Building SourceCred for the Wider World (Pt. 2: Get to your god damn point!)

TL;DR

NOPE :smiley:

Purpose

In my background post I did my best to express why I’m a big advocate of great soft cred mechanisms/cultural practices, with hard cred / plugins as axes for growth.

Here, I want to explore how we can add a fork of props / did-a-thing to allow more communities (including ourselves) to:

  1. More deeply embed their values in SourceCred, making SourceCred accessible to a wider range of communities.
  2. Keep their values from being diluted amidst growth while still allowing for evolution and decentralization, allowing them to be more inclusive without compromising the essence of what makes their community special.
  3. More consistently and precisely measure value.

Disclaimer // Motivations

I like SourceCred’s community and value it a lot. It has been a wonderful breath of air from the more bro-y feel of much of the crypto space. Part of the motivation comes from wanting to sustain this culture amidst inevitable growth/participation. I’m not sure if this is right or not, but I thought I should put it out there.

What to Expect

In the last post, I explained how in our current props / did-a-thing system, two contributors can and likely will receive different cred scores depending on who they are (not to mention when the post, how they phrase it, and a whole host of other factors). Solving this property is the essence of this post, so let’s start there. After that, I’ll explain some major benefits of the approach and look at them in a few examples.

Holding Contribution Consistent — Who dunnit? Don’t know, don’t care!

If we want Alice and Bob to receive equal cred for the same contribution, we need some rough consensus about how much a certain kind of contribution is worth. We have this for our plugins, but not for “soft cred”. In the plugins, a post is a post, a reaction is a reaction, a commit it a commit. In other words, our plugins have some accounting of what contributions are worth in relation to each other. We also have the ability to change these configurations to align with the state of the community and what we value. This holds contributions consistent between contributors. Brilliant.

So how could we replicate this for “off plugin” contributions? Rather than leaning into complex cryptoeconomic voting mechanisms and that jazz, I suggest the simpler approach of establishing some rough consensus about what’s worth what. This holds contributions consistent between contributors.

Inspiration

The Token Engineering community uses a “praise” system, similar to our props. However, one key difference is that they weigh contributions based on a rough tier system. Certain contributions earn more “impact hours”. These might correspond to the intensiveness of a contribution or how valuable it is. Doesn’t really matter. The point is to get a general, very broad strokes sense of a contribution’s worth. It’s more of a way of making sure that contributions don’t fall out of some fair range than trying to account for things perfectly. Cool.

The issue they have is that they have to review these manually, which besides being a headache is centralized and not scalable. But these are exactly the issues that SourceCred is in position to solve.

We already have a “review process” in did-a-thing and props where the community will react and thus flow cred to posts. In the current case, reactions = varying degrees of cred. In this model, we can use different emoji’s to signal a certain tier of contribution. Rather than scaling cred for more reactions, we can scale the trust that the contribution fits a certain tier.

Let’s see this in a few examples. But first, remember that the point isn’t to track every last thing, but to cover the foundations that the plugins might miss.

Case Studies

Community Case: Dog Whisperers of America

The DWA has decided to use SourceCred to track and reward contributions to this much neglected field. They know exactly what they want to reward people for and roughly how much. Inventing a new trick, obviously a lot of cred. Volunteering for a conference, great. However, they’re frustrated as they don’t have a clear way of tracking these alongside their normal Discord and Discourse plugins. They like the plugins they use, but the DWA is very clear about their values and want a better way to reflect that. They are tough customers.

By providing them an easy way to account for the value of contributions, they can more closely align their SourceCred instance with their values.

This is the tip of the iceberg, but you can begin to see how simple consensus around how valuable some basic non-plugin contributions are would help a much broader range of communities derive value from SourceCred.

Community Case: Potentially SourceCred

As more people join the party, we’re going to face a lot of noise entering the project. By noise, I mean that larger percentage of props and did-a-thing reactions, messages, and thus cred will come from less “core” contributors, many of which will not have spent the time to become informed, engaged, or value-aligned.

By establishing some rough consensus about non-plugin, soft-cred contributions we can better position ourselves to gain from inclusiveness.

Individual Case: Casper

Casper does great work for their community. However, as an introvert, Casper doesn’t engage much with others. While they’d love to hear from Casper more, it’s just not gonna happen. Bonnie works alongside Casper and realizes how much impact they’re having. When Casper occasionally posts on did-a-thing, Bonnie is always there to flow cred to Casper. Bonnie will often dish props to Casper, but since most people aren’t familiar with what Casper’s doing, not much reacting happens.

In this model, the community has been more specific about the value of Casper’s work. When Bonnie spotlights Casper, it’s not a question of who else will flow cred. The contribution clearly fits in a tier.

Individual Case: Alice

In the last post, I highlighted how Alice often found herself being rewarded less for the same types of contributions as Bob. Here, we can see that this problem gets handled up front. Because we have some rough consensus about that contribution, this can’t happen. The tiers of contribution are broad enough such that there’s no room to argue about what tier something fits in.

Bonus Points

  • One emergent property that I really like about this is that this allows anyone to emoji the shit out of other peoples’ stuff without having to think about if cred is flowing or not. This feels more natural and would allow a more natural onboarding to using SourceCred without any strife around scarcity fears.
  • It’s easier to give trust out then to get it back. It’s easier to establish consensus around what’s valuable when the circle of trust is tight, and then include people liberally in a more robust system.

From here

I’m going to tinker with building this out, and test how it performs/feels. It’s certainly worth a try, so I’m looking to for feedback to include as this develops. If you prefer to chat about it, hit me up on Discord.