PGM-33 : Allocating PAL to incentivize delegates

Paladin Governance Management (PGM) PGM - Approved
1

TL-DR: Allocate 100,000 PAL to delegate incentives.

Context:

As Paladin grows, it inevitably needs more work than the Core team can provide, and while we expect new team members to pick up some initiatives as revenue ramps up, our treasury is not at a point where it can support this.

Which is why we have started giving out retroactive grants to significant DAO contributions with PGM-26.

Another important part of DAO dynamism is having a high percentage of users utilize their voting power directly or indirectly in order to govern the protocol.

We do not believe it should be expected from all holders to participate in governance, this is simply not feasible when considering the number of governance tokens most users own.

Instead, non-active stakeholders should delegate.

In the current setup, Paladin has ~700 stakeholders, 129 of whom have staked or locked and own hPAL, the governance token. Roughly 15% are active participants in governance and 16% have delegated to active delegates (most original delegates have now ceased contributing).

This means close to 70% of voters are not participating.

As a governance protocol we should strive to have above 60% total participation. For this reason we are suggesting to launch a delegate incentive pilot program inspired from the Hop’ Protocol’s delegate program.

Rationale:

The idea of this grant program is to reward delegates for their participation, depending on the number of users they represent as well as their weight in the protocol.

Eligibility:

  • Minimum Participation Requirement: To remain eligible for delegate incentives, a delegate must maintain at least a 80% participation rate. Participation is measured as voting in any governance vote.

  • An obligation of transparency: the delegate’s position / goal over his tenure must be clearly explained when presenting his delegate initiative and each vote must be justified on his own delegate thread. The delegate must also respond to comments that concern him/her and communicate if he/she does not vote on a proposal by motivating his/her choice.

  • Minimum Delegation Threshold: To remain eligible for delegate incentives, a delegate must maintain at least 10,000 PAL from at least 2 stakeholders excluding themselves.

  • Registered member of the Paladin Association. Delegate roles are essential for Paladin. As the Association represents the general interest of the DAO in a trustworthy environment, it is necessary for Delegates to be a part of it. Delegates may apply for honorary membership in the Association under a pseudonym.

If you fit all of these criterias you will be added to the UI and considered a Recognised Delegate. All other delegates will be Shadow Delegates. This system is heavily inspired from Maker and will strive to create high integrity delegates for Paladin.

Specs:

Delegate incentives increase with the amount of PAL delegated, but in a decreasing fashion. This means that a small delegate will receive a greater incentive per voting weight than a large delegate, but a larger delegate will receive more.

The formula for delegate incentives is:

I = 0.846 * log(h) + 996.617

I = Incentives to be received
h = lowest level of PAL delegated that month

This formula targets ~10,000 PAL for a delegate at the minimum threshold. At the time of writing, 2 delegates qualify for the minimum eligibility threshold. Our hope is to grow this to 5 over the course of the trial.

This formula’s slope was heavily inspired by the Hop proposal. It was originally chosen to provide the targeted levels of minimum and maximum compensation and is open to revision at the close of this trial. In general, it was selected to provide a slower accumulation of incentives for larger delegates than smaller ones, while preserving a noticeable increase at all levels (under the maximum cap) for any delegate that can attract more delegated PAL.

All Recognised Delegates will have to self-report on a monthly basis in this thread in order to get their incentives.

Means:

  • Up to 100,000 PAL from May to August;

  • All inactive delegates will be offboarded from the governance interface;

Sustainability:

We will review this program during the summer to explore its efficiency.

Success will be defined by the following metrics:

  • Total percentage of voters voting or delegating;
  • Participation of delegates;
  • Number of active delegates;

Vote Options:

For/Against/Abstain

Community poll:

  • For

  • Against

  • Abstain

0 voters

4 Likes
2

Thanks for starting the discussion, glad to finally see this topic being considered !

Definitely agree that incentivizing delegates make sense, however it needs to be done carefully to prevent opportunists to take advantage of this program. Very nice to see a good amount of safety measures to prevent abuses.

This can also include comments on the snapshot if no thread imo

In addition to the fact that it seems unfair for larger delegates, this might also encourage “sybil delegate” attacks if the incentive is greater per voting weights for smaller delegate no ?

Not sure to follow what means the 0,846 & the 996,617 -can someone explain please ?

Really glad to finally see delegates being considered and rewarded, however it should not mean that the delegate contributions so far were worthless. To prevent this, I’ll write another proposal (with a smaller budget) to retroactively reward active delegates over the past year, following the framework voted for the next 3 months if approved.

What happens if some inactive delegates come back after this proposal ?
Should they represent themselves and explain why they were inactive ?

Looking forward to learn what the community thinks about the proposal :fire:

1 Like
3

I really love this idea, but do we have some feedback on how this program benefit for the Hop community ?

1 Like
4

I agree, participation is not sufficient, and it’s a great idea to bring this up to discussion.

I think 700 stakeholders is a bit optimistic. You could probably ignore smaller positions (<100$ in value) for which it would not make sense economically to pay 10$ in transaction fees to stake their tokens. So if you apply that filter, you are left with less than 120 addresses that could potentially be encouraged (are the incentives still sufficient?) to stake their PAL tokens (but i guess that would we out of scope of this proposal).

There are roughly 170 addresses that have already staked (hPal holders). You mentioned the 2 participation options:

  • direct participation (which is gasless on snapshot, so it’s really a shame we don’t have more voters) or
  • indirect participation via delegates (again, users will have to pay the delegation tx of 5-10$ so I would apply the same filter as before and consider only hPAL positions worth more than 100$, which brings down the number of wallets to approx 85 addresses).

On average, we seem to get 20-25 votes per proposal on snapshot. If I am not mistaken, delegates represent an additional 22 addresses, so that means that we have about 42-47 addresses represented in the votes out of a total of 170 if we consider all hPal addresses or out of 85 if we only consider the holders worth > 100$.
It would be interesting to recalculate how much that represents in terms of overall voting power.

I might be rambling a bit, but what i am getting to is whether we really have 70% of non participation or are we closer to 50%?

How do you measure the 60%? Is it the percentage of the circulating Pal supply, percentage of voting power, or looking at the number of wallets holding the tokens?

I would like to see more participation and discussion in the forums, so I think it is an interesting idea, but i wonder if the monetary incentive is really going to change that? For the association, there was also a monetary incentive, and there wasn’t exactly a ton of candidates.

Also as Nelson mentioned, does hop have any feedback already on their program? Did it really boost the participation?

It would be interesting to also get a view of delegates that have gone inactive or from some who applied but then never became active… If they moved on to other protocols, did they have any explanation as to why they lost interest in Paladin? Would a small monetary incentive really have made a difference?

We talk a lot about rewarding the delegates (who are doing the work), but is there anything else that could be done for the holders? For example, would a “gas refund program” make sense to reimburse the transaction cost of delegating your tokens to one of the Recognised Delegates?

I am terrible at maths… can you confirm if this calculation is correct?

  • assuming a delegate has the strict minimum of 10,000 PAL delegated to him during the month, is the result just 1,000.001 PAL of incentive for that month?
  • and if the delegate controls 100,000 hPAL, the result would be 1,000.847 PAL of incentive?
    If that is the case, the impact in PAL controlled is insignificant. You could just go for a flat 1,000 PAL incentive a month.

I am curious who are the 2 delegates that would currently qualify… Do you mean that they meet the 10k delegation threshold or all 4 criteria (participation, transparency, delegation threshold and member of Association)?
I would guess dydymoon might be one, but would be the second?

2 Likes
5

Imo we should decide on one way of communicating vote explanations otherwise it will be too messy

It depends what we see as the goal of the proposal:

  • Reward current delegates
  • Attract new deleagtes

The latter is a priority, the former is necessary for durable participation

Its a logarithmic formula meant to achieve 1000 PAL for anyone who reached 100,000 hPAL delegated to him with a limit to 5000 PAL when trending towards infinity.

We can use their v2 proposal as reference: RFC: Renewal of the Hop Delegate Incentivization Trial - 🐰Hop Ecosystem - Hop

Another opportunity is that both @fig and @Kene_StableLab participated in the creation of this framework and could help us in ourt endeavour.

None of them are association members yet, but the delegates are Dydymoon and Kene from StableLabs.

There is probably a way to create a cleaner formula, we remain open to tailoring it. Small note is that we’ve already received interesting from community members in launching their platform.

1 Like
6

Has Kene received 10,000 PAL from at least 2 stakeholders excluding himself, or does he just control the 100k hPal in his wallet?

7

You are right, they only have one person delegating to them. They would need to get at least one more in the current situation.

8

I believe that this is an excellent initiative and I am glad that this framework has taken learnings from the Hop Framework.

It is important that while maintaining an 80% participation rate, the participation rate should be reset at various intervals to ensure new delegates have a timeline in which they can join Paladin Governance. For example the case could be that the participation rate resets every six months.

I believe that having more than one source of delegation is an excellent idea, my only concern is that it many potential delegates may not be able to reach this quota, getting delegation may not be easy for a delegate without the required network and reputation. I believe that if we are trying to attract professional delegate teams and popular individuals then this could work well, but if we are looking to encourage new and organic delegates from the Paladin Community and beyond, this could be a hindrance.

Delegates joining the association is also a very important step in showing alignment with the long term vision of Paladin.

Finally, I believe that each delegate should self-report on their votes in their delegate platform; a good example is the StableLab delegate platform for Paladin; below your commitment should be your voting history in Paladin, as well as your rationale for each vote.

At the end of every month, any delegate can start a thread for delegates to report on their participation and communication rate for every vote. This is an example of what this would look like.

3 Likes
9

I would say that the participation % for a delegate should be assessed over (for example) the 10 most recent proposals or from the time when he applied to be a Recognised Delegate (whichever period is the shortest). I think that would be better than having set periods of time with a reset.

I agree but i also think that it is probably quite easy to game: anyone can set up a fresh wallet with only a handful of PAL tokens, stake+delegate, and voila… you have your 2nd stakeholder!
I mentioned it in my previous message but, what about a gas refund for the delegation tx for current hPal holders with no interest governance?

100%. I think it’s great that you post your voting history with a short recap of the rationale. Ideally that could also be a place for users delegating their voting power to you to ask follow up questions if they wanted to.

2 Likes
10

In general, I think it’s sad that we are finding a way to incite individuals or entities to vote.

Now, if the objective is to increase participation in governance votes (voting without delegation and voting by delegation) then incentivization, although questionable, seems to be a coherent solution.

I have identified 2 difficulties (user & delegate):

Problem 1:
To delegate it is necessary to pay a fee on the mainnet → what interest has the PAL holders have to pay a fee to delegate their votes especially if their only motivation is to make a profit on the PAL?

Problem 2:
How to encourage delegates to vote acidly with the delegation power they are given?

How can new delegators make a place for themselves (as @FrenchTony pointed out, also it is important to have a diversity of delegators)

Here is what I propose:
(it is more or less radical so it is adjustable)

To get a APR on the PAL it would be unconditional to vote or delegate your voting power.

FOR THE USER

-Direct Voting (user own here voting power):
If the user has a participation rate of more than 80% in the voting, the user can claim his rewards. If the user’s participation rate is less than 80%, then no rewards will be given.

-Voting by delegation:
The user can choose to delegate his voting power instead of voting himself. At that moment he chooses a delegate to vote in his place. (I think it would be interesting to expose the participation rate of the votes in order to challenge the delegators between them, it can be an idea)

Voting by delegation 2 possible outcomes:
1-if a delegate has a participation rate higher than 80% then the users who have delegated their power to him receive their return on their PAL .(the totality of the APR no matter the participation rate of the delegator as long as it is superior to 80%, OR in proportion to the participation of the delegator)

2-If a delegate has less than 80% participation, users are not eligible for rewards from their APR (OR in proportion to the delegate’s participation)

In case a delegate is not doing his job, we should migrate his delegations

Other ideas for funding the system described above, would be to vote an allocation of PAL to be distributed to users who vote or delegate their votes. This allocation is offset by a reduction in the APR on PAL.

FOR THE DELEGATORS

I think that it would be interesting that users have a better return (temporarily) when they choose a delegate who is new and with little power. This would give everyone the opportunity to present themselves as a delegate and the reputation would not be the only argument because the participation rate would be visible

The delegate, would be rewarded with an allowance voted by the DAO. As mentioned in previous posts.

Also, it seems that PAL is coming to arbitrum, and in order to bring maximum flexibility to the users I think it would be interesting if they could claim their reward on arbitrum

What do you think?

Also, I take advantage of this proposal to share with you my great motivation to become a delegate, unfortunately the difficulty I am facing is the harvest of the delegation.

4 Likes
11

I like a lot this approach, and it’s something I’ve been considering a lot lately. I guess the biggest obstacle to this is the technical feasibility. @Kogaroshi any opinions?

Rewards can’t be claimed on Arbitrum because they are distributed on Ethereum.

You should create a delegate thread to start campaigning =)

12

Hello, I am an EPFL student and member of the BSA association at EPFL.

I studied the formula for delegate incentives and it seems that it is not optimal to attract more new delegates and neither to encourage someone to do a lot of research to then post on the forum and then accumulate delegated tokens.

As FrenchTony said the difference to having 10 000 tokens delegated and 100 000 tokens delegated is smaller than 1 in reward (less than 0.17 dollar if 1 pal = 0.17$) which is not really encouraging to try to accumulate tokens.It will clearly not be worth it to accumulate 100 000 delegated tokens but it will be worth for someone malicious to create multiple accounts (let’s say 10 accounts) and post multiple messages in the forum with his different accounts so if he gets at least 10 000 tokens delegated on each of his account he will receive approximately 1 000 pal in reward (170 $ ) on each of his account (so 170 x10 = 1700 $ in total).

Instead of earning only 0.846 * log(100 000) + 996,617 = 1000,847 pal =170,14$ if he only had one account with 100 000 tokens delegated (So there is a difference of 0,14$ between someone who got delegated 10 000 incentive tokens and someone who got delegated 100 000 incentive tokens).This does not encourage anyone to participate more in the governance once he got 10 000 tokens delegated.

I agree that this formula is good to attract more little members to participate in the governance but I think we could create a better formula that at the same time encourage people to try to get more tokens delegated to them but also that attract more little members that can join the governance more easily.

Here is a proposal for a formula with a square root instead of a logarithm (I don’t think my formula is optimal at all and it can be improved or neither do I know if it makes sense) :

if x < 10 000 : f(x) = 0

if x >= 10 000 : f(x) = sqrt(x) + 996

(we can maybe lower or increase the exponent of x to improve the formula)

5 Likes
13

Thank you for launching the debate on the intricacies of the formula. You are spot on the current shortcomings of the formula.

As you mentioned, we should target having a good incentive growth between small and large delegates, maybe even with a cap?
We could target 1000 PAL floor for small delegates, then ~ 2500 for 100k hPAL and a ceiling at 5000 at ~500k hPAL?

So something like:

if x < 10 000 : f(x) = 0

if 10 000 < x < 500 000 then f(x) = ?

if x >= 500 000: f(x) = 5000

1 Like
14

I think f(x) = sqrt(x^(1.27718))+641.607 is a good formula for what we are trying to do, and it’s close to the values that you gave me :

    f(10 000) = 1000
    f(100 000) = 2200
    f(500 000) = 4999.78

I can always improve the formula if you think we can do a better one. But what’s interesting is that the growth of the function is slightly decreasing as you accumulate delegated tokens but it’s still really interesting to accumulate the maximum delegated tokens possible. Also the constant gives an advantage to small delegates so more people will want to participate in the governance and the governance will be less centralized.

3 Likes
15

Thank you @Baptiste for your formula! Here’s a little help to visualize it properly so that everyone can express his opinion according to the numbers.

Figure_1
Figure_11000×600 26.3 KB

As you can see it works a bit like experience points in videogame, at the beginning is easier to get more and more PAL incentivizing smaller delegates to keep being active but gives good returns also to bigger delegates. (Shoutout to @Beguin1 for the comparison).

The only con I can see at the moment is that it might even be a bit too easy for bigger delegates to increase their revenues but this is something where I’d like to hear what the community thinks.

4 Likes
16

I understand I just created an other formula thanks to the polynomial interpolation of Newton and it’s maybe better. I just don’t know if it’s maybe growing to fast at the beginning (but I can change that) and is also quite complicated in comparison of the other one :

f(x) = 499267/758131920000000000000000000000x^5 − 23910389/25271064000000000000000000x^4 +9819703/18953298000000000000x^3−41722579/280789600000000x^2+1154669357/37906596000x+2241989500/3158883

f(10000) = 1000
f(100 000) = 2700
f(200 000) = 3400
f(300 000) = 4100
f(400 000) = 4600
f(500 000) = 5000

here is a graph of the function :

Capture d’écran (835)
Capture d’écran (835)984×803 47.5 KB

1 Like
17

Honestly that seems overly complex, I think we should stick to the original formula as it seems to fulfill our criterias sufficiently.

Vote will be pushed in 24h if no critics arise.

2 Likes
18

Just catching up the discussion, really cool to see good improvement proposals, thanks @Athym & @Baptiste ! :fire:

Agree that the last one might be a little too complex, however imo its great to have more intermediary thresholds.

You mean the one in the initial proposal or the initial improvement formula proposed by Baptiste ?

2 Likes
19

f(x) = sqrt(x^(1.27718))+641.607

This one

1 Like
20

Thanks ! Is it calculated/distributed on a monthly basis or otherwise ?