Increase Proposal Rejection Cost

I totally agree with this. We have to rely on this irresistible truth :

Faking or believing it is not the case will make us less efficient and slower.

I think this idea has a lot of potential and should be further deliberated within the community, particularly in context with the increased proposal cost. If the proposal costs become prohibitive for NNS participants who fear they would lose the ICP on a good-faith proposal, then this would be a good remedy for that. There’s good arguments for us to consider this within the context of the anti-spam proposals, or to consider it on its own merit if the proposal cost increase is approved.

1 Like

Thanks @wpb. We can continue with this one. I think there has been a lot of good discussion and I would hate to lose that by starting another post. Rather than recreate the post I will just clarify my reasons for supporting this proposal.

  1. Do I see this as the final solution?
    No. I consider this to be a stop gap measure that will deter further exploitation by parties who simply want to increase their daily rewards, advertise products, and/or spread disinformation.

  2. Why not wait for a better solution to be implemented?
    Let’s consider the following steps that have to be completed in order for a better (complex) solution to be implemented:

  • Community Activities
    – Brainstorming (1 week)
    – Community deliberates on a proposal (1 week)
    – Vote on a proposal (1 week)
    – Appeal to Dfinity for support developing the solution (1 week)

  • Dfinity Activities
    – Dfinity drafts a design ( 4 weeks?)
    – Dfinity submits design for review (1 week)
    – Community votes on design (1 week)
    – Dfinity plans & schedules development/testing effort (1 week)
    – Indeterminate amount of lead time (months? - quite possible and understandable)
    – Dfinity develops the solution (2 weeks)
    – Dfinity tests the solution (2 weeks)
    – Dfinity submits code update to NNS (1 day)

    So based on that very rough estimate, we are looking at anywhere from 3-6 months (optimistically?). During this time the NNS will continue to be spammed and I believe this will cause stakeholders, and those looking to participate in governance, to lose respect for the NNS and become complacent in their voting habits. Basically, I think we will see less manual voting and more passive voting over time.

  1. Why does this proposal not suffer from the same delays?
    Because the proposal type for increasing the cost already exists. We wouldn’t need Dfinity’s assistance in developing anything. Although, I will likely need some minor support constructing the correct dfx command.

  2. Am I concerned that this will deter meaningful proposals from being submitted?
    I think the term “meaningful” can have different connotations depending on the individual. For me, a meaningful proposal is more than just a good idea. It’s a proposal that has already been deliberated amongst a community of stakeholders (not necessarily on this forum) and already has some backing. If those conditions are met then I do not see 10 ICP getting in the way of that community submitting the proposal.

  3. Would you support a proposal to reduce the cost when a better solution is implemented, or if the price of ICP increases substantially?
    Absolutely. I believe the proposal type exists exactly for that reason. To be adjusted as-needed.

Edit: I have zero insight into Dfinity’s planning & scheduling. These estimates are based purely on my own observations over the last year.

5 Likes

I think another idea, similar to @plsak is proposing, would be increasing the ICP cost per rejected proposal. For example, the first rejected proposal would cost 1 ICP, then the second would be 2 ICP – you could scale this in various ways (e.g., 1 ICP, 2 ICP, 4 ICP, 8 ICP, etc.). This would keep the barrier to submit proposals low for new individuals allowing for people to make mistakes (e.g., submitting a blank proposal) and submit new ideas.

This system could also reset if someone gets a proposal accepted, thus, promoting quality proposals (i.e., someone has two proposals rejected and their third accepted; following this, their fourth proposal is back to 1 ICP). This could also be time gated (six months?) where your proposal cost resets back to 1 ICP after a certain amount of time.

Also, it can be combined with the current flat increase proposal but allow for the increase to be smaller.

One potential down side that I can see is that you could somewhat evade this by creating new accounts (minimum account age to submit a proposal?). I think maybe a mix of this plus the flat increase could make sense.

3 Likes

Long term a more detailed layered approach is needed to address the spam and quality issue.

But for short term practicality this proposal can deliver. After this we will have time to discuss other long term structural solutions.

I agre with this proposal as it is!

2 Likes

Daniel, could you review this please ?

Hey @LightningLad91 do you want any information in the first post changed to better reflect your thoughts? I’d be happy to updated as per your preferences.

Also, do you think deliberation on this is proposal has gone quiet? I know there have been comments, but do you think any of them are in scope for this proposal or are they ideas that need to be deliberated as separate proposals? If you think the conversation has gone quite on this proposal, then I would support you moving forward with making this proposal.

1 Like

Yes I think so. Seems like we can move forward.

@plsak @eluemm I think you’ve both proposed some good ideas and I recommend starting separate topics to discuss further. I’d be happy to spread the word if you choose to do so.

2 Likes

I think 10 ICP might be a bit too much, what about 5 ICPs?

I think 10 ICP might be too little. We are proposing protocol level changes it shouldn’t be easy. I say 50ICP.

2 Likes

It’s not easy at all, there is no guarantee the proposal will pass, the difficulty should be achieving majority not getting the funds to do a proposal.

I agree with the spirit of the proposal but I would price it in cycles to make is stable

Is it possible to add a third option, to signal a spam or meaningless proposal where if majority people vote for it as opposed to “yes” or “no” the penalty is higher than just rejection?

This would take the penalty off of mere rejection as well, I assume it was initially there to prevent spam

Thank you @Kyle_Langham and @LightningLad91 !

I was just going to start a new topic with this update (change Rejection Cost to Spam Rejection Cost) but see that the proposal to increase rejection cost is already out, so let’s wait for the result first:

  • Note, that my personal (strong) preference will be then the check box Is Spam in every proposal, due to increase in administration with new Spam Filter proposals.
    • To have new Spam Filter proposals might seem manageable now, but when SNS is released the governance work will multiply, so simplicity will be very important here.

Also note that if current proposal is accepted, the 10 ICP Rejection Cost will be a blocking factor for me so I won’t be able to propose this (or other) change (until Rejection Cost will be changed to Spam Rejection Cost) and will rely on better ICP situated community members with that.

2 Likes

I thought we were waiting until the 18th?!

2 Likes

Agree with this short-term. Maybe a good long-term idea is to have rejection cost adjust to ensure a target percentage of failed proposals? If a large proportion of proposals are rejected increase rejection cost, and vice versa.

1 Like

@MrPink13

The original proposal included the following language before I took it over:

@wpb and I had observed a slowdown in discussion (before today) so I decided to take it to the NNS.

Personally, I think this proposal should be considered separately from the group of proposals being submitted on the 18th. My reason being that this is a temporary measure meant to reduce the amount of spam we’re experiencing while more long term solutions are developed.

2 Likes

@namyIC recently suggested a similar fix on Twitter. I would consider reaching out to each other to create a forum topic where we can discuss this proposal.

I want to suggest a unifying reward formula for proposal adoption/rejection:
Proposal Reward = (Adopt% - Reject%) * 10 ICP.
That is, if Adopt gets x% of the total voting power and Reject gets y% of the total voting power, then the proposer should receive (x% - y%) * 10 ICP as rewards, which may be either positive or negative.

  1. The simple majority rule should be applied here. Requirement of at least 3% of the total voting power is unnecessary (To see this point, you should deeply understand the principle of neutrality). The absolute majority rule is also unnecessary (It ignores the significance of abstention). Someone would say that x% - y% = 0.001% means nothing, but I totally disagree: Every vote counts. In reality, we may need a minimum requirement, but that is only because counting errors may exist. Obviously, in NNS, no counting errors can exist.
    (According to modern social choice theory, there are no best voting rules because of Arrow’s impossibility theorem. But in NNS, the simple majority rule is indeed perfect: it satisfies anonymity, neutrality, Pareto principle, and positive responsiveness, and there are only two alternatives (adoption and rejection), implying that Arrow’s impossibility theorem does not apply anymore.)

  2. If x% - y% > 0, then your proposal is good and should be rewarded with (x% - y%)* 10 ICP; if x% - y% < 0, then your proposal is bad and should be punished with (x% - y%)* 10 ICP. I feel very strange that the adopted proposals are not rewarded. This seems very unfair to those people who submitted so many good proposals. I think it is obviously a good way of incentivizing people to submit good proposals to improve ICP’s governance. (See my another thread: The Declaration of an ICP Hodler)