Couldn’t agree more here @waldmops and @ATADA and the pledge incentivization has been mentioned previously as well. It would require adjustment to rewards/pledge formula and a HF, correct? It would undoubtedly be a game changer for small pools though and worth keeping momentum around. I also agree with @Colin_Edwards that the fixed fee should not be viewed as covering costs. It’s up to a SPO to evaluate their own resources, margin, and costs. And to his point… that does not need to be nor should it be public information
1 Like
@Colin_Edwards and the rest of the thread.
I need everybody to understand that there is a difference between K as a parameter and the actual K, actual decentralization of stake and block production, of the Cardano network. A few epochs ago I started quantifying K-effective, the actual distribution of stake across independent groups. K-effective is the metric you and Charles are looking for to quantify real decentralization performance.
https://twitter.com/DrLiesenfelt/status/1497047709123022848
The current K-effective of Cardano is about 41. There is absolutely no mathematical way to justify an increase from k-parameter of 500 to k-parameter of 750 or 1000 with the network is really at k-effective=41. In fact, if you really want the network to operate more efficiently you will propose a reduction of k-parameter to approximately 100. An increase of k-parameter is only mathematically justified when the effective decentralization (k-effective) of the real network meets or exceeds the k-parameter.
From Epoch 323:
2 Likes
Hi Martin,
I have an industry background, so I believe I know what you mean when you talk about “business is business.” However, I also believe that Web3 is different, and that we need to challenge assumptions that no longer work in order to go forward and “make the world work better for all,” as the tag line on the cardano.org home page states.
The spirit of Cardano is different than “business is business.” To be successful, Cardano requires a more cooperative approach. To be successful, Cardano also requires that the change in behaviour start from the ground up.
For example, a simple solution to the gap between building a stake pool and earning rewards would involve large pools choosing to delegate rewards to smaller pools. This does not require a change in parameters-- it requires a change of heart.
I am committed to delegating ADA to smaller pools, if and when my pool (CHG) has a stake large enough to earn rewards. I’m not in any rush to get there, and if no one wants to help along the way, then I’ll get there eventually myself anyway.
If more people decided to do the same, then I believe Cardano would grow more quickly. When the ADA market grows, then all SPOs benefit.
If you are talking about “wealth” with cardano, that is not coming from the SPO side. SPOs are not the “big fishes” in the ecosystem. Some maybe yes, but they run private pools. Running a pool is a business, for some a small one, for others a big one. TBH, we need more delegators, not more pools/operators.
4 Likes
I think this is a great point (but I hate the naming!)
I totally agree that K is not the same as a measure of network decentralization. To me, K represents the optimal size of a pool (for maximizing rewards) and it doesn’t have any other significance beyond that. So, I am not sure calling something “K-effective” adds any clarity; it’s imbuing significance and a relationship with K that isn’t there.
I don’t really follow your point about K-effective needing to be larger than K; I tend to think of network decentralization along the lines of “how much adversarial stake does one group control” (which I think you do as well), but that feels almost independent of pool size.
For example, if we had four pools - one pool of size 300 and 3 pools of size 100 with MaxSize=300, then if MaxSize moves to 100, and we have 6 pools each of size 100. In the short term, K has really very little relationship with the network decentralization.
There is some longer term impact - being near the optimal size does protect the three smaller pools. If the maxSize remained at 300, the larger pool could split into two 150 sized pools, and those two pools would offer better returns than the 100 sized pools.
Anyway, I’d love to dive into your ideas more in a different context; this looks like a really nice approach. (other than the linkage to K, which is a bit indirect at best in my mind.)
1 Like
Alot of SPOs that are against the min fee change are missing the point. I feel like the k parameter and minPoolCost come hand in hand. Say that k is increased to 1000 or 750 like we all want. Does that mean that just because k is increased that smaller pools suddenly are able to compete more? No, this just means the desirable amount of pools is increased. What’s the incentive then for more small pools to be delegated to? The fact that the parameter wants it? The golden pot of ₳ will not last forever.
If you want to see the network expand then it will need to move to more sustainable models of rewarding operators. At some point, 340 ₳ is going to be too expensive and essentially unnecessary.
1 Like
yep exactly, thats the illusion of many. just because of an k increase out of nowhere delegators coming to “your” pool. they are not. delegators are loyal, if an spo is doing a good job on updating its delegators with news, answers questions, etc. they most likely stay with them. and without an incentive also to pools with higher pledge to not split up, we will just see more multipools than ever before. if someone is having currently for example 60M in the pool and k increases to 750, well splitting up into two pools and try to hold as many delegators as possible will be the game. but that is not changing anything for the tiny pool with 20k in delegation. i am a operator supporter, trying to help each and every one of them that are having problems on the technical side. but, just being able to setup a cardano-node instance will not automatically bring in delegators and rewards. cardano is a “DELEGATED proof of stake” blockchain, not proof of work. so the person behind a pool that people are actually delegating to should be more in the focus.
k does not represent the number of pools that are possible on cardano, it is the target number of max. SATURATED pools on the chain. we have over 3000 pools on cardano live, more than 1000 of them have made a block so far in this epoch. those who will not make more than 1 block in this epoch are super happy with the current 340 minCost parameter, because its giving them a decent amount of rewards to start with.
2 Likes
The proposed changes will not improve the decentralization and reliability of the network.
The minCost and K parameters are meaningless.
The pledge should affect the possible level of saturation. The pledge should be blocked for the period of the KES certificate.
2 Likes
I absolutely agree with ATADA.
Reducing minFeeCosts will have hardly any inpact on a delegators pool choice. There is no incentive for them to leave a big pool for a small pool when both reduce their minFee ?
And reducing minFee will hurt small pools a way more than big pools so it’s very likely that big pools will win the race to the bottom.
Raising k without adjusting a0 will cause SPO splitting their pools. We have seen this happening in the past. If we want a decentalized ecosystem an operators pledge should have a higher impact on the pool rewards.
So I fear the proposed parameter changes will wipe out a lot of small pools and make the big one even bigger.
That’s not true, right now a delegator is giving up half their rewards to delegate to a small pool. If your pool can’t start at 10 MM, you have no way to organically grow your pool by offering similar rewards as larger pools. That number is an order of magnitude higher than it needs to be.
We aren’t that far off small pools being unable to offer delegators any rewards (mid-2026, given my projections which include things like growth of fees, increases in on-chain activity.)
4 Likes
@Colin_Edwards:
I think this is a great point (but I hate the naming!)
I will happily call it whatever you want! Ask the IOG team for an appropriate and accurate name.
it’s imbuing significance and a relationship with K that isn’t there.
There should be a very clear cause and effect relationship which doesn’t currently exist.
other than the linkage to K, which is a bit indirect at best in my mind.
The k parameter is currently 500. The actual network decentralization is currently 41. You ideally want 500 equal pools. Currently the network is behaving as if there were 41 equally sized pools on average with extreme tails. The expected versus actual results aren’t even in the same half an order of magnitude.
I propose reducing the minFeeCosts to 30, use pledge amount as a real parameter which matters, setting the k-parameter back to 100 and let the multi-pools save electricity improve relay speed. Then, we can continue quantifying real network decentralization so we can mathematically justify an increase once real decentralization is closer to k=100.
BTW, I love being part of the Cardano community and look forward to a future of growth!
4 Likes
This is true, but also remains true for the Multi pools to split when you change K. That’s the issue here.
They still have the same ability to change their fee to the minimum if needed to retain the current delegation to “maintain their brand”. Hell lets be honest the majority of the multi pools already charge higher fees on their clusters, why do you assume their delegates will just randomly choose another pool with the change of K? They have only grown and expanded each K change, this leaves the same available stake to be distributed to “starter” pools.
There has to be a reason which incentivizes the operators NOT to expand exponentially if the intention is to control a Sybil attack or make the network “fair” you cant just focus on “fixed fee” that’s not the only reason starter pools have little chance to expand. There has to be available active delegation to be gained as well. (not just assumed)
This does not incentivize people to move to other pools outside of the multi pool operation with change of K, again this continues the same previous pattern of K changes and allows them to retain the advantage and continue to over-leveraging the system due to lack of enabled pool expansion limiters.
I’m not saying pledge is the answer this just causes another pain point, but the system was built to function WITH pledge not without.
So are you saying pledge is not required as that is a entry barrier as well?
If so then another mechanism is needed to take it’s place to limit the pool expansion and ability to over-leverage the network.
We should be working towards a complete solution not just playing with the band-aid solution that was added before Shelley launch or add another one in the short-term to kick the can down the road again.
1 Like
Good point… delegators will come.
Businesses can exist for other reasons than purely maximizing profit. Forming a business to act in ways that benefit society as a whole, with a belief that the purpose of a company is not just profits, but also social and environmental good, does not necessarily mean operating at a loss.
I agree with this proposal wholeheartedly. Contrary to what most people think, such a reduction in K will actually provide an advantage to small pools and single pool operators at the expense of the large (currently multi-pool) operators. Here is why:
The larger the pool is, the larger is its VRF score on average for the blocks it produces. Single block forks and slot battles are determined by which pool has the lower VRF score for its block. Getting multi-pool operators to consolidate to one large pool is therefore an advantage to all smaller pools in the fork / slot battle competition. It might only be a few percentage points, but it is something.
2 Likes
First of all, thank you for your clear and open approach in the call and here in the forum.
I can’t help but point out something else here that may be the real core of the problem.
Without praising myself here, I have to say that I as ITN pool operator was invited to a meeting in London in February 2020 - several months before the mainnet launch - where I tried to explain vehemently and also in the weeks afterwards why a min-margin would be more important and correct and why the min-fee will lead to problems. Most of what we see now was there, including a concrete proposal how the min-fee could be conceptualized. It was not taken into account.
Just as little as proposals for the implementation of an item registry (pool ticker, nTokens, sContracts, …) that are operated in a decentralized manner. Ignored… and instead implemented this SMASH server.
Now I don’t want to say that everyone else is always right with every idea. But after almost two years my conclusion is that too much stoic, and MUCH too little cooperative and open for other ideas is acted.
CIPs are a good idea, but if they end up like what we have seen in 9 out of 10 (at least) cases, practically stuck or in an endless loop then it’s “not good” (trying to not use a more direct wordings).
Especially when we can see stuff like max blocksize changes happening out of nothing (aka behind closed door deals) and the same (6 char) entity manages with just a commit and within days to increase max native-token asset name length to 9, while other ecosystem members spent weeks and months to argue why it should become more than 4 characters back in 2021.
Could write up kilobytes of cases and examples here. But that would go a bit off topic.
A general issue is: the clock is ticking. Slots and epochs happen and become history. Delaying or ignoring stuff for too long time is an issue and leads the ecosystem and facts, such as the pool landscape, into certain states. For some of them - I fear - it is not more a question if parameter or ledger change is required, but if they are fixable at all and anymore.
5 Likes
I’d like to echo what some others have said @Colin_Edwards and say I really appreciate your engagement here and on the SPO call. With that said, with this discussion about parameter changes we inevitably come to the topic of multi pools and pledge and I have to say it’s really disappointing that we haven’t heard updates from the incentives team on either of these last two calls.
Reducing the min fee levels the playing field for smaller pools trying to bootstrap but it doesn’t seem like it addresses stake stagnation. Increasing k should technically force delegators to make a new choice about their delegation, however there is a good chance that choice will simply be moving to the new additional pool their operator opens. With radio silence from the incentives team it’s difficult to justify waiting and seeing what happens with this issue. I’m curious if there are reservations from IOG about tuning the a0 parameter, it seems it was plucked out of thin air when shelley went live and is just going to stay that way at this point. I think even demonstrating a willingness to adjust it could yield benefits, even if the change is small.
I want to believe to believe that if each of us asked ourselves “Is the network currently as decentralized as we’d like?” that the answer would be a resounding “no”. If we ask ourselves this question on a regular basis and tune accordingly with the a0 parameter, then it may not go as poorly as some suspect. By that I mean we may not need to significantly increase a0 to get the network to a better state, we simply need to show a willingness and the ability to do so, and we can decrease it slowly if we achieve the results we’re seeking.
As I’ve said earlier in this thread, I think reducing the min fee to 30 gets us to a much better place than we are now. I would like to know if there is any appetite for a small increase to a0 to go along with the potential increase of k. I’m curious of your opinion on this Colin as well as everyone in this thread who has strong feelings about it. I’m also curious of IOG’s opinion on this but I think it’s safe to say they are playing their cards close until the incentives team comes out to play.
3 Likes
That is only partly correct, when you do not consider whale delegators (>~2m stake).
It also gives the bootstrap problem, like catch-22, when all small pools (almost ~94% of current pools) would suffer to bootstrap their pool as they must have a certain size (in pledge or initial pledge+stake) to attract delegators (provide some rewards at all).
I do think and it seems to me the opposite, because all small pool suffer the bootstrap problem, and some of them can attract the delegators by some marketing, fooling, promises (what we see) as the pools can earn much higher ROI then they should. E.g. investing 10K to pledge attract delegators through media etc. they will have ~240% ROI per pool (even when considering their cost), and we see similar trend in pools.
Look at the popular pools. Even with 10K pledge and 99% margin attracted huge amount of delegators earning 35K per epoch, and they’re slowly growing in numbers started as 1 pool, then 2 and now 4.
As I said earlier, hundreds of times, the ideal should be: pool nr. per entity = ⌈available total stake for pledge/saturation point⌉
⌈x⌉, means ceiling
We can smashing around, having opinions that we like, believe and feel in our gut (based on our limited understanding), but first we need to define what we want to and how to achieve it.
The RSS game assumed rational delegators, which does not seem to me the case and it makes sense if you see the social media as a mirror of our behavior as species. Though some individuals are, but the mass seems very far from it.
Ooooh, we can educate them then. Same applies, some would, but the mass. So, I am not sure.
5 Likes
I totally agree with your point, in my opinion it’s completely useless to talk about changing parameteres if first we don’t implement the science as it was designed in the research papers (with an effective pledge mechanism).
Changing the parameters (k and MinFixedFees) will only result in an even bigger prolification of multipools, concentrating even more power in the hands of few big actors.
1 Like
This is the wrong way to think about it. Try to think about it from a delegator’s perspective and how a small pool can grow by being more competitive.
As a delegator, I will not delegate to a pool with less than 10M stake already, otherwise the current 340 min fee will take too much of my rewards. Consequently small pools with less than 10M stake have little chance to grow.
If this min fee is reduced to 30 then a much lower percentage of my delegator rewards are taken by the min fee. With a min fee of 30, I would be prepared to delegate to pools with 800K total stake.
Yes, the multi-pools will increase their number of pools if K increases to 750, but this irritates their delegators through having to rebalance pools and also forces their delegators to reasses. It also increases multi-pool costs, slightly through running more pools and having to actively shepherd their delegators between them. Both are negatives for multi-pools, even if only slight.
So the changes that Colin suggest will benefit small pools by reducing the disincentive to delegators using small pools. The changes will also slightly irritate and slightly increase the costs for the multi-pools.
In the short term, this is a win win for small pools.
We can continue to work on changing the pledge benefit and other factors in parallel, with a longer term fix in mind.
2 Likes
@7.4d4 thanks for your reply. IMHO yours it’s the wrong way to look at it. As a delegator, you’re also an investor in the project: the strenght and value of Cardano is it’s decentralization, without it the value of the project would be way lower, so we should all think about preserving and increasing it. Your way of thinking about it’s a demonstration that incentives towards a useful delegators behaviour are off the right path. The wrong reasoning it’s that lowering MinFees = delegators flowing towards small pools: this in my opinion won’t be the case, without adjusting pledge incentives first.
Also, the hardware requirements for running nodes are always increasing and it’s in everyone’s interest to have validators running good infrastructures if we really want the project to succeed in the long run: with ada price decreased 70% in the last 7 months, lowering fees now don’t seem the right timing to me.
I’ve run a small stake pool since before the ITN and my delegators always had average or above average rewards, but with increased costs and requirements I’m starting to think if it’s still worth. You can try do the math and you’ll see that on 100k ada delegated, the difference it’s not that big at the end, just few ada.
I think that MinFixedFees it’s what kept alive small and medium sized pools so far, without that we’ll see many many pools retiring, while big multipools actors will spin up even more pools with low pledge (the same thing they did with last K increase).
Lastly, proof of stake rewards everyone in a more democratic way than proof of work, a delegator shouldn’t be irritated by having to redelegate once in 18-24 months due to K increase 
active participation it’s the core of the consensus mechanism
1 Like