Reputation-based Weighted Voting on OP Mainnet (TE Academy)

[akrtws]

Any DAO’s growth depends on good decisions. At TE Academy, we’re exploring a new primitive for token-based decision-making: Reputation-based Weighted Voting – we will design a voting mechanism that makes reputation count, and run a voting experiment on the OP Mainnet.

Join us! Take part in the TE Academy Study Season - a free cohort-based education program starting on April 25.

Why Reputation-based Weighted Voting (RWV)?

The design space of token-based governance is huge. In DAO governance’s reality though, only a handful of primitives achieved significant adoption. 1token1vote links voting power directly to the number of tokens held. Vote delegation allows token holders to assign their voting rights to another party, enabling concentrated decision-making power based on trust.

There is a third class of voting mechanisms to complement these concepts: make a voter’s track record count in DAO decisions, and use proofs of expertise and achievements as a signal to define voting weight and decision-making power. In this experiment, TE Academy explores the potential of Reputation-based Weighted Voting in an educational project. Students will have the chance to go over the entire token engineering process and work on a case study with a real voting outcome - deciding on the winner of the first TE Academy fellowship with a $10K prize for the fellowship winner.

Over the past 3 years, TE Academy has established a system of NFT proofs to track individual community members’ achievements and the development of the sector overall. We’ve issued more than 1000 NFTs to students who’ve passed knowledge requirements and to researchers with significant contributions to the token engineering discipline. In our experiment, any community member will be eligible to vote - however, holding Token Engineering NFTs will increase a voter’s weight in this decision. No popularity contest!

We’ll account for:

• passed exams in the TE Fundamentals course (more than 4000 students enrolled)
• sharing knowledge as a speaker at TE Academy events (more than 120 over the last 3 years)
• contributions to our shared body of knowledge in the form of online learning materials (first bachelor-level course in token engineering)
• supporting students as study group host (in one of 45 study groups)
• and more.

The Benefits

Reputation-based Weighted Voting can be applied to any DAO decision-making cases where expertise is required to make good decisions. In RWV systems, voters are encouraged to sharpen their profile and become highly sought-after experts in their domain of decisions (e.g. Retro-Public Goods Funding, Risk Management). For a DAO’s collective, it’s an opportunity to establish incentives for decision-makers and grant voting power to those who were actively involved in good decisions.
Combined with 1token1vote and vote delegation, it’s a new building block to make decisions transparent, and robust against collusion, bribery, and economic attacks.

We offer this program as part of our public goods education, enabled by Optimism RetroPGF! The results of this experiment, as well as the voting algorithm and a simulation engine, will be available open-source.

Learn more about our program in this presentation at the Optimism Demo Day!

Today, more than 4000 students are enrolled in TE Academy’s studying programs, and research initiatives. Currently, we are exploring token-based DAO decision-making via RWV, and AI-Copilots for RetroPGF, our second initiative in collaboration with the Optimism Collective.

Register for the Study Season, and learn how to design, verify, and implement Reputation-based Weighted Voting.

See you at TE Academy!

[akrtws]

Update:

Yey , we’re getting ready for the first session working on Reputation-based Weighted Voting at TE Academy!

Remember, it’s part of the Study Season, anyone interested is invited to take part and work with us on this voting experiment on OP Mainnet!

Here’s our reading list to prepare:

Learning List on Voting and Social Choice

General Voting Theory
Which Voting System Is The Best?
The Mathematical Danger of Democratic Voting
Arrow’s Impossibility Theorem
The Flaw in Every Voting System

Readings
Nice Introductory Overview of Voting System Challenges
Mathematical Overview of Social Choice

Try These Things On Your Computer
Install an IDE. VSCode is a good choice, especially for beginners. Here is a walkthrough on how to get it set up. If you already have an opinion on vim/emacs war, you’re well ahead of what we will be discussing.

Our first session will take place on Wednesday, May 01, 12:00 - 14:00 UTC.
To take part, register here!

[akrtws]

Week 01 Update

Here’s an update sharing our progress in “Reputation-based Weighted Voting” at TE Academy. This project is part of our cohort-based Study Season - an education program enabled by Optimism RetroPGF!

The first week is dedicated to exploring voting theory and social choice. There is no perfect voting mechanism! → Arrow’s Impossibility Theorem
Our goal is to design, verify, test and implement a voting system to cater TE Academy’s own needs by End of June.

We discussed:

• Voting Mechanisms examples, their benefits and vulnerabilities
• Attack vectors in crypto (Sybil attacks, bribing, etc.)
• Goals & constraints (TE Academy’s Fellowship voting, NFT system, proof-of-knowledge, proof-of-contribution)
• Requirements & Timelines

Follow our work:
Session recordings: Track 4/1 session / Track 4/1a session
Slides: Track 4/1 Slides
NFT Information: TE Academy NFTs on Optimism (Otterspace Badges)
NFT Metadata, holders, etc.: Subgraph Queries

Some screenshots below

Exploring the NFT infrastructure to prove reputation at TE Academy

Techniques for collecting and structuring voting design requirements

Querying data about NFT types and number of holders

[FractalVisions]

@akrtws

Over the past 3 years, TE Academy has established a system of NFT proofs to track individual community members’ achievements and the development of the sector overall. We’ve issued more than 1000 NFTs to students who’ve passed knowledge requirements and to researchers with significant contributions to the token engineering discipline. In our experiment, any community member will be eligible to vote - however, holding Token Engineering NFTs will increase a voter’s weight in this decision. No popularity contest!

This is great. Do you have a link to the NFT collection so we can see it ?

[akrtws]

Sure @FractalVisions !

Here’s the link:

Otterspace: BADGES Token | Address 0x7f9279b24d1c36fa3e517041fdb4e8788dc63d25...

Note that numerous communities use this contract to mint NFTs (Otterspace Badges).

To find the TE Academy NFTs, you’ll have to query the subgraph:

Badges Optimism Subgraph

  raft(id:"rafts:29"){
    id,
    specs{
      id
      metadata {
        name
      }
      totalBadgesCount
    }
  }
}

TE Academy has minted two classes of NFTs with various NFT types each:
a) proof-of-knowledge (e.g. for passing a TE Fundamentals exam), query “rafts:29”
b) proof-of-special contributions to the community / the field of token engineering, query “rafts:74”

[skrillah]

Week 02 Update

During the second week of the course, we focused on refining and finalizing the requirements for our Reputation-Weighted Voting (RWV) mechanism. The session was divided into two main parts: reviewing the gathered requirements and discussing various aspects of mechanism design. This week emphasized the importance of accurately defining requirements to ensure the success of our voting mechanism.

Achievements:

Requirements Review:

• Finalized a comprehensive list of 24 requirements for the RWV mechanism.
• Discussed and clarified each requirement with input from participants and TEA members.
• Created a final version of the requirements document.

Mechanism Design Discussion:

• Explored different parameters and voting mechanisms.
• Discussed the importance of weights and gates in the voting process.
• Highlighted the need for testing hypotheses and considering edge cases.

Practical Implementation:

• Joan Presented a Python implementation for creating different voting algorithms (GitHub Repository).
• Introduced a spreadsheet template for modelling voting designs developed by @FtheDev.

FtheDev experiments with the RankedVote Addon for Chrome

[akrtws]

Week 03 Update

In this third week of the program, we started to work on voting designs!

The task:

a) create a first voting design, specify it in a text document
b) create a model based on this specification (use the spreadsheet template, or Python or ipnb…)
The model and documentation must include:

• AS INPUTS:
  • who’s eligible
  • voter information (information/NFT proofs involved)
  • ballot design (what to vote on)
• AS OUTPUT:
  • voting result (fellowship winner)
• FORMULAS OR STEP-BY-STEP DESCRIPTION
  • vote aggregation (how the voting result is going to be computed, voting rule)

Achievements:

Voting Designs - Presentations:

• @joanbp – GroupHug
• @FtheDev / @jade – Quadratic Credibility
• @jonas – Rank n’ Slide

Verification discussion - Presentations:

• @OneLV – secure against manipulation
• @Skrillah – refine design goals @Eren – spreadsheet verification tool
• @octopus – Basic Voting Calculator in Python

With these designs and verification approaches on the table, we enter the verification phase in Week 4: we’ll run simulations to see if the voting designs are robust against attacks, and provide outcomes that fit our requirements!

Follow our work here!

GroupHug Conceptual Design: Voting in stakeholder groups (@joanbp)

Quadratic Credibility Conceptual Design (@FtheDev / @jade)

Attack Scenario - most difficult vs. easiest to manipulate (@OneLV)

basic-voting-calc/ Voting Mechanism Simulation Examples (@Octopus)

[akrtws]

Week 04–05 Update

The last two weeks have been about building a model to run simulations and verify our assumptions about the voting outcome.
Are the mechanism robust against attacks, like Sybil attacks? Can we avoid dictatorship? And – if in conflict – how can we prioritize?

@Octopus created a framework to compare voting designs:

• experiment template
• dictatorship experiments
• max_disagreement experiment
• Nakamoto Coefficient

The Github repo is available here!

Based on our verification, we decided on the voting design to apply for the Fellowship Voting: it’s Rank n’Slide!
Compare the voting design options here!

Additionally, we ran a workshop discussing the voting design beyond the voting rule:

• the information about candidates we provide for voters
• the voting UI
• if preliminary voting results should be visible to voters
• who’s eligible to vote

Learn more here!

We are now working on finetuning the weights for the reputation NFTs, stay tuned!

Verifying dictatorship resistance @Octopus

Verifying dictatorship resistance/results @Octopus

Measuring the state of the ecosystem @onelv

Boosting weights @Octopus