Collator Community Guidelines Revision: Reinstate Maximum of Four Collators per Individual / Entity per Network

, ,
1

Collator Community Guidelines Revision: Reinstate Maximum of Four Collators per Individual / Entity per Network

Summary:

I am proposing to reinstate the maximum number of collators an individual or entity may run on the Moonriver / Moonbeam network to four collators. Since Moonriver/ Moonbeam’s inception the maximum number of collators per individual / entity has been four and was only previously changed down to two as part of a Proposal to increase the minimum bond. For the reasons outlined below, I do not think reducing the maximum number of collators per individual / entity to two is beneficial in the long run for the Moonriver/ Moonbeam ecosystem (including for the smaller collators this change was aimed to protect).

Rationale:

TLDR: Based on the community discussion around this topic, it is clear that the rationale for decreasing the maximum number of collators per entity was that “we need to care about smaller collators” and I totally agree.We need to protect loyal and engaged members of our community no matter their size. Reducing the max number of collators per entity however is not the right solution and it could have harmful repercussions for Moonriver / Moonbeam’s health and growth. I have outlined below why I am opposed to reducing the max number of collators per entity from four.

● Drives away players from Moonbeam:
The reduction of max collators per entity looks undeniably like a protectionist economic policy that could drive away larger players and isolate Moonbeam away from growth opportunities. At least 1 institutional player has indicated they will walk away from Moonbeam if this is enforced. Their customers are mostly looking for stable returns; if the community at any point can unilaterally change the rules; this makes those returns very unpredictable; and in the end- there are 200 other chains they can go after.

We also need the larger institutional collators because:
○ Institutional collators are good for Moonbeam because they help reduce GLMR’s velocity: there is still a huge amount of value institutional collators bring because they lock the token and that overall increases demand and interest in the network (even if these whales aren’t as directly involved in the community discussions, etc.)

○ Institutional Collators Bring Awareness and Users to Moonbeam:
Entities like Kraken, Binance, and others, bring in users from outside the network that otherwise would not be interacting with Moonbeam.

So even if the collator entities themselves aren’t directly as engaged as some would like, this doesn’t mean they don’t serve an important function in raising the profile of the chain (as well as locking up tokens). Losing institutional collators by reducing the max number of collators per entity, or discouraging new entrants, would have detrimental impacts to Moonbeam because we could have greater token velocity and lose future new users and growth potential. So the smaller collators may get a pyrrhic victory in the short term - they may see more slots open up, but everyone’s returns would go down.

● Inconsistent & Confusing rules:
Also, as it stands right now, the rule is very confusing. People on the forum have indicated it would not be enforced for existing entities; but as the proposal is written it says:

“An entity is running more than two collators unless the community determines by governance that it is for the benefit of the network that an entity should run more than two”

…

“The community may use on-chain governance when any of the guidelines described in Part or Part II of this on-chain remark are not adhered to. Specific actions that may be taken (but not limited to) include:

● Removal of an individual collator or set of collators from the active set
● Slashing of collator self-bond
● Permanent ban of the entity from participating in the collating function on the network”

The way that is written, and what was actually voted on; it implies that anyone running more than 2 collators should immediately put in an on-chain proposal and ask for an exemption; or risk sanction. Just a handful of people on the forums saying “it’s fine”, is not a sufficient guarantee; because that’s not what it says on-chain.

Furthermore, laws that require exceptions are typically not good laws and are fraught with inequality and execution bias. I think you also have to consider that from the perspective of someone new to the chain looking at the ecosystem; this looks like a protectionist policy - it’s not exactly “welcoming to newcomers”.

● Uncommon policy:
I tried to find other examples of chains where this was put in place and I wasn’t able to find any. That doesn’t mean it doesn’t exist on some other chain, but it’s definitely not common. The fact that this is not a common decision within web3 is a bit of a sign it’s not something we should pursue.

● Omnibus law:
In a very real way, this rule change got “buried” in the main proposal. The abstract of Proposal 88 explicitly mentions the increase of minimum bond to 2M GLMR; as well as the establishment of common guidelines. The detailed guidelines then do mention that there is now a limit of 2 collators per entity; but it’s one single line in the proposal; and it doesn’t even mention this is a change from the previous guideline. In speaking with people across the ecosystem; the majority of them were not aware of this change until I explicitly pointed it out to them. I feel this should have been split out into a separate proposal; not combined with the increase of the minimum bond.

In closing, I would argue we should revoke the 2 collator per entity limit and put it back to 4.

That being said, we do need to protect the smaller collators because they are important members of the community, they are far more active in governance; and actively partake in discussion and the future direction of the chain. But I think there’s other ways we can address this - perhaps we can ask the foundation to look at increasing the number of collators that can be in the active set; it would seem that would take some of the immediate pressure off. And perhaps there are other ways the smaller collators can differentiate from the larger ones than just a straight-up competition with APY; and the Foundation can help with that.

Reference Material

Community Guidelines

“Collators have a responsibility to the network to act honorably. If any of the following forbidden offenses occur, action may be taken via on-chain governance: An entity is running more than four collators in either network”

Original Forum Discussion that started the discussion (Nov 22)

Forum post with proposal to raise minimum bond and lower max collators per entity to 2 (Feb 23)

Forum Discussion around addendum on the Recent Proposal - Grandfathering discussion (Mar 23)

Onchain Referenda:

Moonriver, On-Chain Remark for Collator Guidelines and Request for Collator Self-Bond Change | Polkassembly
Moonbeam, Moonbeam On-Chain Remark for Collator Guidelines and Request for Collator Self-Bond Change | Polkassembly]

2 Likes
2

As you mentioned, the institutional player can bring their case for running more than 2 collators to an on-chain vote. Are they saying they will walk away if the vote is Nay, or that they are not willing to bring this to a vote?

On your points

Drives away players from Moonbeam
It drives away players that do not want to bring their plans to run more than 2 collators to a public vote. Players that want to run more than 2 nodes can bring this to a vote. You are suggesting that players should be allowed to run more than 2 nodes even if the community is against that, i.e. without a vote. This takes away the “interest balancing” lever the community has and basically opens the door.

We also need the larger institutional collators
Institutional Collators Bring Awareness and Users to Moonbeam
Correct. That’s why the community can decide to let a big player in. Like all things though, there is a fine balance. Should the collator SET be composed of 20 big players with 4 nodes each? I disagree, and I think the community would disagree too.

Just a handful of people on the forums saying “it’s fine”, is not a sufficient guarantee; because that’s not what it says on-chain
The rule does not say to bring it to a forum vote. It says “the community determines by governance”. It’s pretty straightforward imho.

Inconsistent & Confusing rules:
I think you are calling the rule inconsistent and confusing because it has an “exception”. If that were the case, all laws would be inconsistent (see next comment). Moreover, inconsistency and confusion would increase if we change the rule again.

laws that require exceptions are typically not good laws and are fraught with inequality and execution bias
All laws are applied after they are voted which means you don’t get punished if you broke the law before it was voted. This is exactly what this law is doing. It does not apply to those that already have 4 collators. What you are saying basically applies to every law ever made.

Uncommon policy
A policy of max-2 collators is as uncommon as a policy of max-4 collators. Your argument is pro no-collator limit, not pro 4-collator limit. So, this is another discussion altogether.
In any case, the max collator limit makes sense in Moonbeam because of the diminishing marginal returns on increasing collator stake, so, it’s a custom rule for a custom rewards model. The foundation saw this well in advance.

Omnibus law
I do not agree (the discussion had 40 comments and lots of eyes on it), but obviously, this is a subjective view. That’s exactly why Governance exists, to integrate subjective views on-chain, and that’s how this rule was voted - through Governance.

I appreciate your appreciation of what small collators bring to the table. I think it’s important to leave it up to a community vote to decide if 2 collators should go so that a bigger entity can have 4x what each one had. I am sure there are cases where this applies and the network would benefit. I am also sure that not all cases are like that and that Governance should keep the keys to the door.

6 Likes
3

I would like to express my thoughts and point of view regarding the 2 or 4 collators per entity and the importance of involving major players in the development and popularization of the network. while I understand the reasoning behind organizations like PureStake & MF having 4 collators, I am uncertain why this number should remain the standard for all future large organizations.

encouraging the participation of major players who are willing to contribute to the network’s development is important. however, I find it perplexing that they are dissatisfied with being limited to 2 collators while tending to have 4. I would appreciate a clearer explanation on this matter to better understand their perspective. many other collators and community members may also wonder why certain community collators have self-bonded only 10k when new collators need to have 2M GLMR.

as mentioned earlier, any organization can start a discussion on the forum by expressing its intention to launch 4 collators. this process appears to be less time-consuming than setting up 4 nodes. therefore, it seems reasonable for organizations to invest a small amount of time in writing a brief message to the community if they are genuinely interested. from my standpoint, this does not appear to be a significant issue. I would greatly appreciate further clarification on this matter.

as an active community member, I share concerns about the potential recurrence of past problems when collators were frequently knocked out from the active set, leading to dissatisfaction among delegators… It is disheartening that delegators lose 2-7 days of rewards in such situations, which may discourage their further participation. additionally, I have noticed a decline in the activity of smaller collators, with their participation primarily limited to updating and maintaining their nodes.

It is extremely important to emphasize that this proposal will not affect active small collators in the future. In other words, if the collators are actively involved in the network, participate in discussions on forum, vote in governance and contribute to the development of the network in various ways, the presence of large organizations should not be a factor that displaces or reduces the role of these active small collators.

to summarize, as someone who appreciates a vibrant and involved community, I believe it would be beneficial to have an active collators community. If the inclusion of large organizations as collators can bring in a significant number of new users, act as a marketing tool, and enhance the reputation and engagement in Moonbeam, I am in favor of reinstating 4 collators :slight_smile:

3 Likes
4

Hi. Thank you for your suggestion.

I can say that in my opinion the importance of collators as large companies is overrated. Because such collators have almost no involvement in the life of the network. They interested in stable returns? I’m not so sure that we need miners… A good collator/validator understands exactly how important decentralization is. The decision to allow 4 collators per company goes against the ideas of decentralization. And it should be in the best interest of the network first and foremost. Not in maximizing their profits. I have nothing against wanting to cover your costs, it’s quite a right and healthy desire. If a passive collator who is only interested in making a profit from steaking leaves us, but instead 4 new collators come in who are interested in growing the network… Well, that doesn’t look bad to me personally. It’s good to have big companies among the collators, but it’s even better to have active collators.
Most networks have one validator from one organization. Honestly, it’s very rare when you can see otherwise. Yes, there is no prohibition. But you’re unlikely to see support for something like this from the community, the team, or other validators. Moonbeam has a rather unique situation in this regard.

Then again, as far as their involvement goes, they don’t even know that there was such a thing as reducing the number of collators allowed. They could have voted against it that time… And I think it’s a bad precedent to go along with the desires of large collators, whose desires are based solely on their own interests. And not the interests of the network or other collators.

By the way, I agree that the decision to reduce the number of collators should have been a separate proposal. It is worth considering this practice for the future and not overloading the proposal with different conditions.

4 Likes
5

I think that’s the most important thing we have to consider. As a rule you don’t want (at all) to watch your steak. You want to put it down, forget about it for a while, and not worry that something could go wrong. And when you find out that your steak hasn’t been active for a long time and you have to wait another week… It brings up a lot of unpleasant emotions.
So the last thing we need is new moves in the active collators network :grimacing:

3 Likes
6

Well said Ioannis. To my thinking the key point is that since existing collators are grandfathered in, the 2 collator rule only applies to new entries and all they have to do is come to the community and present their case. I don’t see how that small step is a significant barrier.

3 Likes
7

Personally I don’t support revoking the the 2 collator per entity limit. Lots of reasons for implementing this have been discussed and mentioned before.

2 Likes
8

after conversing with multiple stakeholders, I am hesitant to deem it as overrated. nevertheless, it is undeniable that certain well-known organisations may not yield any significant benefits. conversely, there are a few notable collators, such as Staked, Binance, which emerge as reliable entities. these collators effectively utilise their extensive user base to promote Moonbeam, thereby contributing to its marketing endeavours. consequently, this aids in expanding the network’s outreach beyond the scope of DotSama.

that is, if an organization with a fairly good reputation wants to launch 4 collators, then, of course, I would not want the new guidelines to interfere with them, but, as discussed earlier, this can be solved quite simply and quickly with the help of a discussion on a forum within the community.

as mentioned earlier, smaller collators may also exhibit waning interest in actively participating in the network’s activities, as evidenced by their reduced engagement in Forum, Discord, and governance.

although the topic of our current discussion primarily revolves around the selection of 2 or 4 collators, I would like to share an additional noteworthy example that sheds light on the potential challenges posed by new guidelines for large organizations undertaking remarkable and special initiatives.

one specific instance that caught my attention pertains to StellaSwap, which is actively working on liquid staking opportunities for stGLMR. however, it appears that the current guidelines could pose obstacles in this regard. the stringent self-bonding requirements, for instance, may impede the launch of 4 collators.

summing up, it is important to recognize that pros and cons exist in every situation. I express my deep concern about the state of the network and emphasize the importance of attracting major players. It can be difficult to understand the mindset of these important stakeholders if you are not directly involved in business. collective feelings and the prevailing mood within the
community can potentially intimidate them

1 Like
9

I really like @turrizt’s take that these possible entities could just speak up in the forum, outlining thier plans and getting an OK signalled to go on.

Other than that I don’t really see the importance here and now.
In the current situation it seems not-too-likely that a new entity just spawns outta nowhere with 4 nodes at once.

2 Likes
10

I do not support reverting to the previous situation that, as we all remember, led to chaos in the active set, damaging delegators that, IMHO, should be prio 1.
We have to find a balance between attracting capital and having a chance for community members who are active in all senses, who develop tools, who offer services for free to stay in the active set. That sweet spot has been reached I’d say, so I do not see any benefit in going back to the previous situation.

2 Likes
11

I had to break out my very old WFF 'N PROOF game to help guide my thinking on this. It reminds me of when a large bank or other big corp opens up shop in a small town – creating jobs, helping fund new infrastructure, boosting economic output, but also puts some mom and pop stores out of business in the process.

As an alternative, why not remove the statement in it’s entirety. Whether two or four, it will only be effective for those that are more likely to play by the rules. Anyone with enough resources to put in 2, or 4, or 6 collators for that matter, can do it and easily obfuscate the fact. The community can always act via on-chain governance no matter what the scenario, and whether that statement is there or is not there.

With regard to evening the playing field a bit for mom and pop, and just thinking on the fly, if there were a holding pen for first time collators that have enough bond + stake (considering that it’s clear any new joiner will have enough to bond + stake immediately after issuing joinCandidates) to enter the set during the very first time they want to join the set and collate, they must wait a set amount of rounds.

As an example, if it’s 5M GLMR to reach 72nds spot, then once a new waiting collator (never created a block) has 5,000,001 GLMR total staked, a countdown of rounds starts before the collator is fully eligible. This gives mom and pop time to make necessary adjustments (and attempt to increase the set via on-chain governance). It also has other advantages. Recently, in another parachain, more than 4 collators entered the active set that were not producing blocks for weeks. This hold-down timer can help gauge performance before a new collator enters the set. Admittedly, this idea would require a lot more consideration and thought.

1 Like
12

I’m not going to argue about this or change your mind about anything. But this is exactly the kind of business I am in.

Yes, that’s right. And I think there was a discussion about it here a few months ago as well.
I’ll just clarify that I’m not saying that a small collator is always good and a large collator is always bad.

13

Just one point that we’d like to add is that the discussion of whether a single entity should be able to run 2 or 4 collators is currently pointless.
There are currently 7 anonymous collators in the set, collators without an identity set, for all we know these 7 could be run by the same entity. We need to first establish the enforcement of the rule concerning anonymous collators, and the relevant penalties that should be incurred when this rule is broken. Then we can proceed to establish the procedure for dealing with delinquent actors.

As a further point, even with identities set, we will never really know whether a group of collators is actually run by the same entity or not. One entity could very easily spin up 2-3 twitter accounts over a span of time and claim each (pair of) collator(s) belongs to separate identities for the sake of bypassing the rule.

1 Like
14

I agree 100% that if we make rules we should at least “try” to enforce them. You are right about the anonymous collators and we should probably do something about that.

I propose that we raise awareness of the expectation to set identities and see if we can get some traction that way.

1 Like