Open discussion for IISS enhancements

#5

Hey @thelionshire I think I can answer those questions.

1.) The range has not been meaningfully modeled out yet. I can say we’ve never discussed anything higher than 10%.

2.) This also has not been modeled in depth. This part of the list of considerations is a relatively newer part of our discussions

3.) Yes

4.) The size of the contribution fund depends on where we end up getting the funds from, but we were thinking this would be adjustable via a Network Proposal based on the necessity of funds and the price of ICX at the time

2 Likes
#6

Thank you. Also had same question as Brian: for the bond would this be self staked / voted and also would it be able to be stored on ledger?

1 Like
#7

Thank you Ben. It’s great that the ICON Foundation has put much thoughts into enhancing IISS.

  1. Bond Requirement: This is a fantastic idea. This will encourage the P-Rep to commit to ICON long-term development. I would also argue it will create a more stable network as larger amount of ICX will be staked. Agree with Brian about having the ability to store the bond in the ledger in lieu of storing in the hot wallet.

  2. Vote Behavior: Rather than incentivizing voters to vote for lower ranked nodes, ICONex should provide a message display on the Voting page, clarifying that the reward is the same whether you vote to the main P-Rep or Sub-Rep. There are many investors who I had spoken with are misled and thought they would receive more staking rewards by voting to the Main P-Rep in lieu of the Sub P-Rep. ICON Foundation could build a recommendation tool to recommend the top 10 P-Reps with active nodes (This could be random or constructed).

In addition, in order to encourage voters to stay active, the reward system should behave like a decay log chart. For example, voters who vote/re-vote within 7 days will receive full reward (100%). However, if the voters are inactive longer than 7 days, their reward should be reduced based of the decay log with 100 days they receive 0 reward.

Contribution Proposal System: Agree. This system will demand active participation from the Main P-Rep. What happen when the Main P-Rep does not vote or abstain.

Contribution Allocation: Agree. Current reward structure is very linear. It would be great to reduce the reward for for higher-ranked P-Rep and use the delta to fund the “Contribution Proposal Fund.”

2 Likes
#8

I think these are all great ideas, I have one question regarding the “Bond Requirement”.

If a P-Rep doesn’t have enough bond would the voters receive also less rewards? It might sound a little too drastic but it could also motivate even more that the P-Rep needs to have the entire amount required for the bond or they will lose voters.

1 Like
#9

It might result in chicken and egg problem.

1 Like
#10

Vote decay is a common suggestion but I disagree with the expected outcome. I don’t see why somebody would not simply come back and vote for the same node or another arbitrary node. There is absolutely no incentive to do research for the average ICX holder. Even if we provided one bullet point for each team, that’s 100 bullet points to read and analyze before voting. People don’t spend this much time

As for providing better information to voters, I agree this should be done but also disagree with the expected outcome. The people who actually care about who they are voting for is a much smaller fraction of the ICX holder population. This information will help some people make an informed decision, but most people are voting for the economic incentive imo. I’m not saying to neglect that small fraction, just that the suggestion will not solve the problem we are currently discussing

3 Likes
#11

A few thoughts.

Bond Requirement: Seems like a good way to incentivize reward hoarding by P-Reps. Especially if the bond is anywhere near 10% as Benny_Options mentioned. That’s way too high. I think this would empower self staking whales more than anyone else. Don’t really see how it fights vote buying either. I don’t really buy the “many voters are scared of being burned” argument. If they are - they shouldn’t be. As most of us know, it would take hundreds (or even thousands) of consecutive down days to fall below 85% productivity and activate the 6% burn.

Voter Behavior: This one feels a little bit backwards. In my opinion, by doing this you would disincentivize contribution and reward passiveness.

Let’s be honest, in this kind of system, most people will vote to get more gains. They won’t look at contributions. I see this as a pretty huge handicap for any new up-and-coming, contributing teams who would be slammed back down as soon as they touch Top 22 because Sub-Preps nodes offer bigger rewards. I like the idea of improved vote distribution but I think the focus should be on getting those votes out to the few Sub-Preps who are actually contributing. I understand that’s the tricky part.

5 Likes
#12

Bond Requirement: Mixed feelings about this one. It won’t influence teams that self-bought in at all, even favors them above everyone else. Set it too low and it won’t matter. Set it too high (above mentioned 10%) and you’ll have teams hoarding ICX nearly for a year to comply with that, where some of them might even freeze their activity completely until they reach it.

Overall, I think teams should use rewards they get to grow the ecosystem. Hoarding coins is the opposite of that. If they can’t use them, then there’s no good reason to mint those coins in the first place.

Voter Behavior: This could just result in shifting the voting pattern from the top few to bottom few. E.g. I want to get rewards, but don’t want to bother with anything – I’ll vote for the last P-Rep, because it’s the least likely one at the moment to become a main P-Rep and lower my earnings.

Contribution Proposal System: Sounds good. It makes it easier both for voters and projects. Seems that bootstrapped forum CPS paid off.

Contribution Proposal Fund: Why “set amount of ICX available on a monthly basis”? Shouldn’t it accrue whatever excess there might be?

Contribution Allocation: I’d prefer to see the total budget for P-Rep rewards slashed by the intended contribution allocation and the rest distributed more evenly than it is now rather than to tax the top few.

I also hoped this would include increase in Main P-Rep count and better reward distribution across all P-Reps that was mentioned months ago.

4 Likes
#13

Hello @bwhli,

In terms of the alternate wallet address, it looks fair point. We are discussing internally to implement key dualization. This function will make two types of keys, the key for consensus and the key used for governance.

3 Likes
#14

Voter Behaviour:

In my personal opinion, I don’t see how restructuring the reward rates would benefit ICON, PReps, voters and more importantly, a healthy ecosystem.

Scenario A: PRep X ranked #53 is a sub-rep who’s a whale and self-stakes with no intentions to contribute to the project, has never engaged with the community or has any interest in ICON’s developments but happy to run a node, take in the rewards to sell to pay for his lamborghini (thus creating inflation & sell pressure)

Voter B decides to vote for PRep X because he doesn’t care who he chooses to vote for, as long as it’s a sub-rep, he’s happy with his higher rewards.

Instead, it would be nice if ICON considers promoting sub-rep teams accomplishments on ICON’s website or the icon.community dashboard or a newsfeed integrated to the ICONex wallet staking interface > A weekly promotional billboard flashing ‘ReliantNode presents a desktop wallet with staking > myiconwallet.com’ #vote #spread #disrupt

New ICONists/investors visiting any of these websites or accessing their ICONex wallet to stake/vote, will instantly be introduced to a sub-rep team listing their accomplishments or published articles.

It may be extra work for the ICON team to publish/upload a new sub-rep team each week but it’s a win win for everyone!!

4 Likes
#15

I feel like we’re discussing solutions without first finding consensus on what the issues are. Could we please officially agree upon/document what the problems/limitations are so as we can all collectively focus on solving them. I’m sure ICON has done this internally, but for the community and P-Reps everyone has their own ideas about what is working and what isn’t. This should be achievable in a very short timeframe (ie a few days/week) and I believe it would help us all to come up with productive solutions together.

I also want to voice my personal opinion that currently the redesign is being tackled in a very centralised manner. We’ve had an IISS upgrade forum topic here for a very long time, until today ICON hasn’t commented about it publicly. We know that the team has been working hard behind closed doors, but there needs to be more transparency. There has been considerable public criticism of top P-Reps lately, yet they have been left out of the discussion process until today.

ICON also has the expertise, control, and voting weight to make unilateral change, so the reality is P-Reps and the community actually have zero leverage or a binding influence on policy here. We’re only able to make suggestions and work together, so let’s actually work together as efficiently as possible in our shared endeavour.

…

To my mind there are a couple of obvious issues:

  1. A lot of voters/votes have been placed on the top teams listed in the voting list by default.

  2. Voter apathy. There is little reason for passive ICONists to revote/move their votes.

  3. The ability to revote is not technically possible, yet.

  4. The voting interface needs upgrading to address people that vote for the first teams listed.

  5. Velic - A top P-Rep who doesn’t interact much/if at all on governance or with the community.

…

I like the bond idea.

I think giving higher rewards for voting for subpreps is a VERY BAD IDEA. It does not address the root cause of the problem and will lead to people voting for passive nodes. I think Parrot 9s suggestions for improving the voting process were far better.

My number one wish is that we all need to think outside the box on how to achieve a revote once we have an improved voting interface. I refuse to accept that it’s not possible solely because it interferes up block production. We could have ICON ready to take over block production via 22 newly created nodes whilst we all voluntarily take down our nodes and start again from zero. There could be an exciting marketing campaign put behind this to let the community know they have to revote, and we can provide additional financial incentive to encourage a revote within the first few days/weeks.

Let’s not let our current technical limitations limit our creative thinking.

I’m very excited for the future and in compete agreement that we need to upgrade things. Thank you everyone for your contributions! I’m extremely proud of the ICON Foundation, our community and my fellow P-Reps. :facepunch:

5 Likes
#16

What if the main p-rep teams had to chose a “running mate” from a team outside the top 22. They would have to share a certain percentage of their rewards with this team. Therefore they must research a good team that is actively helping the ecosystem. This could start highlighting some of the work of the other p-rep teams which, honestly most people really don’t have much time to research.

This would also show which top p-rep teams are doing their due diligence and choosing quality partners.

1 Like
#17

Bond requirement: P-reps should be accountable for their actions, what will the bond be used for? Will, there be a possibility of ICX in this bond being burned if P-reps actions are hurting the ecosystem? Who will decide on that? Also, the requirements for the bond shouldn’t be high at the start or it may completely halt the progress of P-reps. Maybe a gradual increase to ease the system in. If we put 10% of delegated votes the lowest main P-rep would have to provide 250.000 ICX into a bond which is a huge amount to start. Also, what would happen if a P-rep got a lot of support for the work that they have done and suddenly they have a lot of votes dedicated by the community? Does that mean because they don’t have funds to allocate to the bond that they shouldn’t receive rewards?

Voter behavior: Incentivizing people to vote for sub P-reps is not necessarily better for the ecosystem. If something like this is implemented I’m sure there will be sub P-reps promoting themself without contributing by stating “you get this bonus if you vote for us as a sub P-rep”. People who don’t do a lot of research on P-reps are familiar with the ICON Foundation and they think if someone is a top P-rep they can be trusted which concentrates the votes. The way ICONex is designed by listing P-reps according to the number of votes doesn’t help with this. If you look at similar situations on some other applications like google search or amazon, people rarely scroll to the 2nd page and they usually buy stuff from the first few listings. Education of iconist is important here as well as providing information in a manner that doesn’t take an iconist to much time to keep track of what’s going on with all the P-reps.
I think the topic here should be how to incentivize iconists to be an active voter instead of leaving their votes allocated to a single P-rep of course if a P-rep is doing a good job those votes shouldn’t be reallocated. So maybe some sort of system that incentivizes taking another look at the P-reps you have voted for when you claim your I-score. (I’m just thinking on the fly here, I’m sure there are better solutions to this but it’s just a basic idea).

CPS: I’m looking forward to the CPS paper, but generally support it with maybe some minor adjustments like the contribution allocation posted by blockmove. No need to have ICX rewards going into the hands of top P-reps just for them to send it to the fund. Isn’t that what blockchain is about, cutting out the middle man?

3 Likes
#18

I like the idea of the re-voting, mimic a similar democratic voting style. Maybe we should implement a re-voting annually.

3 Likes
#19

Higher voting rewards for sub P-reps can have a negative impact on the ecosystem
There may be many sub P-reps that do nothing but simply receive a lot of votes and share rewards.

Bond as an option - a good idea, but there are a lot of questions about that.

1 Like
#20

  1. Bond sounds interesting. I’m for having skin in the game (I myself do). I do have some concerns with implementation and I also do not think it will stop vote buying. I also have concerns that it will foster an environment of hoarding and not spending- it is good to save some rewards but some should be spent on contributing. If even 10% is required, this would be a lot, especially as price goes up, and limit some teams that may not have started with capital. Perhaps if It was less than 5% and implemented as a penalty for having no skin in the game (such as 10-20% less rewards), with a forgiveness period while first getting started, would work? My other concern is how the bond would be verified, and can this icx still be staked? I would not support a system where the bond cannot be stored on a ledger and also staked. But would we need to give up the address and lose privacy?

  2. I echo that I am against higher rewards for lower p-reps just for the sake they are lower. You’ll end up with teams marketting they and only that. Also, if teams gained votes to get into the top 22 and be a main P-Rep, their voters should not have to decide whether to continue support or earn less.

3-4: I am for the contribution system and getting rid of the previously planned eep and dbp. As we are seeing, voters seek rewards and most do not take the time to perform proper research. Relying on voters to determine what DApps and eep’s gain funding could be a dangerous game and end up with a few “cool names” or logos as the top.

I think it’s an excellent use of the p reps to make selections, however, I do have concerns over the efficiency of this. There are several teams in the top 22 that are completely inactive. Can we really rely on them to go through proposal submissions and vote?

3 Likes
#21

From our point of view having developed the first E-Commerce integration for ICON we have yet to receive significant enough vote to pursue further development on our planned initiatives. Once we receive more vote or accrue enough funds from our current rewards we will begin the ICONPay Shopify integration.

We retracted our grant request because fundamentally it should be achievable using the rewards currently being distributed without the foundation having to effectively sponsor additional development.

Self Bond - We agree with this proposal however it would need to be done in a way that the signing key is separate as we originally mentioned during mainnet launch with having a management key and a signing key. It might also be difficult for some teams to have the necessary self bond but that will always be the case to some degree. The other option is to make the minimum fairly low but the amount self staked would create a certain capacity the P-Rep can receive in total votes. For example 100k self bond x 10 would allow 1 million vote, to receive more vote the producer needs to increase self bond.

Vote Behaviour - The proposal to increase rewards to voters to vote for lower ranked nodes would at least help average out the vote and producers who do actually want to contribute will be able to do.

Contribution rewards if they exist should come from the existing rewards these are already very substantial however they do not necessarily get allocated to strategic development. From our perspective we had to fund much of the ICONPay development at a loss and only recently broke even based on our current rewards. We look forward to an improved DPoC where voters will get additional rewards if they vote for teams who actually contribute, there is very little incentive to vote for anyone different if you get the same return and as a result teams outside the top >22 are not able to contribute as much as they might like. There is no easy way to solve this problem at the protocol level alone as arbitrarily giving voters more to vote for less ranked node does not ensure they contribute.

A possible solution would be to have a quarterly vote, each P-Reps could vote for 1st(3 points), 2nd(2 points) and 3rd(1 point) of who they believed contributed the most in that quarter, each P-Rep would have equal vote (as per a vote in Parliament irrespective of there weighted vote in the network). The top 10(or X amount) would receive additional rewards, 50% shared with there voters. Voters would then receive more rewards by voting for teams who really contribute. It is not perfect but probably better then what is currently implemented to spread vote and ensure real contribution occurs.

3 Likes
#22

I think this idea merits discussions and research. Great idea Metanyx! P-reps can see day in and day out what is being done. This would encourage voters to vote for performing p reps. While still somewhat biased, i think this could be very powerful. Also, if it was just a portion of rewards, It would not overhaul the system but would be a neat addition. Lots to think through and consider but I’d love to hear the foundation’s thoughts on this.

1 Like
#23

My position on this matter may differ from everyone’s because I’m not evaluating the proposed changes based on their merits. They do seem well-thought-out and it’s clear that the good of the network is the goal.

But I think it’s important to remember the actual problem we are trying to solve:

———-
Ubiquitous voter disinterest, complete lack of participation and complete lack of attention.
———-

Any incentive change that is still contingent on voter attention and participation may not work well and will likely not generate the behavior changes we are seeking.

Over the last 4 months we learned that the vast majority of voters vote once, and never return.

Without a powerful incentive to re-vote, they simply never will. And as long as they continue to receive rewards without any consequences, they have no reason to change their behavior.

Do the proposed changes fix the real problem, or do they just introduce a new set of rules that assume voters will both participate and take interest? If the latter, we may be missing the mark.

—————

For this reason, I still believe that ideas like incentivizing a periodic revote by expiring voter rewards could help solve the problem.

And it does so in a way that requires no major changes to the system.

If people aren’t getting rewards, they are guaranteed to return and re-vote. Maybe the votes don’t have to expire. But the rewards could.

And there’s only one thing to know: you have to come back periodically and vote. Simple. Basic. And a fool-proof incentive mechanism.

If they’re presented with a simple list of team accomplishments (or lack thereof) it could go a long way to encouraging more intelligent voting.

While I realize there are specifics that would need to be ironed-out to perfect this approach, it’s the one I lean towards. And what I think should be tried first.

In summary: Let’s not forget what the real problem is.

2 Likes
#24

Yet we expect them understand the new incentive structure and react to it the way we hope?

Both approaches require education. But to me, skimming a bulleted list requires vastly less.

If one genuinely feels that voters will do zero research and zero learning, then one must also believe that the new set of rules being introduced will fail.

A simple sentence explaining the importance of voting for active teams is more likely to work, because they’ll finally be aware of who is contributing. Assuming we present that information as simply as possible.

——
I understand the point that people aren’t going to read all the bulleted items in depth. And I agree. But do we just not tell them what the teams are doing?

The entire system is contingent on team contributions. And the entire system relies on the voters knowing what teams are doing.

A bulleted list is the most simplified way to do that. One can technically find issues with any form of information-conveyance if they really want to. At the end of the day, voters need to know what the teams are doing.

———

Let’s look at a real world example:

If I’m looking down a list of bulleted items and I see 5 to 10 words over and over, describing things that seem meaningful to me, I’m going to quickly digest what each team is doing.

I’m also going to see teams on the list that have done a little or nothing. That’s extremely meaningful. Currently were providing zero information and that is obviously unacceptable.

Nobody is going to dig into every single bullet item and research them all. To imply this is necessary is incorrect, imo.

The benefits of having a list of accomplishments is obvious.

2 Likes