It was a wish - not a confirmation
In the CIP Editors Meeting on 23 February 2021, both CF Technical Project Manager @Frederic and IOG chief technical architect Duncan Coutts agreed upon and planned to address this issue. They seemed enthusiastic about it at the time but I haven’t heard of any progress made about the issue after about 2½ months, so I hope they will respond here or in a new thread.
We were discussing the issue that I commented upon here, regarding @shawnim’s earlier proposal to mitigate the problem by (I’m greatly simplifying here) crediting pool performance somewhat towards pool ranking, instead of ranking mainly by size. Even though that proposal has gone by the wayside, this issue is still being discussed, and needs to be discussed more. I was at this meeting to represent another CIP but was invited to speak on this issue based on this comment:
It should be a relief to all the small SPOs, excluded from pool rankings no matter what their rewards, to hear Duncan confirm the reason for that exclusion and that it’s not the goal of the developers but a consequence of the design parameters that they were given: i.e. to very sharply limit delegated stake to K saturated pools and leave all other pools with nothing but their pledge (I’m “R” and “Robert” in this discussion, not fully credited in the meeting roster):
As recorded in these minutes, Duncan @ IOG says:
- the change needed is to the actual infrastructure. Not shooting the messengers, but rather requesting/campaigning for changing to the actual incentives of the system.
- it may make sense of reevaluating wether the ranking actually reflect the proper underlying incentives of the system.
- we shouldn’t have a cliff edge at k, we should probably have a slopeoff.
To get the full resolution from that meeting, we have to go to this video… since from there on, the minutes only mention “flag to researchers” and there the trail stops. Below is a partial transcript of the discussion summarised above plus the emerging plan to discuss this issue with researchers and develop an updated CIP to document and correct this issue:
- (46:11) Duncan: starts speaking (long passage already paraphrased in minutes)
- (48:09) (me) it’s about the “slope” of the curve [of pool delegation] that happens inside & outside the boundary of the K largest pools.
- (48:37) Duncan replies and says there should be a smooth drop-off instead of a “cliff edge”
- (50:48) Duncan: I think it would be perfectly reasonable to write a CIP that addresses that issue.
- (52:15) Frederic: (re: where do we go from here) back to researchers? Duncan: ask original CIP submitter if he still wants to pursue it?
- [briefly mentioning] a reference to what I said in Github comment above; i.e. the problem was not addressed by raising K to 500, nor would it be fixed by K = 1000, because curve drop-off shape would still be the same
- (53:28) Frederic says he can connect researchers to the issue about changing the underlying incentives that cause the cut-off to be so sharp between the large and small pools. Duncan offers to cooperate while also saying the goal would be better achieved if it were coming through Frederic.
Today, I believe there has been no progress about this issue because no new pull requests have been submitted that meet this description (anyone interested can subscribe to this feed on Github):
As I’ve said before on this forum, I think the Cardano Foundation support of small pools is extraordinary. The main reason our inclusion in the CF’s pilot delegation group in October 2020 left us back where we started, as it has done for most other small pools, is that the delegations were only done for 4 epochs.
Conclusion: If a 15 million ada delegation were to return to us for a period of 3 months it might well be enough to assure our pool’s viability… and it well might not, as long the problem of the K-pool all-or-nothing Nash Equilibrium remains a core component of Cardano’s design principles.
A provocative statement: @Cardano-Foundation With current ADA price levels, your delegation methodology creates a strong incentive against co-operation. As a result, we have 2000+ pools that are nearly useless to delegators.
ADA holders can co-operate to contribute to the PoS system in one of two ways:
- Delegate to someone else’s stake pool
- Combine their ADA to fund a pledge and jointly run a pool
Both approaches can result in pools having enough funds (delegations+pledge) to mint blocks more or less regularly, at least once every other epoch.
In practice we see a lot of (1) and very little of (2). Instead, there are hundreds or even thousands of pools with very little pledge and only a few delegators, even though these pools can’t generate consistent rewards for their owners. This means that there are many, many SPOs with not enough funds to regularly mint blocks who nevertheless choose to neither delegate nor pool their money with others in order to fund an attractive pool.
In other words, even small ADA holders deem it more attractive to become solo pool owners than to cooperate in one of the ways mentioned above.
Indeed, the current delegation programs by @Cardano-Foundation and IOG strongly incentivise such behaviour and I wonder if this is by intention and if not, how we could improve upon the program.
An illustrative example: Suppose I own 25’000 ADA. By delegating, I can expect to receive 1’250 ADA per year if the average RoS is 4.5%.
If I create my own pool instead and put the money as pledge, I would expect to mint a block every 40 epochs on average. One single block reward amounts to currently ~700 ADA, so the expected annual return would be about the same as with delegation.
So why bother with running a pool? Because I also would be able to participate in a lottery for which the ticket costs nothing and the chances of winning are substantial. The @Cardano-Foundation delegation program is such a lottery. Suppose there are 1’000 eligible pools, 50 of which are chosen every three months to get a 15M delegation, so the chance for a pool to win the lottery is a whopping 20% every year.
How much is the prize? Well, for a pool with 1% margin the expected value would be 20% * 15M * 4.5% * 1% = 1’350 ADA.
This means that, thanks to the CF delegation program only, running a pool can yield more than twice as much per year for an investment of 25’000 ADA than by just delegating the same amount.
Not only that, a pool with 15M stake will advance in the rankings and attract more delegators than a small pool, increasing the profits even more.
I argue that this is the main reason why there are 2000 out of 2600 pools with 100k pledge or less. Does that pose a problem to the stability of the network? I’m not competent to judge that. However, what I can say with confidence is that running a lottery and handing out free tickets is probably not the best use of @Cardano-Foundation’s resources.
So, assuming the Foundation agrees with my reasoning, what can be done about it if the goal remains to promote a healthy, diverse SPO community from which nobody is excluded because they can’t afford a high pledge?
One option would be for the CF to delegate pledge instead of stake. Instead of being delegator, the Foundation would become temporary co-owner of the pool and be entitled to the full pool reward times their proportion of the total pledge. An SPO using this instrument would thus have to pay a price in terms of a portion of the pool rewards.
Example1: SPO pledges 25’000, CF pledges 15’000’000. Then, CF would be entitled to 99.8% of the pool rewards.
Example 2: SPO pledges 1’000’000 , CF pledges 1’000’000. In this case the rewards would be shared evenly between the two.
A variant of this could be for the CF to set a policy that they would pledge a maximum of, say, 75% of the total. So, if someone runs a pool with only 25’000, the CF could add another 75’000 to that pledge, but not more. This would encourage SPOs to cooperate by pooling their pledge in order to get more external pledge from the CF.
Some SPOs might object and say “But what about my costs? I have to do marketing”, to which I would respond that no, you don’t. Money spent on pool marketing is wasted money. Any pool with sufficient pledge and reasonable fees will attract delegators. You only need marketing if you wish to set your pool apart from countless other, equally unattractive pools. If you cooperate with others, however, everybody would win.
It is a common misconception that pool rewards should be split to the same ratio as pledge contribution of each owner in a multi-owner pool.
By only having passed all this criteria, a stakepool will automatically be included in rondomize selection?
Or an application will be required?
Thanks in advance !
Kudos to Cardano Foundation, Frederick Gregaard, and Hinrich Pfeifer
Early adopters helped build the network. They risked everything to bring this plan to life. I don’t see why those who saw the vision and acted should not be rewarded for that. Sorry…
People can always delegate and grow their ADA. They can develop on the network. They can build a Stake Pool and find investors to share the price burden. The reason for the increase protects the network from bad actors. As price goes up more and more will try to attack the network. So actions must be made to secure it. Increasing pledge is one of those actions.
Edit: Forgot to add it as a reply.
I can see where you’re coming from, a bit exagerated but sure, early investors had an advantage and that’s playing out, most early SPOs have successful pools. And I do agree, the higher the prices the more bad actors will be attracted to the network… But bad actors get advantage by having less competition, e.g. less pools minting blocks.
The idea of this program is to make more pools mint blocks, so delegation will be as distributed as possible, making it nearly impossible to most won’t protect the network. Bad actors most likely already have the capital to make their pools mint blocks.
Thank you. As the ADA price goes up, this criteria discriminates against heavy investors that have less ADA in quantity but at the same value as earlier investors. Is it intentional to make it hard for new community participants to share in these opps?
If I paid average of 1.50 fiat per ADA for 5000 units, I have significant stake at risk to pledge.
Someone else bought at .40 fiat per unit for 5000 and has far less local fiat at risk.
This diminishes the meaning of their pledge and stake, yet they have greater chance of getting SPO work offered.
Why? What am I missing? They are a riskier member of the community. Thank you!
This is a well thought out comment.
I feel the same. And I’m starting a 100k pledged pool tomorrow…
Hi @Cardano-Foundation, you mentioned that a delegation normally stays for three month. Today our Pool AZUL welcomed a CF Whale which somehow send all his ADA to another Wallet just three hours later . pool.pm
So while the ADA are gone the whale with empty pockets is still there.
Ist that “normal” during redelegation process ? I am a bit confused …
Here is some writeup of the Cardano foundation which explains it.
Bottom line: it’s based on their improved delegation process…. And the whale will come back
Glad you found the answer @jochen. Worth to mention that the ten pools that, for a short time yesterday, received a delegation which was withdrawn, will be included in the next cycle. Provided, of course, they don’t change their pool parameters.
Thnx the quick answer, and letting us know what’s caused the events .
Thanks. I’ll read this if i have not already. I appreciate your reply.
I think i am missing something in the design, because my current understanding would make it inevitable that the Cardano network would be heavily centralized long-term. I believe they are specifically architecting against that reality, but I can’t get help understanding it. I’ll keep digging! Thanks for your response.
Hey @James_Rupe - i think you may have missed my main point. I am not implying that any people long-term invested in Cardano should NOT be able to benefit from that - and you are if you were here since the beginning. This is a concern about me paying for a system to run, an SPO system that is being well-managed and professionally administered - using power and an wasting time on an asset to the decentralized network that is being unused. It’s wasteful to not have clarity in which systems have a chance to participate. There needs to be clarity. If decentralization is the goal, then only letting wealthy people joining the community run nodes works against that goal and invites bad actors to put your investment at risk. The community should let lack of skills and poorly managed systems get fleshed out but reward highly available stake pool operators that can be a part of the throughput without it all going to a central group. Hope that clarifies? My goal is not for me to make up for your long-term stake in the dev teams, etc. Thanks!
Hi guys, when will be next round of delegations from Cardano Foundation to the small Pools like ROG ?
É possível trocarmos umas ideias por email?