Proposal: Experimental Incentive System for Active ArbitrumDAO Delegates

[cattin]

Proposal: Experimental Incentive System for Active ArbitrumDAO Delegates

Abstract

We introduce an Experimental Incentive System aimed at the ArbitrumDAO delegates with a duration of six months. The goal is to assess the impact of the incentives on the delegates’ active participation and involvement in the DAO’s governance. This proposal is subject to changes.

Motivation

This proposal represents an improved version compared to previous initiatives and requests for comments (RFCs):

• [RFC-2] Delegate Incentive System for ArbitrumDAO
• [DRAFT] Experimental Incentive System for Active ArbitrumDAO Delegates

We have developed this new proposal considering the comments and observations provided by other delegates. Our additional motivation lies in the recent increase in proposals and increased activity within ArbitrumDAO. We believe that delegates are the core base of the DAO and should be properly incentivized to maintain active participation, both in voting and feedback on proposals and in the ideation of governance processes.

Delegate Incentive System

Before reading the proposal, note how we have decided to structure it

This proposal is structured into four main sections:

• Option 1: This is our main proposal detailing all the specifications of the incentive system.
• Option 2: Introduces the implementation of “Karma” to automate certain processes. It’s important to note that this option entails additional costs.
• General Summary: An overview of the proposed items.
• Next Steps: Voting deadline

Option 1 - Main Proposal

Specifications

Duration of the Incentive System

We’ve allocated an initial trial period of 6 months. This timeframe should allow us to gather initial metrics to gauge the system’s functionality and its anticipated impact.

Total Number of Delegates to Receive Incentives

We have set the number at 50 delegates. This figure isn’t based on specific reasoning, but we believe it’s a reasonable starting point.

Funding

• Incentives for delegates: 1,500,000 ARB
• Incentive Program Administrator: 25,000 ARB

Delegate Selection Process

Requirements

Based on the feedback received, we’ve opted for a single-tier instead of a multi-tiered program, allowing a wider array of participants, the requirements would be as follows:

• Voting Power: >50K ARB, corresponding to 194 delegates. (Source: Arbitrum Delegates and Voting Power - Dune Analytics).
• Delegates to Receive Incentives: 50.
• Budget Allocation: 1,500,000 ARB (5,000 ARB per delegate per month).
• Historical Participation Rate (Tally): Over 25%

Note:

• So following this model, the top fifty delegates - defined by their final score - within the range, would be compensated
• The 5000 ARB per month is the maximum amount a delegate could earn, but this amount could be lower depending on their score.
• Certain votes won’t be counted towards the Tally Participation Rate, as detailed in the following links:

1. Core Test Proposal
2. Treasury test proposal
3. [OLD] AIP-1.1 - Lockup, Budget, Transparency

Incentive Program Application

Delegates matching the requirements must confirm their participation in the Incentive System in the forum (a dedicated channel will be established for this). They must post using the template provided below, within a 7-day application window.

Incentive Program Confirmation Template:

• Forum Username (Link):
• Twitter Profile (Link):
• Snapshot Profile (Link):
• Tally Profile with Exact All-time Participation Rate % (Link):

Note:

• Failing to send a confirmation message will exclude you from the incentive system, regardless of eligibility.
• Applicants should ensure accuracy in numbers and links, facilitating quicker verification.

Eligible delegates for the incentive program will be announced on the forum.

Scoring and Framework

To determine which delegates will be paid on a monthly basis, we will use a point system and a framework that will be made public.

Framework - Experimental Incentive

Details: Terminology, Symbols, and Formulas

We explain in detail the framework and points system:

• Delegates (DD): Delegates with right to access to incentives
• Ranking (TOP): Delegate’s position in this Dune table https://dune.com/pandajackson42/arbitrum-delegates-and-voting-power
• Funds ARB (FARB): The amount of ARB allocated per month to the delegates’ payment
• Activity Weight (%): Represents the weight assigned to each key activity to be measured in delegates.
  • Participation Rate (PR) - Weight 20: Percentage of the total participation of the member in votes. This parameter is extracted from Tally. This is the only parameter that is not reset monthly.
    • PR% formula: (PR * 20) / 100
  • Snapshot Voting (SV) - Weight 15: Percentage of delegate participation in snapshot voting. This parameter is reset at the beginning of each month.
    • Tn: Number of total proposals that were sent to snapshots for voting in the month.
    • Rn: Number of proposals the delegate voted on in the month.
    • SV% formula: (SV(Rn) / SV(Tn)) * 15
  • Tally Voting (TV) - Weight 25: Percentage of delegate participation in on-chain voting in Tally. This parameter is reset at the beginning of each month.
    • Tn: Number of total proposals that were sent to Tally for voting in the month.
    • Rn: Number of proposals the delegate voted in Tally in the month.
    • TV% formula: (TV(Rn) / TV(Tn)) * 25*
  • Communicating Rationale (CR) - Weight 25: Percentage of communication threads with the justification of the delegate’s vote on the proposals sent to snapshots and Tally (if necessary if the vote does not change). This parameter is reset at the beginning of each month.
    • Tn: Total number of proposals that were submitted to a vote.
    • Rn: Number of real communication rational threads where the delegate communicated and justified his/her decision.
    • CR% formula: (CR(Rn) / CR(Tn)) * 25
  • Commenting Proposals (CP) - Weight 15: Percentage of proposals where the delegate asked questions or generated important discussion for the advancement of the proposal. This parameter is reset at the beginning of each month.
    • Tn: Total number of formal proposals posted on the forum.
    • Rn: Number of actual proposals where the delegate made a genuine and quality contribution. Spam messages will not be considered.
    • CP% formula: (CP(Rn) / CP(Tn)) * 15
  • Bonus Point (BP) - Extra +30% TP: This parameter is extra. If the delegate makes a significant contribution to the DAO, he/she is automatically granted +40% extra TP. This parameter is at the discretion of the program administrator. This parameter is reset at the beginning of each month

• Total Participation (TP): Sum of the results of activities performed by the delegate. A TP% of 100 indicates full participation.
  • TP% formula: PR% + SV% + TV% + CR% + CP% + BP
• Payment ARB (PARB): Final amount of ARB that the delegate will receive based on his TP% and his position in the table.
  • PARB formula: IF(TP% >=100; 50000 ;(5000 * TP% / 100))

Parameter summary

• Activity Weight (%):
  • Participation Rate (PR) - Weight 20%
  • Snapshot Voting (SV) - Weight 15%
  • Tally Voting (TV) - Weight 25%
  • Communicating Rationale (CR) - Weight 25%
  • Commenting on Proposals (CP) - Weight 15%
  • Bonus Point (BP) - Extra +30% TP (Total Participation)
• Total Participation (TP):
  • TP = PR% + SV% + TV% + CR% + CP% + BP

Weight and activity considerations

The activities that have been considered in shaping the framework and scope are measurable activities that we believe can positively impact governance.

We have distributed the weight according to the type of activity, as we believe this is the fairest way to do it at the moment.

Surely both the activities and the weightings are far from being the final parameters, ideally after these 6 months we should have a clear idea of how to modify certain parameters or which ones to add/remove.

Note:
We have not included the governance call in the list of activities because we feel that it may be difficult for delegates in different time zones to attend because of the specific timing of the call. Plus some delegates might not want to dox their voices.

Total Participation (TP)

The 50 delegates who conclude the month with a TP score of +60% will be the ones to receive incentives based on the result provided by the PARB parameter.

Should more than 50 delegates have a TP score of +60%, only the top 50 scores will be selected.

If fewer than 50 delegates have a TP score of +60%, only those meeting the criteria will be compensated

Note:
In the exceptional circumstance where scores are identical or more than 50 delegates exceed +90%, the entire month’s budget may be distributed amongst all delegates with such scores. This decision is at the discretion of the program administrator.

Bonus Point

Should a delegate propose or actively participate in an enhancement proposal for Arbitrum DAO and execute it, like the Arbitrum’s Short-Term Incentive Program (Arbitrum Improvement Proposal), they will be granted an additional +30% to their TP score for making a valuable contribution to the DAO. This could also be a governance process enhancement proposal or a template for grant selection.

We understand that this is one of the weakest points of the proposal and we believe it is very objective. To assess this point we will try to answer these questions

• Is the proposal of interest to the ArbitrumDAO, does it cover any vertical that has not yet been worked on?
• Is the proposal well described, does it have details of the objectives it wants to achieve?
• Does the proposal generate debate among delegates, is there interest among DAO members (for or against) about the proposal?
• Does the proposer exhaust existing resources to demonstrate that their proposal adds value to ArbitrumDAO (governance calls, responses in the governance forum, etc).
• Was the proposal put to a vote: was it a close vote or a resounding no?

When evaluating the bonus points we will try to answer these questions and the administrator will have to upload a post to the forum with the answer and a justification of why he/she grants the bonus point. We also look forward to hearing the community’s opinion on the matter.

Parameter Snapshot

When we refer to a month, we consider from the first day of the month 00:00 UTC to the last day 23:59hs UTC of each month. Within these days and hours, delegates have the opportunity to perform all actions to meet the parameters. The month’s activities conclude on the last day of the month, so delegates must be punctual.

Delegate Responsibilities

Delegates are responsible for keeping track of their actions. At the end of the month, they have until the 3rd of the following month to submit a forum post with the following requirements:

• A copy (Link) of the framework with their activity data for that month.
• Additionally, the post must have a summary with all the links including:
  • Snapshot voting link
  • Forum post link explaining the reasoning behind the snapshot vote
  • Tally voting link
  • Forum post link explaining the tally vote reasoning (if you have changed your vote)
  • Commenting on Proposals link (only those that went to snapshots)
• The summary should also include:
  • Delegate proposal links
  • Links to any significant contribution (discussion, questions, suggestions, etc.) made to a proposal.

Note:

• Delegates failing to post the requested information will not receive incentives.
• To avoid forum spam and disarray, we will ask the facilitators to add a channel for the incentive system so delegates can input all their information there.

Incentive System Administrator Responsibilities

After the delegates post and present their results in the framework, the incentive system administrator has 10 days to gather all the information, review, verify, and unify it. Then, present all the unified data on the forum in the framework and attach a summary explaining that month’s final result.

Once the final result is presented, delegates have 2 days to file a claim in case of disagreement.

Dispute

Delegates have a 2-day window to dispute if they disagree with the results presented by the Incentive System Administrator.

To raise a dispute, they must do so via a forum post with the following template:

• Title: Dispute
• Username
• Reason for dispute (be detailed)

The system administrator must promptly address the issue, resolving it within a maximum of 2 days.

Incentive System Ban

Should a delegate or any community member attempt to deceive or game the incentive system, they will be banned. This decision is at the discretion of the program administrator.

We must remember that this system is experimental, and we hope for community members’ cooperation for its success.

Payment Execution

If all goes according to plan, the delegate payments are expected to be sent as a proposal to Tally, with full details of the payments, between the 16th and 17th. This will be repeated for 6 months so that the DAO can control the incentive program and cut it off whenever it wants.

Payments are made in ARB tokens.

Considerations

Like every new and experimental system, there may be delays, so please be patient.

Estimated schedule:

Administrator Perspective

1. Start of the month: Track delegate participation in governance.
2. Last day of the month: Begin collecting delegate posts from the forum.
3. Review all the information, check for accuracy, and compile it into the framework.
4. Establish the results within the given timeframe.
5. If everyone agrees, send a proposal to Tally with the payment details, to be voted on.
6. Keep the DAO informed

Incentive system administrator

Given the short duration of the program and the lack of a DAO Facilitator, SEED Latam is willing to assume the administration of this program. This means that our delegation won’t participate in it and that the administration won’t be carried out by myself, but by the organization SEED Latam.

However, we’re also open to a group of delegates or potentially a third party administering this program, in which case we would obviously provide any assistance necessary.

For the administration of said program, the budget would be 25,000 ARB. This payment will also be divided into 6 months and will be sent in the proposal to Tally along with the payments to the delegates. In this way, the DAO will have control over the administrator.

---

Option 2 - Karma automation for delegate incentive program

Features summary

Karma will develop features to enhance administrator efficiency and provide transparency to the Arbitrum community. The features that will be built include:

1. Automated fetching of all delegates with Voting Power (VP) over 50k, along with their details.
2. Profile display for each delegate, showing their voting history and voting reasons they post on Snapshot, Tally or Forum.
3. Calculation and display of PR, SV, TV, CR, CP (TBD if CP can be automated) for each delegate as specified in the proposal.
   Karma will implement automatically fetching delegates’ forum data
4. Functionality for the Administrator to allocate Bonus Points.
5. Computation and presentation of the final Total Points (TP) and PARBB for each delegate on: arbitrum.karmahq.xyz/delegate-compensation.

This automation eliminates the need for delegates to manually report their monthly activities.

Timeline

• Week 1: Collaborating with administrators and community members to finalize the project requirements.
• Weeks 2 and 3: Initiating the sprint to design and develop all the features
• Week 4: The product will undergo comprehensive end-to-end testing and be deployed to production. At this stage, the community will gain full access to all features.
• Week 5 and Beyond: Implementation of bug fixes and enhancements based on insights and feedback gathered.

Payment Schedule

• Payment 1: 10k ARB when the proposal passes and the project is kicked off.
• Payment 2: 10K ARB on the successful launch of the product (After week 4 per the timeline above)
• Payment 3: 10K ARB 2 months after launch as we continue to fix bugs and make enhancements (week 12 of program kickoff)

Karma payments will also be executed from the DAO and tracked by the program administrator.

Cost

• 30000 ARB

SEED Latam Commitment

Seed Latam is committed to monitoring the proper functioning of Karma and collecting feedback from the community on the dashboard’s performance. We also commit to collaborating with Karma so that the dashboard meets the needs of the incentive system.

As with Option 1, we are also open to having a group of delegates or potentially a third party administer Karma’s progress and the program, in which case we would obviously provide all necessary assistance.

The budget for the administration of the program for this option would be 20,000 ARB. This payment will also be divided into 6 months and will be sent in the proposal to Tally along with the payments to the delegates. In this way, the DAO will have control over the administrator.

Disclaimer

SEED Latam does not have any commercial relationship with Karma nor does it receive any payment from them for incorporating them into this proposal.

The only link between SEED Latam and Karma is that both want the success of the Arbitrum ecosystem.

Overview

Proposal overview

Duration of the incentive system

• 6 month

Total Number of Delegates to Receive Incentives

• 50 delegates

Selection process

• Voting Power: >50K ARB, corresponding to 194 delegates. (Source: Arbitrum Delegates and Voting Power - Dune Analytics).
• Delegates to Receive Incentives: 50.
• Budget Allocation: 1,500,000 ARB (5,000 ARB per delegate per month).
• Historical Participation Rate (Tally): Over 25%

Eligible delegates must confirm their participation in the incentive system by posting a message on the forum.

Key Points

Scoring

• Option 1:
  • Framework - Experimental Incentive
• Option 2:
  • Karma Delegate Dashboard

Minimum score to receive incentives

• TP= +60%

Bonus Point

• +30% TP

Parameter Snapshot

• Last day of the month 23:59 UTC

Dispute

• 2 days

Payments

• On the 16th-17th of each month a proposal is sent to Tally for delegates to approve payments.

Incentive system administrator

• SEED Latam / Any delegate or entity willing to collaborate

Total Cost

• Option 1

  • 1,500,000 ARB Incentives
  • 25,000 ARB Admin costs

• Option 2

  • 1,500,000 ARB Incentives
  • 30,000 ARB Karma development
  • 20,000 ARB Admin costs

Note
Costs may vary, since we probably won’t give the max amount to every delegate every month, so the cost mentioned would be the max this program would cost assuming we compensate 50 delegates per month for 6 months with the maximum allocation.

Reports

Whichever option is chosen, SEEDLatam commits to deliver two reports, one after the third month and one after the sixth (and final) month of the program with metrics on its impact, delegates’ performance, and expenditures. They will also collect feedback from both the community and participating delegates.

The reports will be a detailed forum post and also accompanied by a community call explaining the metrics and feedback collected.

ArbitrumDAO has the power to:

• Change the direction or parameters of the program when it deems necessary, with a snapshot vote.
• Terminate the program at any time, with a snapshot vote.

Conclusions

Please note that this initial incentive system is experimental and is not intended to be the final version. With this initiative, we aim to evaluate the performance of the DAO and determine its impact. We also anticipate collaboration from both the DAO and facilitators to ensure its success.

At the end of the six months, SEED Latam commits to delivering a report containing the gathered data and feedback from the governance.

---

Next steps

• Snapshot
  • Vote for Option 1
  • Vote for Option 2
  • Against

[mmurthy]

Hello everyone! - I am the founder of Karma, we have been working with a number of DAOs over the last 2 years building governance tools (primarily contributor reputation tools). I am excited to collaborate with SeedLatam on this proposal. I am happy to answer any questions on the Option 2 of the proposal which involves automating tracking of delegate activity metrics.

[maxlomu]

gm everyone,
I’d like to extend my thanks to @cattin and the SEEDLatam team for their diligent work on this proposal.

I agree that creating a framework to incentivize delegates is essential for our DAO. It not only attracts top talent but also ensures their dedicated commitment.

A critical aspect that should be a cornerstone in every DAO is diversity. In light of this, I want to offer an alternative viewpoint on the proposed framework.

The current framework predominantly rewards those who can engage in numerous discussions, vote, and participate actively in the forum, a method I refer to as “horizontal quantity”.
This approach particularly benefits delegates who can dedicate all their time to the DAO, often those deeply involved in political activities and backed by organizations that can divide tasks among multiple members.

However, the current framework falls short in acknowledging the role of “builders” – individuals and organizations possessing specialized, high-caliber expertise, or what can be termed as "vertical quality " knowledge and skills. These builders play a pivotal role in enhancing the Arbitrum ecosystem, notably through protocol development, attracting new users, and generating revenue for the Arbitrum DAO. Nor does it favor individual delegates who, realistically, have less time to review, discuss, and vote.

While I anticipate the evolution of party systems within DAOs over time, I would like to see a framework that encourages diversity without necessarily leading to complete politicization. This isn’t a critique of political organizations; rather, it’s an aspiration to cultivate a DAO characterized by a rich multitude of perspectives. The key lies in crafting the right incentives.

Incentivizing individuals with deep vertical expertise, rewarding quality contributions over sheer quantity, and encouraging impactful actions are pivotal. Here’s how I envision implementing this:

1. Base Compensation for All Delegates: Implement a baseline compensation for every active delegate. This would be way below the 4000 ARB initially suggested, yet still proportional to their Tally score and other relevant criteria - and without the need for a 70%+ pre-requirement.

2. Rewarding Proactivity and Initiative: Place significant emphasis on proactive involvement. Develop a system for endorsing and co-sponsoring proposals. Delegates who sponsor proposals that gain approval should receive additional rewards. This incentivizes not just participation, but meaningful, result-oriented contributions.

Thanks and looking forward to some feedback on this.

[Bob-Rossi]

+1 in acknowledging the work that SEEDLatam has put into this with multiple forum detailed form posts & addressing feedback with each iteration.

In totality, I would strongly vote in approval of a proposal of this type. I think that being an active and effective delegate for a DAO of this size / importance to the space can be time consuming… and to reward those who take on the task in the capacity noted above isn’t unreasonable. Long-term the more delegates we can get active in the forum the healthier & more decentralized the DAO can become. Incentivizing delegates to be active in the DAO should help both retain and hopefully entice new delegates to join as the time goes on.

The foundation of this proposal is a good one. 6 months is a good timeframe. Moving to ARB as the budget is perfect for ease of administration, as well as gives incentive for delegates to make decisions with the health of the ARB token in mind. The cost & scope - 1 Million ARB is IMO a completely fair, arguably low, cost for a project like this. Considering other approved projects of this scope. I agree with avoiding an overly complex system, especially for a 6 month trail run.

I’ll add, forcing delegates to actively apply and holding themselves accountable with reporting threads is a must for me.

Tiers: It looks like the limits were designed to put about 50 potential delegates in each tier. (Or maybe that is just pure coincidence?). Either way, probably my only real reservation / concern with the proposal would be that at 10 delegates a tier it’s seems narrow in scope. I’d rather see more delegates covered, balanced by lower ARB incentives. i.e. I’d personally rather see something like 20 delegates a tier with the chance at 4k / 3k / 2k ARB in each tier. At current prices, those are still fairly hefty monthly incentives considering the role.

If anything is taken from this post… I do strongly think the 10 per tier limit being bumped up is something to consider if not addressed in prior discussions. Bumping up delegates paid gives opportunity to entice more delegates to meaningfully participate. I simply bring this up as I’d hate to see a situation where a tier has something like 12 delegates who are all doing a really great job, but only 10 get paid. (and yes, I know a cap of say 20 has the same problem if 21 people apply… but at that amount it’s seems less likely to have 21 delegates all doing near perfect delegate work for 6 months straight.) And the two who are doing near equal work are getting nothing, resulting in them just stopping their work out of frustration… Or the flip side, where there are 40 delegates in a tier that apply and the bottom half feel like it’s pointless to even try because they have 0 chance of making top 10 anyway.

To offer a solution, maybe the alternative is as simple as adding some flexibility in the count depending on applicants per tier? Or maybe the 95% rule moved down to 90%? I just want to avoid the STIP scenario where we are making assumptions on participation rates with 0 historical data, only to find the actual participation wildly out of whack for reason’s not foreseen.

Scoring System: The formula looks to be pretty fair to me, and well thought out. I think the bonus points is a nice idea to resolve the previous issue of ‘proposal spamming’. It also allows for a good mix between objective standards with a little bit of subjective leeway for extraordinary performance.

The rule about exceeding 95% score I’d almost rather see as a fixed rule, versus as a discretionary tool. To hit 95% you’d have to be a longstanding delegate (due to the historical Participation Rate being 20% of the weighting) with an essentially perfect month of voting + communicating. If we see 11 or more delegates hit that, they all deserve to be paid.

Karma: I admittedly don’t know a ton about them. From what I read in the post, I don’t have issue with it as a general idea. However, if this - Delegates of Arbitrum DAO (karmahq.xyz) - is the dashboard I’m not sure if it’s super accurate, at least based on reviewing my profile. My voting statistics seem way off…? Although it sounds like that is to be updated as part of the 20k cost (which seems reasonable from what I can tell, if they do what is noted in the OP). It looks like they can also add Snapshot votes as well in their dashboard, which is a good feature if we want to go that route.

In summary, I think this is at a minimum worth putting to Snapshot vote to get larger community feedback. I would vote “For” this if it did go to vote, as on the whole it seems like a mostly well thought-out and discussed proposal. My only real critique would be I’d rather see more delegates rewarded as I noted above in the discussion regarding the tiered incentive system, but if it the community at large is okay with 10 I wouldn’t let it stop me since it is ultimately a trail run.

Edit: For transparency, I’ve left my original wording about Karma above. But a quick update since then - they did reach out super quickly and addressed the issue of counting the Tally votes improperly. My dashboard looks correct now and the experience / talking to their team has led me to believe they would do a good job if we used their platform. And having a chance to use look at it a little more I’d be for it.

[pedrob]

Hey @maxlomu, very interesting perspective!

I agree with the idea that constructive and valuable contributions should be rewarded. However, I’m not entirely sure (speaking from my personal opinion) that the delegate program is the way to implement such compensation.

Mixing the role of reviewing and voting on proposals with a benefit for supporting and approving them can lead to a situation of collusion, where the interest in obtaining the reward for proposing / the endorsement outweighs the critical and constructive scrutiny. Ultimately, this could lead to a deterioration in the quality of the approved proposals.

That said, it would be interesting to discuss your idea of developing a system for rewarding individual builders that bring value and generate revenue for the Arbitrum DAO.

[Patrick_J]

Hi Cattin (and SEEDLatam) ,

I was excited to see this proposal for Arbitrum delegates to receive compensation for their time and work. This is a really tricky aspect of DAO governance that I have spent a huge amount of time personally considering, so please take any criticism as it is intended, to be constructive in order for you to improve your proposal. I believe that ArbitrumDAO should be striving to make the best implementation possible of delegate compensation.

I suggest taking the time to properly think through the implications of this proposal will lead to a much better outcome. Once a version of compensation has been launched, it will become much harder to change it later due to the vested interests of those who are benefitting from the implementation.

In reflecting on the STIP process, I believe there are valuable lessons to be learned, particularly in the realm of governance and decision-making for ArbitrumDAO. While STIP’s intent was commendable, the execution demonstrated some challenges, notably during the voting period. The compressed timeframe for assessing and voting on numerous proposals highlighted the importance of thorough evaluation in governance. This experience suggests that a measured, careful approach is beneficial, especially when dealing with proposals involving significant decisions or expenditures. With this in mind, I would recommend a more deliberate pace for the delegate compensation proposal, ensuring ample time for consideration and discussion. This approach aligns with the principle of ‘measuring twice and cutting once,’ prioritizing thoughtful deliberation over speed in the interest of robust and effective governance. I don’t believe that the experience of STIP should be used as a justification for this proposal, as in your Motivation section, as the aim should be to avoid a repeat of that experience in the first place.

My Relevant Experience

Until August this year, I was Facilitator of the GovAlpha Core Unit at MakerDAO. One of our responsibilities was the management of the Recognized Delegate compensation program at Maker, which was later removed as part of the transition to the Endgame Plan. Maker was the first major DAO to implement a form of delegate compensation, and the initial program led to a number of Recognized Delegates coming onboard before the DAO voted to move in a different direction.

GovAlpha was not the primary author of the compensation scheme at Maker; it was written by a community member, although we provided feedback and, over time, updated some language and mechanisms. We were, however, responsible for its maintenance, monitoring, and ensuring payments were included in executive votes (bundled onchain proposal at Maker) in a timely manner.

The Maker model has been adapted and refined by a number of DAOs, although it is by no means a perfect model. I would not advocate for it to be adopted by Arbitrum DAO, as I believe there are many ways the original design could be improved. However, I believe our hands-on experience running a delegate compensation program at one of the oldest and largest DAOs in the ecosystem gives me, and the rest of the GovAlpha team, a unique expertise and perspective to provide feedback on this proposal. I aim to offer constructive feedback on this proposal, and I will certainly highlight areas that I think are strong, as well as areas that could be improved.

Feedback

Initial Implementation as a Trial

Maker also started with a trial of delegate compensation, first proposed by our team over two years ago. Through this trial, we were able to identify multiple pain points and iterate on the design to improve the draft of the MIP prior to implementation. I think starting with a trial of your proposed implementation is a very good idea. Six months seems like a reasonable amount of time if you are committed to actively iterating and improving the methodology through learned experience.

Total Number of Delegates to Receive Incentives

@cattin

We believe that keeping delegates active and incentivized is important as they are the basis of decentralization for ArbitrumDAO.

I could not agree more with this statement from your Motivation section. However, I am somewhat confused by the choice of the multi-tier system described in your proposal. I believe it fails to respect the will of Arbitrum delegators while remaining potentially open to griefing attacks.

@cattin

The idea of setting the tiers is to encourage delegates to seek more delegate votes or to set a healthy %VP for both themselves and the governance - while also enabling smaller delegates to be compensated for their active contributions to the DAO.

I am not convinced that the proposal necessarily achieves this, as smaller delegates are still potentially at risk of foregoing any reward for their time.

The Issue with Tiers

The primary issue with your design for tiers is that it places arbitrary criteria for deciding which tier delegates should fall into. Delegates and their delegated votes represent the liquid will of ARB token holders. The more votes delegated to them, the more trust that holders have placed in a Delegate to represent them. Due to the cut-offs of the tier system, your model does not respect the will of token holders.

Imagine a scenario where the 11-15th delegates fall within Tier 1; they will not receive any reward for the size of their delegation. The 16th delegate falls within Tier 2, and they will receive compensation. A scenario hereby develops where delegates may be perversely incentivized to lose a portion of their delegation to fall into the lower tier to gain access to rewards, which is not a healthy scenario for delegates to find themselves in. Or, a scenario exists where the number 1 delegate in Tier 2 will want to actively discourage new delegations to their platform because it will push them into Tier 1, but they will not form part of the top 10, and they will no longer receive any compensation.

Potential Griefing Attack

Similarly, there is a potential griefing attack here, where if the 16th place delegate is close to the Tier 1 threshold, the 11th placed delegate in Tier 2 may be incentivized to grief them by pushing them into the Tier 1 bracket through delegating their own tokens to them, in order to become eligible.

Choice of Amounts for Tiers

Upon reviewing the Dune dashboard linked in your post (which is great, by the way!), you are currently proposing to reward the following ranks of delegates by ARB delegated (ignoring other qualifying factors and assuming everyone opts in):

• Tier 1: Rank 1-10
• Tier 2: Rank 52-61
• Tier 3: Rank 99-108

These ranks seem quite arbitrary. Do the delegates between 11 and 51 provide no value? Why is rank 61 valuable to the DAO to the extent of receiving compensation, but rank 62 is not?

There is nothing wrong with trying to foster a spirit of competitiveness among delegates, but achieving this through the arbitrary setting of compensation tiers seems counterproductive and is likely to lead to frustration and disillusionment amongst the delegates.

Similarly, it is not logically consistent to state that this empowers smaller delegates, since there are many more small delegates that do not qualify for compensation than those that do under this proposal.

Remedies for this may include increasing the size of the tiers, but I would seriously suggest rethinking how you decide which delegates qualify for compensation in the first place and consider scrapping the tier system entirely.

Alternatively, you could have an infinite number of (larger) tiers that automatically adjust their proportion of allocated ARB as more tiers come online using a pre-defined mathematical formula. Although this is more complex it would remove any subjectivity from the process.

@cattin

For the moment we believe that it is not necessary to make the incentive system complex, we should try to start with a system as simple as possible so that ArbitrumDAO members can evaluate if this type of incentive has a positive impact on Arbitrum.

I don’t think this is a sufficient justification for adopting a system with the flaws outlined above. It is evidently difficult to come up with a workable solution; we certainly faced challenges at Maker, and other DAOs have also been grappling with this. However, this solution has significant flaws, and its perceived “simplicity” is not enough to counter the issues present. That said, the proposed system is already quite complex, so it seems like an inconsistent position to leave the tiers as they currently are to preserve simplicity.

Tally Exclusivity

Why is Tally participation a measure for qualification, but not Snapshot? Is this an issue with data availability? Surely, participation at both stages of voting is part of the actions of delegates that you want to reward?

Applications

Making this opt-in is a good idea. I would suggest the addition of something like requiring delegates to agree to follow a Code of Conduct that we had at Maker. A similar code of conduct is currently used by Aave.

Scoring System

In general, I quite like this section and the breakdown of contributing elements. It all makes sense, and these are mostly items that you would want delegates to be performing.

It’s unclear to me whether a delegate’s TP can exceed 100%; you may wish to clarify whether it is capped at 100% or whether it is possible to get up to 125%. This may become relevant in the event of a tie-break scenario.

One note on communication tracking: this is a surprisingly large amount of work. We spent a significant amount of hours tracking this at Maker, and you are looking at doing this for more delegates, and across more criteria. You should seriously consider whether you could use some automated system such as SourceCred or Karma to generate scores for the Commenting Proposals segment of this, otherwise, it may quickly become unmanageable.

Monthly Resets

I would suggest you consider using a rolling measure over the most recent 3 months rather than monthly resets. This strikes a nice balance by rewarding both recent performance and also rewarding delegates who have been consistently active, although I think this is a minor issue.

Bonus Points

As you note, this is a very subjective category for rewarding delegates. In general, I would suggest trying to make the criteria as objective as possible to avoid disagreements further down the line.

Multisig Usage

Arbitrum has an active Tally instance which is more than capable of executing payments to delegates through an on-chain vote. The use of a multisig adds the possibility that multisig signers will wield undue influence over delegates, either knowingly or unknowingly.

Clawback

In the event that the multisig signers are corrupted, there is almost zero chance that the DAO will be able to claw back funds by drafting and executing an on-chain vote without the corrupted signers simply transferring the funds to a new wallet. This protection is a paper tiger and should not be relied upon.

Signers

I would urge you to try and identify signers who are not delegates. This is a potential conflict of interest.

Incentive System Administrator

Please correct me if I’m wrong, but to my knowledge, the role of Governance Facilitator is currently not official in the ArbitrumDAO. Therefore, this proposal effectively gives SEEDLatam control of the incentive program. However, as written, the proposal does not reflect the current status of Arbitrum governance.

Similarly, as someone with experience in both governance facilitation and facilitation of delegate compensation programs, I can attest that both roles require a significant time commitment. Your proposal appears more complex than the system I have experience with. I am not sure that governance facilitators will necessarily have the time to do this properly. However, I concede that the governance facilitation role at Maker is significantly different from the role in other DAOs that have adopted the nomenclature used by Maker without the same roles and responsibilities.

Karma

I actively support the use of tools such as Karma for measuring delegates’ contributions to governance. This will save a significant amount of time for the administrators of the program, while also removing potential sources of bias. If the DAO votes to support this proposal I am of the strong belief that they should also approve the Karma component. The price is also very reasonable for a six-month trial period.

[cattin]

Hi @maxlomu,

We really appreciate your feedback and we’re working on finding a way to implement it in our proposal, but in the meantime here are some thoughts:

I would like to emphasize that our framework does not seek to favor any particular type of delegate. It is intended to encourage active participation in the DAO, regardless of who the participants are (builders, protocols, individual delegates, etc.). It is important to recognize that this framework is experimental and far from perfect. However, starting with a baseline is essential, evaluating the results over a given period, and then making adjustments for improvement. We are aware of the difficulty in achieving a fair balance, but we are confident that, over time, we will be able to establish the right incentives.

Regarding your comment on the high entry participation rate and its possible impact on reducing the number of participating delegates, we agree with what you mentioned. We propose to modify the following parameters:

• All-time participation rate - Tally: +50%, taking into account votes from a 4-month period (current rate is +70% considering votes from day 0).
• Total Participation (Total Participation - TP): + 50% this parameter determines which delegates enter the top 10 (Currently set to +70%)

In addition, we are considering the possibility of increasing the number of selected delegates from 10 to 15 per level. However, this would also imply requesting a larger budget, on which we do not yet have a final decision. We would like to receive feedback from other delegates in this regard.

Regarding compensation for proactivity and initiative, we completely agree with you. It is a challenge to distinguish actions that benefit the DAO and avoid incentivizing initiatives that only favor a small group or the delegates themselves. For this reason, we have included in the proposal a bonus point system (+25% in participation points - TP), aimed at rewarding those delegates who contribute with initiatives beneficial to the DAO. Initially, these bonus points will be awarded by the program administrator. We are interested in evaluating this dynamic and, in the future, proposing a more structured process.

@cattin

Bonus Point

Should a delegate propose or actively participate in an enhancement proposal for Arbitrum DAO and execute it, like the Arbitrum’s Short-Term Incentive Program (Arbitrum Improvement Proposal ), they will be granted an additional +25% to their TP score for making a valuable contribution to the DAO. This could also be a governance process enhancement proposal or a template for grant selection.

Such scenarios are highly subjective and will be left to the judgment of the incentive system administrator. For this reason, we have not included it as a parameter within the table since it can be challenging to assess what genuinely adds value to ArbitrumDAO. Thus, we have decided to handle them this way.

However, I do want to emphasize that we’re open to other ideas on having a less arbitrary framework for these specific scenarios to better incentivize proactivity.

[cattin]

Hello everyone,

We would like to thank @maxlom0, @Patrick_J, and @Bob-Rossi for their valuable comments and feedback. Special thanks to @Patrick_J for his thorough response; we at Soveign Finance AVC, created by SEEDLatam within MakerDAO, are big fans of his contribution and dedication to MakerDAO.

We would like to share some thoughts:

• I would like to emphasize that our framework is not intended to favor any particular entity. It aims to encourage active participation in the DAO, regardless of who the participants are (builders, protocols, individual delegates, etc.). We try to make it as neutral as possible.
• It is important to recognize that this framework is experimental and far from perfect. For this reason, we understand Patrick’s concerns and are very thankful for the feedback we received.
• We believe starting with a baseline is important, evaluating the results over a given period, and then making adjustments to improve. We are aware of the difficulty of achieving a fair balance, but we are confident that, over time, we will be able to establish the right incentives for ArbitrumDAO.

Tiers and Number of Delegates: Evaluation and Proposal

We recognize and share the concerns raised by @Patrick_J. Initially, we implemented a tier system to mitigate incentive manipulation. However, it is clear that this approach requires significant adjustments. In response to this, and also addressing the concerns of @Bob-Rossi and @maxlom0, we propose the following structural changes:

Delegate Requirements:

• Voting Power: >50K ARB, corresponding to 194 delegates. (Source: Arbitrum Delegates and Voting Power - Dune Analytics).
• Incentive Eligible Delegates: 50.
• Program Duration: 6 months
• Budget Allocation: 1,500,000 ARB (5,000 ARB per delegate per month).
• Historical Participation Rate (Tally): Over 30%, considering votes from a 4-month period (current rate over 70% since inception) ~ *note, this would be the historical participation needed to qualify for the program.

Tally Exclusivity

We use Tally as the main source of data to make it easier for delegates to calculate their participation rate since it already provides it, considering that now, due to the STIP, there are also more than 100 proposals on Snapshot. But we are open to changing this (maybe karma could help aggregate snapshot participation).

Scoring System

We just want to clarify that the maximum TP currently is 125%, but the limit of the amount paid to delegates would just be 5k ARB (corresponding to a TP of 100%). The remaining TP would just help them improve their rank.

Note that delegates are responsible for reporting their activity in an orderly fashion, thus avoiding overburdening the program administrator.

@cattin

Delegate Responsibilities

Delegates are responsible for keeping track of their actions. At the end of the month, they have until the 3rd of the following month to submit a forum post with the following requirements:

• A copy (Link) of the framework with their activity data for that month.

• Additionally, the post must have a summary with all the links including:

• Snapshot voting link

• Forum post link explaining the reasoning behind the snapshot vote

• Tally voting link

• Forum post link explaining the tally vote reasoning (if you have changed your vote)

• Commenting on Proposals link (only those that went to snapshots)

• The summary should also include:

• Delegate proposal links

• Links to any significant contribution (discussion, questions, suggestions, etc.) made to a proposal.

Bonus Points

In response to @maxlom0’s proposal on compensation for proactivity and initiatives, we are incorporating a bonus system (+25% in TP points) for delegates who contribute significantly to the DAO. These points will initially be allocated by the program administrator, but we do think we need at least some sort of rubric that the administrator could use to determine how many bonus points a delegate should receive for specific actions.

Our original intention was to encourage delegates to be more proactive, and ideally compensate those who manage to push proposals like the following to snapshot with bonus points:

• Proposals related to governance processes and templates. For example: [PROPOSAL] - Consolidate Security Proposals into a RFP Process
• Proposals related to initiatives that benefit the DAO. For example: Arbitrum's Short-Term Incentive Program (Arbitrum Improvement Proposal)

Note:

• The program administrator should post a brief justification for the bonus points awarded in the forum.

Specific Considerations

• Applications: We agree here, we can get them to accept Community Guidlines or work on drafting a Code of Conduct.
• Use of Multisig: We agree with implementing recurring onchain votes for monthly payments to delegates, although we know this could be tedious for the DAO. We seek additional feedback on this aspect.

Incentive System Administrator

In the current context, the Arbitrum Foundation has two Facilitators. However, we are aware that we cannot assign additional responsibilities to them as DAOs. The initial uncertainty about the magnitude of the workload to administer this system has been clarified thanks to the feedback provided by Patrick. Consequently, we consider it pertinent to seek the collaboration of another person, delegate or entity, who, in conjunction with SEEDLatam, will contribute to the effective implementation of the program.

We are willing to allocate a specific budget for this collaboration, but we wish to confirm if this proposal has the governance endorsement. We also want to emphasize that we offered to administer this program from SEED Latam in case we don’t find a facilitator willing to run it ~ and if this happens our delegation would obviously not partake in the program, regardless we do understand the concerns raised plus potential conflicts of interest.

This approach highlights the relevance of roles such as facilitator. Previously, we proposed an RFC to discuss the integration of the Facilitator role in the DAO. At that time, the activity and initiatives were not as significant as they are today, which is why we did not move forward with the proposal. Currently, we have a draft for the election of governance facilitators, which we share below for your review and comments:

[DRAFTV3] Introduction, selection, and guidelines for the Governance Facilitator...

Karma

We agree that if we can automate procedures, it will be easier for the program administrator. We can ask @mmurthy if Karma has any problems or suggestions with these changes.

[Bob-Rossi]

Appreciate the update / revisions that take into consideration the feedback.

I think after hearing more thoughts about the tier system, scraping it is worthwhile for scenarios mentioned by @Patrick_J. I also see the eligible delegate pool has increased to 200 or so, and those who get paid out from 30 to 50, which I think is good as well. Among the other changes, I think it is in a good state for voting.

I saw the discussion regarding snapshot. I think something to add is that the Scoring System does include Tally Votes, so though it is not part of the historical Participation Rate metric, it still is reflected in 15% of the weighting. So while Tally has more weight overall, voting on Snapshot has not lost impact completely. If there is large pushback to include the Snapshot metrics, my suggestion would be just ignoring the any votes related to the Round 1 STIP. I know that isn’t a great solution, but the large vote count massively throws off Snapshot metrics to the point I’m not sure it accurately reflects the spirit of a historical participation rate. i.e., someone who voted only in the STIP will have a far better participant rate then someone who voted in everything but the STIP. And beyond even all that, makes tracking a nightmare. While not perfect, it may be a worthwhile compromise. As those who want Snapshot history included do get it included, while also keeping tracking complexity down.

[glitchc]

This proposal is not just about tokens; it’s about our shared destiny.

In the heart of our Arbitrum DAO lies a fundamental principle: decentralization. It’s not just a buzzword; it’s the very essence of our community-driven ecosystem. As we deliberate on proposals, we must remember that every active delegate contributes to our shared vision. Let’s explore why rewarding all participants, regardless of voting power, is not just fair—it’s essential for our collective success.

The Power of Inclusion:

Our DAO thrives when everyone has a seat at the table. Excluding delegates based solely on voting power undermines the very fabric of decentralization. Imagine a vibrant marketplace where every voice matters—a place where a newcomer’s idea can be as transformative as that of a seasoned delegate. By embracing all contributors, we foster innovation, resilience, and diversity.

The Fallacy of Centralization:

Centralization creeps in when decisions are monopolized by a select few. High-voting-power delegates wield influence, but at what cost? We risk becoming an oligarchy, where the fate of Arbitrum rests in the hands of a privileged minority. Decentralization, on the other hand, ensures that power is distributed widely. It’s a safeguard against undue concentration and promotes transparency.

The Unsung Heroes:

Consider the delegates who diligently participate, even if their voting power isn’t astronomical. They attend meetings, contribute ideas, and engage with the community. Their commitment fuels our progress. Let’s recognize their efforts. After all, decentralization isn’t about the loudest voices—it’s about amplifying the quiet ones too.

A Proportional Approach:

Rewarding higher-voting-power delegates isn’t inherently wrong. But let’s refine our approach. Why not distribute rewards proportionally? Imagine a system where each delegate’s share aligns with their contribution. It’s a win-win: influential delegates receive their due, and active participants feel valued. This balance ensures that no one is left behind.

Our DAO is more than a collection of wallets; it’s a living, breathing organism fuelled by passion, ideas, and collaboration. Let’s champion decentralization, celebrate all active delegates, and pave the way for a future where every voice echoes through the halls of Arbitrum. Together, we’ll redefine what it means to be decentralized—one reward at a time.

[fADFER]

This proposal is excellent, but the incentive scope needs further consideration. I believe incentives should cover a broader range, not just limited to the 50 delegates holding over 50,000 votes. Those holding a significant amount of ARB tokens already have a greater motivation to care about ARB and may not require special incentives. We must not forget those who hold fewer ARB tokens but still actively contribute to the development of ARB. If we only reward those with large holdings, it may create a perception of making the rich richer.

[Sinkas]

It’s important to keep in mind that delegates are not necessarily token holders and token holders are not necessarily delegates.

[L0tus]

Option 1 presents a comprehensive and manually driven incentive system, involving detailed scoring criteria and community involvement. The proposed scoring system, although complex, aims to fairly evaluate delegates’ participation. It allows the community to assess and vote on delegate performance, promoting transparency.

Option 2 introduces automation through Karma, providing a more efficient process but relying on predefined algorithms. While automation reduces the burden on delegates to report, it may lack the nuanced evaluation that a community-driven approach offers. Karma’s features streamline the process, but may not capture the full spectrum of delegate contributions.

Considering the experimental nature and the goal of assessing community impact, Option 1 aligns better. Its detailed framework encourages community engagement, ensuring a more inclusive and subjective evaluation. However, continuous feedback and adjustments are crucial for refining this experimental system over time.

[Rainy]

While I appreciate the merits of this proposal, I have concerns about the method for selecting the 50 delegates. If the criteria rely on the Dune Dashboard, I find it suboptimal. To illustrate, 21 out of the top 50 received delegation from fewer than 100, and 13 from fewer than 10.

[axlvaz_SEEDLATAM.eth]

Pre-selection is based on a few criteria, just to avoid farming and spamming. Delegates have to express their willingness to participate and the 50 who receive rewards have to compete for income as long as they meet the minimum participation.

[nanopunk]

Opting for Option 1 as it offers a comprehensive point system, ensuring a nuanced evaluation of delegate contributions. Transparency and detailed assessment are key for fostering a fair and accountable incentive structure within ArbitrumDAO.cheers

[thedevanshmehta]

I support this program & applaud SeedLatam for taking lead on a framework for delegate compensation.

Framework → Service provider is the right template for ArbitrumDAO to enter new contractual agreements and i am glad to see a well thought out delegate compensation framework

I have voted against inclusion of karma in this season, mostly because the metrics will likely change after season 1. So much better to keep this as a (cheaper) trial run and if the metrics hold up, we make it automated from next season onwards

[0x_ultra]

Fully in favor of the Experimental Incentive System proposal. The clear structure, detailed framework, and commitment to fairness make it a promising initiative. Excited to see the positive impact on delegate engagement.

[cp0x]

I have some questions about formulas:

1. Snapshot Voting formula is (SV(Tn) /SV(Fn)) * 15, but there is no abbreviation Fn. Also, according to this formula, the less we voted, the higher the indicator, but it should be the opposite.
2. Tally Voting formula is (TV(Fn) - TV(Rn)) * 25, which should match the formula above (after correction), but it is different. I don’t see the point in minus.
3. Communicating Rationale formula is (CR(Fn) - CR(Rn)) * 25, however there is no abbreviation Fn (only Tn and Rn). Also, there is no point in subtracting since Rn is a positive indicator.
4. Commenting Proposals formula has the same problems as in paragraph 3.

[cattin]

Hey! Thank you so much for pointing this out, since this proposal underwent multiple modifications, we made some typos with the formulas and it should actually look more like this, to reflect the formulas we used in the excel:

1. (SV(Rn) / SV(Tn)) * 15
2. (TV(Rn) / TV(Tn)) * 25
3. (CR(Rn) / CR(Tn)) * 25
4. (CP(Rn) / CP(Tn)) * 15