Distribution of rewards method for PoS seems to be undecided as of now - should the community have a say?

From: https://www.adatainment.com/_downloads/docs/delegation_design_spec.pdf

Page 47-50 discusses the ruled out approaches and feasible approaches for reward distribution.

Ruled out:
A.2.1 Manual Sharing
A.2.2 Automatically Issue Transactions Each Epoch
A.2.3 Let Members Collect Rewards

Feasible:
A.3.1 Automatic UTxO Updates
A.3.2 Lotteries per Stakepool
A.3.3 Reward accounts per stake key

My own comments:
A2.1: Agree terrible approach to manually distribute rewards from the stake pools themselves as that creates a lot of potential trust issues and could lead to media stories like “staking pool owner ran away with stake rewards.”
A2.2 Agree bad approach as super-linear growth.
A.2.3 Let Members Collect Rewards: Yes this creates a need for storing a lot of pool history until members collect rewards, so seems a bad approach.

A.3.1 Automatic UTxO Updates: Bad because it could be perceived as effecting immutability and as a community member I believe this is one of the most vital things. Trust in the immutability is a big big deal. It could be much easier to hack a node code and create a weak spot in the whole system.

A.3.2 Lotteries per Stakepool (per epoch it seems by the wording of this)
So with 73 epochs a year I assume this means 73 chances a year to win the lottery inside a pool and for a community member to receive its stake relate to its staking proportion. This seems like a bad idea as it would create the possibility of users going without a reward for a long long time in bigger pools.

A.3.3 Reward accounts per stake key
Bad idea as the last paragraph suggests:
“• Unless people stick to a single staking key (which would immediately mean they give up
all privacy, not a choice most people would be comfortable with I suspect), we basically
end up creating lots of staking keys, to which we would only deposit once, and withdraw
from once – in other words, we’d have reinvented UTxO entries, and the accumulation
does not help.”

Question: Why cannot a pool have a single wallet that receives the rewards that all users can collect its share of the reward including the pool owner (A variant of A2.3 but different in that the wallet itself is on the ledger so no chance of no immutability and also that there is no need to store a cache as the pool has an address itself. The user then is the one who updates the node when collecting the reward updating it has collected its part of the rewards. There needs also to be some interface for paper based wallets for this as well but should be possible.) You reduce transactions to number of pools and you remove the risk of nodes tempering with any information and you let users collect the rewards at users leisure - perhaps also with a minimum amount withdraw to remove the problem of dust.

In any case the fact that the way rewards are distributed depend on the users trust in my mind means that the community needs to be involved in this discussion.

1 Like

Distribution of rewards method for PoS seems to be undecided as of now

Everything is decided

the way rewards are distributed depend on the users trust

Does not depend on any trust

the community needs to be involved in this discussion.

nope

5 Likes

Ok. What is the decision so there can be some clarity?

Several places in the design document this is mentioned Vantuz. Perhaps you are referring to a more strict “Trust” than what i put in trust (like public perception of a given choice if it is good or bad.)
P.48: “(Public Perception) This may be perceived as subverting the notion of immutability of the
blockchain (at least in its UTxO model)”
P.49: “. An individual stake pool member will on average receive the same rewards as with any of
the other approaches, but it will be much less predictable. This might be problematic from
a Public Perception perspective.”
P.49 “. (Public Perception) although (in the limit) this is the same outcome as sharing, apparently
most humans don’t see things that way15 – they would prefer known outputs (even if
smaller) to unknown ones. An additional indicator of human response might be to look
at a similar mechanism (random rewards for depositing a fixed stake) has run since 1956.
Premium Bonds16 – computer nerds / crypto nuts should note who helped create the
original ERNIE). The public might like the gambling aspect, businesses might not!”
P.50 “Unless people stick to a single staking key (which would immediately mean they give up
all privacy, not a choice most people would be comfortable with I suspect), we basically
end up creating lots of staking keys, to which we would only deposit once, and withdraw
from once – in other words, we’d have reinvented UTxO entries, and the accumulation
does not help.”

I am sorry I do not agree with you and to think any reward mechanism is not dependent on how it is perceived by the users is naive! (Oh and also given how a good deal of tought has gone into this in the document I believe you are contradicting yourself or the ones who wrote the document)

I disagree wholeheartedly!

Please keep in mind I DIG Cardano. I have invested in Cardano. I am puzzled how this document has been released in its current shape. Either this should have been discussed in public from the start or kept private until a solution was reached. Then again at this point it seems to be in the public ( I got this from a link on the forum in this post and have no prior knowledge apart from what I read from the link pdf) so at this point it is definitively healthy to just discuss it given its public nature. And I only see good in community being involved and having the ability to have a voice on this given how staking reward distribution is both dependent on what is good for the Cardano Network but also for each individual community member. For example a lottery model where a user would risk having to wait for months for a reward would probably kill off any interest in staking. (if I understood the lottery model correctly that is.) So clearly there is a relationship between the two things.

2 Likes

In a more positive note this could mean the document on the forum is just an old draft and does not reflect the current state. And clarity wise it is more important with clarity on the monetary policy and how we should interpret “as of now” in the monetary policy portion than the exact reward mechanism as that can wait for the pool staking period. But I do think it is good for the community to be exposed to these ideas as any nuances for users can only discovered when the ideas are exposed to the users at large. And yes it will be nice for the public to know what is the reward method.

Oh one more thing: The transparency in Cardano is miles ahead of other projects. The fact that I can have these discussion on this board at this type of detail level is something unheard of in many many other projects. I believe that is a great thing.

4 Likes

It’s rili a good question ! Maybe we should rasied it in next AMA or something.

It’s called “speculation”, you can look it up here: Theoretical philosophy - Wikipedia

cool!

It does not imply any trust inside the finally selected reward mechanism itself.

I am puzzled how you would consider finding a draft of a document in a forum-comment left by a community member, and downloading it from a 3rd party website not related to authors in any way as “being released”.

Ok.

Doesn’t change the fact that authors can use GitHub for their collaborative work, and doesn’t change the fact how tons of people are thankful for them doing the work in open, so we can gratefully observe the work process.

And doesn’t imply that public should have anything to do with the process of designing complex consensus protocols.

You are definitely free and welcomed to discuss anything, and maybe try to educate the public on why different mentioned solutions are good and/or bad in certain aspects. I am not trying to forbid the discussion of public research documents, I am trying to correct the wrong ideas from your description of the document, and point out that these discussions have entertaining/educational nature and will not affect the results of the protocol design.

Everyone will know when design process is finished and published. I think it will be nice for everyone to understand that you are taking 3rd party renderings of old revisions of design process documents and presenting to the public your own take on how you imagine the design process to work, which probably has nothing in common with reality.

And I don’t say that it might not be interesting for the community to play a game where they try to come up with their own ideas on how stuff might work, and then compare it to actual results and try to figure out why their solution would not be selected.

My points are simple:

  1. Reward distribution method is finalised (and will be available in actually released specification)

  2. When you say “the way rewards are distributed depend on the users trust” - other readers might think that it implies they will have to trust other people with their wallets or rewards, and this is not the case.

  3. People might think that you know something they don’t and that when you say “the community needs to be involved in this discussion” you mean that protocol developers actually require some sort of a public vote in order for the protocol to be finished, and this is also not the case. Protocol designers apply their deep scientific knowledge and decades of experience in cryptography, formal methods, game theory, and engineering in order to come up with what they think is the best, secure, fair, trustless, and sustainable solution for the presented problem. And they happily take and consider input from contributors, when it’s presented in a competent way thru appropriate channels.

I just point out my corrections. Does not mean that entertaining discussion might not continue.

10 Likes

Hello,

I always considered this doc as draft. Whenever I shared it directly I emphasized that, like for example @RobJF can confirm.

But I have to admit that in the quoted post it may was not clearly enough. That’s why I edited the post with a note on that.

Anyway it’s interesting and worth sharing.

4 Likes

This document was shared by Charles in the TG channel. It’s 50 pages as he referred to. It was the right document to review by Eystein.

1 Like

For example, if you want to be a researcher for a bit - one of the most important things researchers do is falsifiability.

In the specification draft they have described multiple possible options that might be used to solve the problem, and they have carefully provided pros/cons and also described how non-researchers might perceive
each solution. This is part of being a researcher.

You have provided your own idea of a possible solution. Can you now try to come up with pros/cons and possible reasons why this option might not work or might not be the best solution? This might be fun.

3 Likes

Does not make it not a draft.

3 Likes

Of course it doesn’t!!! :)))

It was shared on TG before he travelled to Berlin for the review.

But let’s give Eystein also some credits as his comment on the document’s state is absolutely valid and he might have made the fair assumption that Charles does travel so long not for an early draft review, but more a final draft review especially considering that Shelley is scheduled to be released in Q1.

1 Like

Never “underestimate the Community”, there might be some cryptographers, game theory experts, mathematicians or just amazing minds among 50k+ of us, who may anytime come up with some brilliant idea, which would make sense to be included in the protocol design. Likelyhood is very low, but let´s not rule that out and always consider utilizing the power of the Community.

You are absolutely right we should praise IOHK for this level of transparency and allowing such level of insight into their work process & draft work products. This is truly exceptional and highly appreciated. Most of us believe in Cardano among many other things because of this unique level of transparency and openness.

2 Likes

Modern research is a collaborative process, not something done by lone geniuses. There’s no way to know that an idea is brilliant until it’s been thoroughly discussed and tested. And participation in such discussion and testing makes the researcher one of the insiders already, an outsider has no chance. Having worked both sides of that fence I know what it’s like, and I understand the reasons for it. Allow the insiders to get on with it.

2 Likes

Thanks for your reply Vantuz. Yes its speculation regarding Public perception. My point is simple if community is engaged we can give you a better idea of Public perception.

Regarding trust I think we simply are talking about different things as I ment a broader trust where public perception is part of it. But yes I agree with you on what you are talking about as trust in this regard. And I can see I should have worded myself better in this so fair point.

Regarding the document it was not clearly described in one way or another. I just think it is better for Cardano if community members instead of reading speculations on Public perception could have a say in public perception.

You have a point on GitHub. If this was how the document was used I understand and it is correct. And I agree wholeheartedly with you it is great it is public. But the status of the document and where it was in the process was not clear. We as the public only have a say when it comes to public perception and should not dabble into the technical aspects of experts and owners of design (unless competent and through appropriate channels as you point out.)

Regarding wrong ideas of description what in particular are you referring to as you seem to have a wrong idea of what what I meant with my description as well. For example in how I am looking at trust from a community perspective while you are discussing trust on a reward. Regarding you comment that these discussions have no effect on the results of the protocol design this makes me sad. When community perception is important the community should be engaged. You risk the community giving you a thumbs down when these things are put into actual implementation if not. So I stand by my comment I think this is naive. But each to his own, transparency can be used to further a design and it can be used to just inform the public. Choice is of course the ones who design to make. But I believe community engagement simply is a better path.

Regarding taking 3rd party renderings of old design process this is absolutely important to know. As I said in another post I have only the knowledge of the link and I appreciate to know more about the status.

So to your main ending points:

  1. Great. I am happy to know this and I am sure others in the thread are as well.
  2. This was not what I intended and I think it is pretty clear from what I am discussing this is not what I intended. But it is fair you address these things. But you seem to have a mental state of thinking I want to write negative things when it is far from it as you interpret what I write in the very worst ways. I think when a document like that is thrown out in the forum it is fair to discuss it and discuss status of said document. Now we have a clearer view of the status and I think that is a good thing for everyone.
  3. I think community should be involved in the aspects that are part of public perception. Especially when such things are speculated in a document it would be a better path (but of course this is just an opinion) to simply engage the community in such discussions. When it comes to the experts in cryptography, formal methods et al I agree as well. Of course this should be left to the experts.

Alright fair discussion and thanks for clearing some things up.

3 Likes

Well given 2 master degrees and heavy focus on research with top grade and consistently over time and in different disciplines in particular in such subjects (quantitative and qualitative research, statistical methods etc) yes I know a thing or two about falsifiability so no need to lecture me on that. (Then again I do not use falsability in computer science so you could lecture me on that as I always want to learn more ^^ but from a theoretical standpoint no.) I have not seen many black swans as of late. I only mention this since your tone seems to be I have no knowledge of the subject.

And no I do not feel competent to provide falsability for such an idea as I have not the competence to look for potential pitfalls. For example I have only a shallow understanding of what dust entails. I have a shallow cryptography knowledge. Ask me on law or psychology and I can give you a more detailed description on what I know and what I do not know (incidentally for me expert knowledge is also being aware of what you do not know.)

With no understanding of what the intent was with the document and how it several places mentions public perception I felt it was an OK thing to do to come with suggestions as well on another approach. I used basic logic and copying methods described and circumventing the cons already described in the document while trying to emphasize the pro’s in the document as well. But yes I agree with you this should be left to the domain experts. But it seemed to me a better approach and easy to combine approach between what was already suggested in the document.

2 Likes

I agree. And when it comes to public perception aspects the public perception needs to be tested as well. You can have a great idea on how you think people will behave but it is far better to try it out with models and get actual data if your theory is correct.

2 Likes

Every part of Cardano is built on Peer Reviewed Research Papers, which have gone through the most rigorous, high assurance process to be finally accepted by globally acclaimed, external 3rd party subject matter experts and presented on prestigious conferences.

This paper might not really fit into that category due to it´s nature and purpose and guess it will not go through the standard academic Peer Review process (?), why Eystein might be right that a Community Peer Review via public consultation is a very valid approach in this particular case, also considering the Community “owns” the Cardano venture and should also take some “best effort” responsibility in such situations.

1 Like
  1. Game theory is well tested, and minor discrepancies can be adjusted using parameterization, which is built-in here.

  2. Ideas and speculation of community members regarding people’s behavior are certainly no better, probably much worse, than those of professionals in relevant disciplines.

1 Like

Oxymoron: community are not peers.

  1. Great yes but you cant suddenly change (Well you can but it would damage your standing potentially in the community) from one reward mechanism to the other without a clue on what you think the community prefer. There has been no game theory on this at least not in the document provided so I do not see your point. I am not saying we should decide. I am saying I hope there are other methods employed than speculation.

  2. Agree. That is why you need to move on from speculation. As is what was described in the document or atleast what Vantuz said it was here. There are many tools for this as I am sure you are aware as well. Cheif among them is collecting data. Even game theory models need to collect data and adjust parameters as you point out. Models are always models, not more, not less. You need to try them out and collect data. Do they describe reality in a good way? Are they a good fit to my hypothesis?

1 Like