone in web3 to understand; a quantifiable number that you can glance at and say “I can trust this person” or “I should be cautious with them.”

Similar to a credit score, this credibility score is calculated from a variety of sources, including:

• onchain history and activity

• reviews written for you, and the credibility scores of the people who left those reviews

• vouches (Ethereum staked) to back you, and the credibility scores of the people who left those reviews

• slashes in you, if you happen to step out of line of community consensus

• who invited you to Ethos, and their credibility score

🟡 Please note: Anything said here is subject to change, as we expect to learn and make adjustments to the first iteration of the Ethos Credibility Score.

What’s the scale?

In the current pre-access phase of Ethos, we’re imagining a scale system that is similar to ELO ratings. ELO was created for chess and is primarily derived from competitive results; it is used to predict the likelihood of who wins when two people play each other.

While Ethos participants won’t be competitively dueling, when a user with a high credibility score reviews someone with a low credibility score, it should have a greater impact than two users with similar scores.

At present, Ethos scores are split into five levels:

0-800 - Untrusted800-1200 - Questionable1200-1600 - Neutral1600-2000 - Reputable2000+ - Exemplary

We expect to see a distribution resembling a bell curve, where:

• New wallets are likely to start in the Neutral range

• A smaller percentage of users land in Untrusted/Exemplary than Questionable/Neutral/Reputable

How do we determine an initial credibility score?

Credibility of web3 participants is incredibly challenging to measure today. Not only is it difficult to measure, but it’s cheap to spoof. How does Ethos give an accurate initial “credibility” score for users? While we are still working on this, we expect the vast majority of users to land in neutral credibility.

The strongest signals of how to build or lose credibility will be the mechanics that Ethos offers (reviewing, vouching and slashing), and these will be the most important drivers of credibility scores changing.

Here’s how we’re currently thinking about additional metrics that will influence the initial credibility score

Positive Influences:

• Your wallet has “significance”

  • history, participated in specific events, claimed specific rewards, etc.

• Your attested accounts (Discord, Twitter, GitHub, etc.) have “significance”

  • non-zero number of followers, git commit history,

• The person who invited you to Ethos has a high credibility score

Negative Influences:

• Your wallet and connected accounts have 0 significance

• The person who invited you to Ethos has a low credibility score

Each of these items is very purchasable; you can go buy a Twitter account with 10k followers and 10 years of history for a few dollars. This is why we intend to minimize their initial impact on your credibility score and start most people at zero, while maximizing the one clear signal we do have - who invited you?

Ethos' early access will start out as an exclusive, invite only product -  we intend for invites to be very limited for early access. Similar to someone vouching for you with capital, an invitation acts as a signal to the network that “this person can be trusted.” We believe that receiving an invite from someone who is already highly credible will serve as the best signal for your initial credibility score.

How does the credibility score change over time?

Participation in Ethos using our mechanisms is how your credibility score will change over time. These mechanisms are peer-to-peer signals to the network of who can be trusted and who can’t, with different orders of magnitude.

Reviews

The most simple form of Ethos mechanism, a review is negative, neutral or positive and has a minor impact to a user’s credibility score. Reviews can range from simple interactions, like an over-the-counter deal with someone, to more severe situations, such as when someone performs a rug-pull on their project.

Since reviews don’t have a strong connection to financial upside or downside, they reflect only minor change to a user’s credibility score - but can still be built up over time.

Vouches

This is the most clear way to signal to the market that you believe in someone- would you back them with capital? Vouches are only used positively, and can have a major positive impact to a user’s credibility score.

Let’s say you have a friend who has just joined the space and is looking to create a new web3 project. You know them personally, but the rest of the space doesn’t yet. As a reputable person on Ethos, you can vouch for your friend to help signal that they can be trusted. Vouches are a higher signal to the network because they require you to put capital against someone’s name, and thus your friend can be backed in a more meaningful way.

A user’s credibility score earned from vouches is likely to be earned over time, meaning not all credibility score is given up front. We’re still experimenting with and learning how vouches should impact credibility scores!

Slashes

Similar to vouches, slashes represent a more significant instrument than just an ordinary negative review. Slashing, commonly used in proof-of-stake systems, enables Ethos participants to initiate a vote on if another participant has stepped out of line. If consensus agrees with that vote, a slash will not only incur a financial penalty, but also have a major negative impact on the user’s credibility score.

Slashing is not yet implemented in our pre-release, as we are still working out how slashing will impact the credibility score. If you vote in a slashing event the same way as the outcome, should that count towards your credibility? Should being a successful slasher increase your credibility? We’re still learning and figuring this out!

Other Mechanisms

While these are the current primary mechanisms from which we expect credibility scores to derive, we have imagined other potential future mechanisms. We’ve been thinking about things like “promises” which would enable people to make a public promise, and Ethos users determine if that promise was fulfilled.

Transparency & Governance

The credibility score will be calculated off-chain to start, thought it will be 100% transparent and observable in how it’s calculated. Considering changes will need to take place as we learn and adjust the algorithm, it’s important to keep in mind that it will take revisions to reach a true baseline.

While the Ethos team will define the initial credibility score, our vision is that it will be calculated on-chain, governed by its participants, and adapt to how the community’s opinion about credibility changes. This requires a journey to get to that point, and we intend to introduce these concepts over time. We have already imagined ways to introduce light-weight governance on how credibility score weighting can be determined by those who participate.

Other Credibility Scores

We’ve created these mechanisms in reviewing, vouching and slashing with the intent of anyone being able to observe and create their own formulas for how credibility is calculated.

Just like morals and ethics differ between cultures, we expect people to want to measure credibility differently for different use cases. The Ethos credibility score is intended to be a generic source of truth for who can be trusted, but we’d love to find ways to extend the primitives of the platform to other scores too.

If you’re interested in chatting more about creating your own credibility score, please reach out to us directly on Twitter.

You can also join the waitlist for Ethos here: waitlist.ethos.network