Jim, thank you for raising this issue and making us pay attention to it, I join the vote. All the guys who were selected by the team and took an active part in the early testing deserve a place in the active set. Time has shown that community collators are in no way inferior to large companies in professional terms, this is confirmed by uptime and the overall stability of the network, which many of us provide. My voice is 2m 
4 Likes
I can hardly see the proposal can help diversify the set with smaller independent collators as they are unlikely to attract such a large investment. But I agree that it at least can make it harder for larger entities thus preserving those independent collators that we still have in the active set. Moreover, given the drop in prices the current dollar value of minimum bond is too small. So I voted in favor.
6 Likes
I also voted in favour of a 2m bond, although it will shut out independent collators forever to enter the active set. On the other hand, smaller entities already stand no chance to enter the active set and we now also have the orbiters program, to which smaller entities can apply (it has, however, its own difficulties, as seats are already allocated). Weighing in both pros and cons, I still find it essential to raise the min bond. Otherwise we end up in a situation where VC funded staking teams and venture funds take all active seats with 4+ collators each.
5 Likes
I agree with Jimâs proposal. I think that due to the current price the amount should be increased to a minimum amount. I think it is a correct way to delicately take care of the stability of the network on the collators, and as Sik comments, if the price of the glmr token increases in price, it can be reverted again or make a new proposal. I agree with an increase in the minimum amount, since this will make people who enter have to think about it twice, since an increase in the min bond will make them see that they have an effort and responsibility. We also saw some checkers without producing blocks for many rounds, this will mean that due to the amount that will be needed to enter, the least these things will not happen, since the investment is greater.
5 Likes
First of all, I would like to thank Jim for spearheading this conversation, and for everyone who contributed. The foundation has been following the discussion, and weâve been really impressed with the thoughtfulness of the comments - weâve learned a lot from the discussion in this thread.
Weâre taking the discussion back to the engineering team to talk about the mechanics of making this change and whether this would need to be in a runtime upgrade. We will post back here with an update once we have clarity on that.
Separately, we also saw the 19-person poll; which was great but itâs triggered debate inside the foundation whether we should generate more community involvement and feedback before it goes to a full vote.
We will come back with an update in about a weekâs time.
7 Likes
What does the foundation think about this? that it is mandatory to have a way to contact them in case there is a problem and also the same nominators can contact them if they have doubts, because it seems unfair that when there is a problem, they cannot be contacted.
and most surely contact with entities and establish no more than 2 collators, per entity
In any case, it could be taken to a community vote, but it might create a power struggle in the vote. ( vcs - cex vs comunity)
3 Likes
I see 3 ways to adjust the collator rewards without adjusting tokenomics or enabling slashing (which we should avoid for now).
- Adjust the pool size (collators and/or orbiters)
- Adjust the min collator deposit
- Enable/disable collator deposit delegation rewards
There are some technical limitations around #1, so we should keep that as a constant for now, but it will be an important variable after these limitations are removed.
#2 has been discussed already above, there are multiple reasons why this may be a good idea.
For #3, the collator bond is currently earning rewards as if it was delegated to the collator, this can be turned off to make the minimum bond more effective.
Currently, compared to simply delegating, running a collator will always increase total rewards, no matter what the minimum bond is (since the bond is generating the same rewards as simply delegating, running a collator is just gravy on top. So anyone with enough to post the entire bond is still strongly incentivized to do so.
If the bond no longer earns rewards, it will be easier to dial in the rewards % where itâs no longer as attractive and possibly not worth the effort to run a collator if are in it only for the money.
We should discuss what the appropriate reward for new (and existing?) collators should be. IMO it should be somewhat higher than just delegating, enough to make it viable but not too much people only do it for yield generation. Once we have this target reward number we can solve for the other values.
6 Likes
hey @artkaseman , although what you say seems interesting to me
More than adjusting the rewards of the collators, the aim is to find a way to have diversity in the active set and not be plagued by the same entity or a few large organizations
leaving out the small collators
It has been discussed that raising the self bond, necessary to set up a collator, be increased, so that it is less likely that a company will make a monopoly (more risk by needing more glmr)
In addition, with the drop in the glmr/movr price, the necessary threshold in $$ terms, has dropped considerably, so the above makes more sense.
3 Likes
Did a quick spreadsheet. Looks like at a required self-bond of 900K and self-bond rewards at 0%, there is no incentive to run a collator vs. delegate
4 Likes
Thank you @stakebaby for the calculation.
In general I like the idea of reducing the inflation rewards on the self bond to 0%. This directly translates to higher delegation rewards because the total block rewards per node stay constant but the self bond doesnât participate anymore.
Starting the calculation with the assumption that the expected rewards from collating shouldnât be higher than delegation rewards using the minimum node size to become active is in my opinion misguided because it directly translates to incentives for collating. The whale can effectively lock in the right to draw a stable 20% commission forever while nodes grow over time and dilute infltion rewards for delegators. This starts the moment the node enters the active set since ultimately nodes should tend to grow towards the mean total stake.
I argue that at the time of bonding the collator shouldnât have a direct incentive to collate. Over time the mean total stake on every node will rise and every existing collator will enjoy higher future rewards compared to delegating but that doesnât favor a single operator. This is already incentive to start collating but the additional rewards come later in time.
Also the assumption that in case of delegation the user would be able to capture the highest APY of the active set is not relistic for the same reasons. Itâs more likely that the mean APY can be expected if the delegations are spread out across existing active validators and ech one of them tend to move towards the mean APY over time.
If I use the geometric mean of the total stake and APR I get to the 1 million self bond as break even point
4 Likes
Another consequence that should be considered with this approach (thanks @Goldberg_StakeSquid) is that the APY for new entrants would much higher than other collators since their bond would not be receiving rewards.
This might make the tactic of delegating a high amount to stay in the set but revoking to pump APY until 3rd party delegators make up the difference and then repeat with a new collator more effective. Instead of removing bond rewards, redistributing them across the entire set of delegators (of all collators) could offset this but I suppose would break the model of each collator generates rewards for its own delegators.
3 Likes
That is a good point @Jim_CertHum, the new collator if self funded and the bond is not counted toward the delegators total, would have much higher APY compared to others, likely guaranteeing their slot in the active set. However, it still would raise the barrier to entry and discourage reusing delegations to boost in multiple collators.
On the flip side however, I am concerned we are raising barriers to entry and effectively engaging in rent-seeking behavior by making it harder for any new collators to join, whales or independents.
The Orbiters program is a way to maintain participation of independent collators, but with the current limitation on increasing the active set combined with the current bear market, makes that not as attractive as it could be.
Looking ahead a bit, the limitation on increasing the collator pool size will be removed, is there any combination of total slots / collator slots / orbiter slots that could solve this problem?
3 Likes
Regarding raising the min bond, @stakebaby and @Goldberg_StakeSquid have made a good case for this.
Regarding removing rewards for collator bond, this also seems like an excellent option, but @Jim_CertHum raises a good point about what happens to those rewards?
Perhaps we can consider:
- leave all reward calculations the same, but simply send the rewards on the collator bond to the treasury for community needs.
- leave all reward calculations the same, but âburnâ the tokens instead of delivering them to the collator.
4 Likes
I do like those suggestions, especially sending some more funds to the treasury, however they are veering toward tokenomics, and a proposal that involves them will take a long time to discuss and pass with the entire community.
If we keep the scope on collators and short term changes we can get something passed much sooner.
Weâre taking another look at the max pool size limitation, and a small increase may be available now. Letâs assume for this discussion we can add 4 more slots to Moonbeam, to match Moonriver.
Should we increase the pool size and if so, how many should be regular collators and how many orbiter slots?
3 Likes
If all of the spots (68) were Orbiter slots it would solve the problem but I donât think that would be a popular opinion.
You are right that when the active set is able to increase in size then a higher minimum bond will make it very difficult for a smaller independent collator to join. As of right now the barrier to become active is approx. 2.7M GLMR and 19k MOVR., so raising the bond certainly seems more reasonable as an interim measure until the set increase limitation is gone.
However, if we look at the past when the set expanded we see that it only acted as temporary measure until whale collators pushed out smaller collators and increased their total slots. I think as slots increase there is a greater potential that one backer would add more anonymously, squeezing out anyone else trying to enter.
So perhaps the answer is for the set to just continue to expand until we achieve some sort of natural equilibrium, but that will most certainly involve more than 4 collators from one entity.
2 Likes
I hope we do raise the min bond AND drop the rewards on the bond.
Regarding additional slots, my perception is that recent additions to the active set have all been anonymous/whale types and not active community participants. Maybe Iâm not looking in the right places⌠but that is my perception. In which case, I would think that increasing the orbiter slots would help us acquire more smaller independent collators which would (hopefully) contribute to the active community.
3 Likes
it seems to me that increasing the active set helps only for a short period, unless the orbiters get there. since, based on past experience, whales from time to time begin to launch several new nodes into the active set and thereby knock out small independent collators⌠when the active set increases, itâs only a matter of time â when the new whales decide to join the game 
3 Likes
I think that one day we will wake up and 10-20 slots will be gone to one entity, and there will not be going back from that because any measures introduced thereafter, will also protect those nodes too. It happened to Shiden recently. From the total backing numbers I see at stakeglmr.com, itâs only a matter of when.
Arthurâs point for having a sufficiently low bond to allow entrance of smaller players during set expansion makes sense. But set expansions are rare, and from experience the only time that the self-bond was the barrier to entry was on launch. Every other set expansion saw players lining up to grab the spots with many times the self-bond. So, the self-bond seems to have become more of a theoretical barrier than a practical one.
I think the heart of the problem is that whale holders have only upside in running a collator and no material risk. Fixing that requires introducing some downside or risk. Possible solutions:
- slashing on their self-bond
- no rewards on their self-bond (forgone profits)
- longer illiquidity/lock imposed on their self-bond
- public shaming, ehmâŚ
Compare that to a community collator who has a huge risk of losing his spot in the set if he misses a couple of rounds due to delegators running for the door (thatâs risk/reward done right and working FOR the network).
I am not saying this is unfair (it is), it is just a constant force toward the de-decentralization of the set. We see it at work every day.
Expanding the set by 4 spots will allow players that can come up with 2+M to enter that is whales, or maybe just one whale (it has already lined up). Total backing would also be slightly diluted and the small player that are community backed would lose delegators. The only small positive effect the expansion would have (towards making it less enticing to collate vs delegate) is to reduce collation rewards, but if you wanna go for that then why not reduce the 20% directly.
Huge conflict of interest here, so please donât take me seriously.
4 Likes
correct me if Iâm wrong but I feel that the last time a community collator entered the active set was Polkadotters and they are out for a good while now. Itâs hard to define community collators. I consider LEGEND as one of us and he has 100k self-bond but everyone else I assume âsmallâ started from the testnet and has 10k self bond. After Polkadotters, if itâs not a collator funded by a known wallet or a collator that adds /n to his name, we donât know who they are. We donât hear anything from them either.
The main mechanism for orbiters to prevent whales is to make it permssioned by KYC. I donât necessarily think that this is a future proof way to grow a healthy collator set. I believe economic incentives are much stronger mechanisms. And I believe that the mistake that we try to correct here was made long ago in the private sale and seed rounds. I derive that judgement from the absense of those problems in Moonriver that didnât have allocations for VC.
I can not totally rule out those problems in Moonriver but at least in Moonriver I was never forced to set my alarm every round change to check and the network is running much longer. Others might have a different experience.
I support everything @stakebaby posted right above me now. The root problem is the higher rewards. If we bring that down for new collators it will have chilling effect. If we have to bring that down for any collator then we start yet another race to the bottom where the better funded party uses economies of scale to run as many nodes as he can using the same infrastructure during the bear in expectation of profiting big in the bull. The more nodes you have the easier it becomes to reuse the stake to establish yet one more.
For the sake of completeness, I fear that the commission at which the extra income through collating will become irrelevant for whales lies somewhere around 2% but that directly incentivizes taking bribes by MEV.
2 Likes
Had a good nightâs sleep, so I wanted to add this to my essay aboveâŚ
Increasing the rent, i.e. the opportunity or actual cost of having a collator, is not rent-seeking if that is done to protect the networkâs interests. There is already a rent to be paid, and that is the illiquidity of a 100K bond. Many of us here think thatâs too low.
âRent seeking (or rent-seeking) is an economic concept that occurs when an entity seeks to gain added wealth without any reciprocal contribution of productivity .â
So, if anyone is rent-seeking, that is whales that are taking over 5 or more collator slots. Not that I need to do any fancy transaction tracing and count wallets to know that whales ARE taking advantage of the system to its detriment. I can simply count the number of whales raising arguments in this blog.
Zero.
2 Likes