Further decentralize the set of LKSM-nominated validators by replacing 4 P2P.ORG validators with 4 well-established independent validators.

Thanks for making the proposal. Do you have any stats indicating how are you selecting which validators to remove and which to include? I just want to make sure this is unbiased.

I do agree we need better validator selection mechanism and happy to see the community is helping on this area.

@xlc Thank you for your reply.

Being that this is my recommendation it's impossible to remove my bias - I obviously have a vested interest in this proposal as it recommends my validator among the 4 to replace the P2P validators. I suppose I'm an advocate for Thousand Validator Programme validators hence the reason why I picked those four individuals.

In any case you're right, we should examine the facts. Let's take a look at the current state of LKSM nominated validators:

+-----------------------+--------+----------+------------+--------------+------------+
| Stash                 | Active | Inactive | Commission | Active Stake | Self Stake |
+-----------------------+--------+----------+------------+--------------+------------+
| DSbh... (WEI)         | 30     | 633      | 0.10%      | 4,149.32     | 107.3435   |
+-----------------------+--------+----------+------------+--------------+------------+
| G7Ur... (Stake.Zone)  | 48     | 1145     | 0.00%      | 4,149.29     | 250.8581   |
+-----------------------+--------+----------+------------+--------------+------------+
| DzSj... (P2P)         | 5      | 72       | 0.75%      | 4,159.20     | 2          |
+-----------------------+--------+----------+------------+--------------+------------+
| DQX3... (P2P)         | 8      | 55       | 0.75%      | 4,148.59     | 1          |
+-----------------------+--------+----------+------------+--------------+------------+
| DMYL... (P2P)         | 9      | 70       | 0.75%      | 4,148.98     | 1          |
+-----------------------+--------+----------+------------+--------------+------------+
| J4GL... (DotScanner)  | 5      | 115      | 3.00%      | 4,149.38     | 70         |
+-----------------------+--------+----------+------------+--------------+------------+
| CsKv... (Polkachu)    | 34     | 677      | 0.00%      | 4,148.69     | 6.7821     |
+-----------------------+--------+----------+------------+--------------+------------+
| CczS... (NodeEasy)    | 66     | 602      | 1.00%      | 4,161.45     | 50.1       |
+-----------------------+--------+----------+------------+--------------+------------+
| GcwN... (KeepNode)    | 22     | 119      | 2.00%      | 4,148.36     | 1          |
+-----------------------+--------+----------+------------+--------------+------------+
| DB2m... (Jaco)        | 3      | 88       | 2.00%      | 4,150.41     | 1          |
+-----------------------+--------+----------+------------+--------------+------------+
| G7qf... (Jaco)        | 3      | 44       | 2.00%      | 4,148.67     | 1          |
+-----------------------+--------+----------+------------+--------------+------------+
| Hc74... (Jaco)        | 2      | 53       | 2.00%      | 4,147.92     | 1          |
+-----------------------+--------+----------+------------+--------------+------------+
| J9gd... (KeepNode)    | 20     | 118      | 2.00%      | 4,149.13     | 1          |
+-----------------------+--------+----------+------------+--------------+------------+
| Cg5j... (Ryabina)     | 3      | 82       | 3.00%      | 4,149.56     | 0.15       |
+-----------------------+--------+----------+------------+--------------+------------+
| E2pS... (Jaco)        | 3      | 35       | 2.00%      | 4,148.68     | 1          |
+-----------------------+--------+----------+------------+--------------+------------+
| GZTC... (P2P)         | 4      | 62       | 0.75%      | 4,150.77     | 1          |
+-----------------------+--------+----------+------------+--------------+------------+
| DaBh... (BlockDaemon) | 1      | 1        | 2.00%      | 4,149.72     | 0.1        |
+-----------------------+--------+----------+------------+--------------+------------+
| GpyT... (Polkachu)    | 96     | 657      | 1.00%      | 4,148.43     | 5.6008     |
+-----------------------+--------+----------+------------+--------------+------------+
| GLv5... (Jaco)        | 3      | 56       | 2.00%      | 4,148.93     | 1          |
+-----------------------+--------+----------+------------+--------------+------------+
| FkWk... (Paranodes)   | 34     | 384      | 0.00%      | 4,148.67     | 10         |
+-----------------------+--------+----------+------------+--------------+------------+
| HsxG... (P2P)         | 7      | 62       | 0.75%      | 4,147.97     | 5.174      |
+-----------------------+--------+----------+------------+--------------+------------+
| EUuS... (Jaco)        | 3      | 29       | 2.00%      | 4,149.73     | 1          |
+-----------------------+--------+----------+------------+--------------+------------+
| D8jY... (Ryabina)     | 3      | 26       | 3.00%      | 4,149.87     | 0.15       |
+-----------------------+--------+----------+------------+--------------+------------+
| EsNZ... (Paranodes)   | 21     | 252      | 0.00%      | 4,148.52     | 3.0096     |
+-----------------------+--------+----------+------------+--------------+------------+

Jaco holds the most nominations from the LKSM Karura nominator. However he's the highest voted member of the council and a robust figure in the Kusama ecosystem, so it's safe to say he represents a huge portion of token holders, but perhaps nominating 6 of his nodes is overkill. Maybe this is something we can remedy after P2P.

P2P has the second most number of nominations and is one of the biggest centralization culprits when it comes to the Kusama validator set. They are operating 100 nodes (compare that to Jaco's 41 nodes) yet their self-stake is among the lowest.

The rest of the validators are not of concern in my opinion - all reliable node operators who adhere to decentralization principles. They are also not as heavily favored by the LKSM nominator as Jaco and P2P

Now let's look at the four validators that would replace the four P2P validators:

+-----------------------------+--------+----------+---------+--------------+------------+
| Stash                       | Active | Inactive | Comm    | Active Stake | Self Stake |
+-----------------------------+--------+----------+---------+--------------+------------+
| Hf8C... (bLd Nodes)         | N/A    | 9        | 10%     | N/A          | 50.0363    |
+-----------------------------+--------+----------+---------+--------------+------------+
| Ez4A... (Alice und Bob)     | N/A    | 191      | 0%      | N/A          | 56.0025    |
+-----------------------------+--------+----------+---------+--------------+------------+
| GcYj... (Bricksz)           | N/A    | 63       | 0%      | N/A          | 50         |
+-----------------------------+--------+----------+---------+--------------+------------+
| GqC3... (Adam_Clay_Steeber) | N/A    | 32       | 0%      | N/A          | 50.75      |
+-----------------------------+--------+----------+---------+--------------+------------+

Though none of these validators are active, they have a good track record in the Thousand Validator Programme, plus it's very likely they would be elected by the LKSM nomination alone. These individuals are people I've personally seen all over the Kusama ecosystem providing great value to the network in various ways - all people I can personally vouch for. bLd Nodes has 10% commission only because he's struggling earning enough rewards from his sparse elections into the active set and has expressed that he will decrease if this changes.

These 4 validators might not be the "best" but that's a subjective argument since there are many tradeoffs to consider when nominating validators. It's my opinion that the financial benefit from nominating P2P nodes is far less than the benefit gained by introducing more independent entities in the validator set. Karura has the opportunity to make these well-established independent validators relevant in the active set which is something that would improve the security of the Kusama network.

Just my 2+ cents.

Dear All,

I think than to have independent validator is something important for the long term of the ecosystem.

But 2 questions :

• The BlD node is at 10% of commission and I think Karura can't select a node at this % of commission. So to be select it will need to be reduce for me.
• It's request to remove 4 P2P on the 5 select by Karura, and so keep only 1 P2P node.
  But currently Karura select 6 Jaco, 5 P2P, and other at 1 or 2 node.
  Why reduce only on P2P and not also on Jaco ?
  As example -2 P2P -2 Jaco, to have 4 Jaco, 3 P2P, others at 1 or 2.
  I think it will be more logic than to remove 4 nodes at P2P only.

No Risk - No Fun

Hi all!
I will try to state the most objective point, but it will obviously be biased here since I am concerned with our validator being proposed (thanks Adam btw even if we didn't talk about it together before).

About our actual commission:

• The validator bLd Nodes commission was raised 10 days ago from 2% (the usual one) to 10% for reasons we stated here: https://twitter.com/bLdNodes/status/1465455958000345088 .
• Our main objective has and will always be to be in the active set. Of course, we will never run a 10% commission being constantly active, this is why we just lower it back to 2% regarding this proposition.
• In agreement with the principles of the DOT Validator Alliance we co-founded and also the rules enforces by the 1KV program, we will not run more than 2 validator/collator nodes on the same network and always make our best to benefit the whole network and independent validator community to support diversity and decentralization.

About validators commission in general:running a validator at 0% commission is not a healthy way in our opinion:

• Validator doesn't show skin in the game.
• This is not sustainable for an independent who supports infrastructure costs (saying that, we've been running this validator at loss for a year now).
• This brings unfair competition for other validators and as a result centralizes the network and reduces diversity > SPOF weakens security also.
  For more information about validator situation today, please read Paradox's post here where I stated my view for improving the protocol also: https://kusama.polkassembly.io/post/1163

I think with liquid staking is a wonderful opportunity for Acala's community to be an actor of relay chain decentralization and security, I would love to see talks going with Acala team about criteria to select those validators. Obviously profit has to be the main one, but many others deserve to be taken into account in my opinion, especially decentralization and security risks.

About other validators, I don't want to target a specific operator here, especially not P2P which despite being a 'big' node operator is showing involvement with the validator community in general.
My view is that I would like to see the fairest distribution of nomination possible in Acala/Karura liquid staking where nominating one validator per operator (selected carefully of course) would be an engagement for decentralization and validator diversity.

I will second this proposition as this seems an opportunity for us to enter the active set outside of just the 1KV punctual nominations but I would like to see this as an opportunity to raise a voice for other independent validators who are caring about common good of the network besides their own interest.

Cheers.
bLd

Dear Karura community members,

I would firstly like to say a few points in favour of P2P.org (P2P).

1. Anton (Head of DevOps) is a fairly active member in the validator community. He provides good input to conversations and is supportive of other validators, even independents.
2. P2P as with many other larger validators chose to keep own-stake low but establish its share of the risk through self-nomination. As expressed in previous discussions, this offers the validator convenience for management of their nodes. Just a note: it would be great to have P2P establish this risk visually by naming these accounts (P2P.ORG/nom-1).
3. If all of P2P's active nodes go offline we're still not at a risk for a mass slashing. 

I have invited P2P to provide a response in their own defence.

On to the topic:

This particular proposal is imo self-serving and not equitable; it proposes to take away from a single validator P2P and give to a specific four independent validators. The points mentioned that distinguishes P2P from Jaco expresses one perspective, I have (earlier) provided another. As to the benefactors, I am well acquainted with the independent parties but I also believe there are other (or perhaps more) deserving parties. No offense meant to anyone.

Ultimately, the LKSM account should seek to nominate validators based on criteria to maximize profit and reduce risk, to self, or to chain.

I would suggest a fairly dynamic system as follows:

1. Establish a list of parties that the LKSM system will always nominate, this acts as a safety net to ensure that there are always good validators as a fall-back for any issues that may arise.
2. These persons are ones that the council has made prior arrangements with e.g. BlockDaemon or
3. Those the council believes has established a high profit / low risk position by reputation/confidence
4. This list should provide two validators per party at most.
5. The existing nominations of these validators should be taken into consideration. An extremely well nominated validator in this category can have undesirable results.
6. The remaining nominations should be determined by criteria taken from 1KV candidates, this is in favour of reducing risk :
7. Candidate should be valid i.e. up-to-date, no slashes, online
8. Not independently active
9. Low commission 0-2%?
10. High self-stake
11. No faults
12. Favour validators with high backing.
13. This can be done by rotation or the system could always chose the best meeting the above criteria
14. No more than 1 validator per person should be nominated.

If a sufficient number of validators in 2. can not be obtained, then those in 1 should be large enough to service the need with out loss to profit. At present staking levels and nominal values, 1. should be furnished with 16 validators.

Thinking forward towards an admirable next phase would be monitoring of:

• Actual returns over a sufficiently large period of 90 eras
• Participation in GRANDPA voting
• Bad-actors in validating parachain blocks

If the council wants me to put together a sample application to administer the first phase I can do so. As for the second, I am hoping to benefit from Wei's work when he open sources retrieval of this data.

Regards,

Will | Paradox

Summary:

I know the above can be technical, the summary is, vote against this proposal, work on a more sustainable approach.

Definitely agree with your proposal Will, to summarize for non technical ones this aims to provided a fair and profitable approach for liquid KSM based on:

• rentability for lKSM staked
• security of the network
• performance of the validator
• decentralization and diversity of the relay chain
  The only part I would disagree with is the 0% commission, since it is largely used to play the nomination game today but doesn't give trust because no skin in the game besides being not sustainable. 2% seem pretty fair to me.
  Anyway I'll be happy to help if needed.
  bLd

The actual parameters should be configurable and are not exhaustive. I'm sure others can build on the idea, it is just a starting point.

With minimum commission right around the corner (might be in the next runtime) I think we may not have to worry about the sustainability risks associated with 0% commission validators.

All for it as well. Nice write up.

I have few comments about some of the replies in here, great discussion by the way!

1. Why not the Jaco validators? The reason why I focused on P2P is because they are more generally centralized (operating 100 nodes vs Jaco's 41). P2P is not a council member either. I think we should probably move towards decentralizing away from so many Jaco nodes too, but I think focusing on P2P first is a good first step.

2. Paradox, great comments, thank you. I must say you truly live up to your name - I'm very surprised you're advocating for profit seeking and centralization here when you were rebuking that very behavior in a different discussion. Maybe it's because this is a different context? But your philosophy should be applied unilaterally; from nominator to validator, from relay-chain to parachain. Moving from nominating five P2P validators to just one while nominating four validators who adhere the the very guidelines you are recommending here is a good thing and I expected you of all people to support this proposal - it's a small step in the right direction.

Self-serving? Sure, you got me. At least I proposed it on-chain with a non-trivial KAR bond. I also tried to include some other people that I think are deserving. It's my belief that I, and the validators I recommended to replace the P2P validators can provide more decentralization while also providing similar performance. My frame of reference can only go so far and the validators I'm recommending are solid choices in my opinion. You are always free to submit your own on-chain proposal recommending better validators, I'd love to see it.

Once LKSM stakers are able to hand-pick their validators this won't be an issue. This proposal is a short-term solution for the current light-weight version of LKSM staking. All because this proposal is not a permanent solution is not a reason to reject it. Will, why don't you work with Karura on a treasury proposal to implement those phases? It would be great to see you spearhead a more long-term solution. In the meantime, we can help reflect the ideals of decentralization with small steps like this proposal.

In any case, this is an actionable proposal that can be implemented in very short time. It's also relevant to the current configuration of LKSM staking. It would be a shame to stop one small step for the sake of a huge ephemeral step that might happen in the future.

TL;DR Accepting this proposal in no way impedes more permanent future solutions. This proposal takes action now and is a small step towards favoring decentralized independent validators over large multi-node operating validators.

Hello everyone

First of all, I'd like to say thanks to Will from the Paradox team for supporting us and inviting us to say a few words to defend our position. Here are the main points:

1. P2P Validator believes in a bright future of the decentralized crypto world. That's why we support many blockchains from the very beginning investing our own time and money. Worth saying that the number of active P2P validators in Kusama and Polkadot is constantly decreasing helping independent validators get into the active set.

2. For each blockchain P2P Validator dedicates an experienced team. For example, there is a tech team consisting of 3 DevOps who are working 24/7 maintaining our nodes and 2 Developers for Polkadot/Kusama as well as QA, Data engineer, Data analyst, and PM. The team constantly works on mature solutions in terms of security, monitoring, data mining, new products for our clients, and so on. Our team has spent thousands of working hours and still, we have something to improve, learn, propose and we are doing our best!

3. P2P Vaidators has got a customer support team that works 24/7 helping and educating thousands of stakeholders. We created our own help center where clients can get familiar with the networks we support. You can ask any question in our telegram chat or via our website and get a prompt professional reply at any time!

4. P2P Validator is a Grant-holder. We received a grant from Web3 Foundation for Polkadot Multi-blockchain ETL. It's currently in progress.

Actually, I think every blockchain should launch the network with professional nodes operators gradually giving way for independent validators (and it's currently going on in Kusama/Polkadot). An independent validator simply doesn't have enough capacity to do everything in terms of marketing, client support, tech tasks, etc. And of course, an independent validator will not do something for free for a long time.

About LKSM nomination

The nomination looks optimal for the launch of the product and done in the best way bringing optimal APR along with the best performance and security.

About the proposal

From my point of view, the proposal should be rejected due to the fact of bias. Thank you Adam for admitting it in the comment above. There are actually plenty of ways to attract the community to vote for your validator! Of course, it's not easy. And I can understand your desire to get into the active set with no effort :) But there are many other independent validators who might worth to be added.

About the solution

Of course, to have a fair distribution we need to find a solution that could take into account many variables and distribute stake in the best way.

I like the solution proposed by Will. I'd be happy to contribute! So far 2 comments only:

1. I also support changes in terms of min validators fee.
2. Everyone does mistakes so it would be great to be kind to those who were slashed especially if it was 1 level. So this point could be improved by adding a grace period or something else giving another chance to the inexperienced validator.

Idea

Since any mechanism will work based on criteria that are determined by someone else let's brainstorm the idea where we let the consensus implemented in protocol do its job :)

What if we split LKSM funds and distribute all funds equally between validators that fit the criteria mentioned by Will in point 2.

For example, If we have 888 validators (active and waiting) that fit, we just need to split funds between 37 addresses. Then we can range validators in the same manner as phragmen does and vote for 24 validators selecting each 37th validator in the range. Something like:

• address 1 nominates v1, v37,...,
• address 2 nominates v2, v38,...,

Of course, this will require the rebalancing service that is capable to:

1. balance funds between addresses to keep an equal number
2. change nominations based on criteria (if needed)

Here we can make everyone happy and we rely on protocol consensus only. However, it's just an idea.

Best,
Slava

Slava, with all due respect, this is not an attempt to get into the active set with "no effort" and to say such a thing about this proposal and all the work I've dedicated to my validator is quite frankly insulting and not becoming of the company you represent. This proposal is my very effortful attempt to show a potential nominator that I, and the other independent validators I've recommended, are a better choice over many nodes run by your dedicated company from the perspective of improving decentralization.

To reject this proposal solely on the basis that it's biased is asinine and detracts from content of the proposal. I could have proposed this anonymously but I didn't because I wanted to be transparent and didn't think it would be attacked for being biased.

I don't see anyone else making recommendations for other "better" validators. All I see are grandiose ideas that are not quickly nor clearly actionable.

You said,

"An independent validator simply doesn't have enough capacity to do everything in terms of marketing, client support, tech tasks, etc. And of course, an independent validator will not do something for free for a long time."

That is a largely speculative generalization that hurts the image of independent validators. To speak for myself, I am more than capable of doing all of the things you have listed - sometimes even more capable than a corporation. Not only that, I've been operating my validator at a loss for months and I have no intention of stopping. I validate out of principle and so do many independent validators. Your judgment on independent validators is probably true for some, but unfair to whose who are intimately dedicated to their nodes.

I do not believe the validators I have recommended have been given their due consideration. I also do not see how this proposal impedes the progress toward the long-term solutions that are being proposed here. Sure, there might be "better" solutions but no one is putting them on-chain and they are not as quickly actionable as this proposal.

Hello, Polkachu here. First of all, thanks to the community for including 2 of Polkachu's nodes in the program!

This is a very eventual discussion. Thanks Adam for bringing up the topic about how to best delegate the LKSM account to worthwhile validators, and thanks to everyone else for joining the discussion.

Rather than singling out a validation operator for subjective reasons such as centralization vs. decentralization, maybe a more productive way is to watch out for a validator's performance. If a validator does not perform, then we can swap them out for a new candidate such as those listed by Adam.

On polkachu.com, we strive to provide more transparency about validator performance. For example, this is the dashboard for Karura's LKSM account here. As you can see, thanks for Karura team's diligence, most validators are performing extremely well. That's why the liquid staking program can pay out ~20% rewards.

However, you can also observe some areas for improvement. For example:

1. "KeepNode/Lithium" has substantially underperformed in terms of APR in the last 90 eras. If you click into the validator for more details, it looks like the underperformance issue has been resolved, but we should've caught it earlier.
2. Many validators do not pay out rewards in the very next era after a reward can be paid out. This prevents the program from the maximum compounding.

My main point is: if we can establish some good quantitative criteria for selecting and de-selecting validators, this can benefit the program in terms of financial returns as well as avoiding rather subjective and political discussions like this one. We can also establish a review cycle for the delegation on a quarterly or semi-annually basis, so the validator selection / re-selection do not depend on ad-hoc proposals like this.

We are happy to assist by improving this dashboard if the community thinks it is a worthwhile effort (for example, the page loading is currently slow due to lack of caching). We can even provide an API endpoint to provide more stats from additional validators, including those listed by Adam. We are very thankful for the inclusion in the validator set, and would like to contribute more to the project beyond providing a good validation service.

Polkachu

Hello Polkachu,

This is useful information and in keeping with the direction I think we should focus. Establish healthy criteria and let on-chain performance guide who stays and who goes. There are however soft-matters that can't be attained on-chain just yet, like slashing insurance that also requires consideration, these can be evaluated by exception.

If I'm not mistaken the roadmap for LKSM eventually puts the responsibility of nominations to LKSM holders. Before this is enacted perhaps a system of continuous review (like the script I mentioned) above might be useful, if it works well perhaps the holders may agree to its long term usage?

Just to get the ball rolling I started on the nomination script (as mentioned above) and it is almost completed. It should be available well before this 'remark' can be enacted :D. All open source, which is very important to ensure data isn't misrepresented.

Paradox

Hey Everyone,

Ilhan from Staker Space, as far as I know the LKSM validators needed to submit a form, we did this nicely and informed the team about this.

The Karura team has some partnerships with Blockdaemon and nominated their validators early stages when LSM was just kicking off. Since Blockdaemon just had 4 kusama validators some other community validators were selected, in which the commission also played a big role in order to guarantee the 16-17% return.

The Karura team took decisions with or without the form to select some validators, which is fine by us as the functionality of being voted in for the validators is still to be built and it was not transparent, but there was a nice selection with Paradox, Polkachu and some more heavy validators such as P2P. Nothing wrong here.

As a user of the Karura platform I know the cross-chain integration and EVM has more priority as it unlocks other parachains to Karura and other possibilities. Glad to see that this is prioritized over voting for the LSKM stash from the community.

Also find it a bit Sketchy that Steeber & Co get selected in the proposal and not pointing out to the 1K kusama validator programme with less then 3% commission. If decentralization is the effort here, that would have been a much cleaner proposal instead of pre-defined validators.

We will be voting against this proposal and hope it doesn't pass and if it did wondering still if this would be honored?

@ilhan Clearly won't be honored if it isn't passed. Does democracy mean nothing?

I'm disappointed that people think that self-advocacy is sketchy instead of judging the validators I selected fairly. The only reason I did this was because there's an incentive for me and for the other three validators who are working very hard and deserve a little uplifting - this wasn't coordinated, this was all my doing. I thought that maybe these incentives aligned with an improvement to the nominations.

I forgot how cold blockchain democracy is - probably for the best. I know we shouldn't allow personal biases in technological decision making. On the other hand, I truly think the four validators I selected are an improvement for decentralization. They are also very reliable individuals who are very dedicated to the ecosystem. They deserve to write blocks.

But this isn't the way to go about it, eh? This is teaching me that personal incentives have no place in democracy which is a shame because at the end of the day we are all human beings with biases. Of course I'm going to fight for this because I stand to gain - is that worse than not participating at all? I truly am trying to find a win-win here, I don't think P2P needs the nominations, after all they look down on independent validators from a place of superiority claiming we aren't capable of the things they are capable of. Is this a validator we want reigning in the active set? One who thinks independent node operators are inferior to themselves?

For the record, all the validators I selected are devout members of the Thousand Validator Programme who intend to keep low commissions once sustainable.

We're getting closer to the finish line!

Hey guys,

I completed the bot a few days ago and I've been doing some internal tests. You may take a look at the repo ->

https://github.com/paradox-tt/TVP_Nomination_Bot, the opening page has some details on how the bot works.

As a wild idea, consider if the settings for the bot (Setting.ts) are held on the Karura chain, the bot can load these variables into the application upon load or upon governance event. If community members wish to change preferred nominee, commission levels or weights they can do so via governance.

This might be a good way forward.