Announcing the stake pools chosen for November 2024

[HeptaSean]

@GrahamsNumberPlus1

True decentralization isn’t just about “enough independent pools to avoid censorship.” It’s about creating equitable opportunities for everyone to participate.

Where did you get that promise? Especially concerning stake pool operation?

@GrahamsNumberPlus1

If 66% of pools aren’t minting blocks, can we really call this decentralization? These aren’t random pools thrown in to skew stats—these are SPOs who followed the rules, invested in the ecosystem, and yet are unable to compete.

They threw themselves into the stats, yes, but it’s still not a meaningful number. It was never promised that you will be able to sustainably operate a stake pool if you just “follow the rules”. Even if they do not intend to skew the stats, it does not change if our decentralisation is “good” or “bad” if more or less small pools are operating that do not manage to attract the necessary stake to do so successfully. It’s just irrelevant.

@GrahamsNumberPlus1

Cardano was designed so anyone could run a pool if they brought value. Are we now walking back on that vision?

First and foremost, “proof of X“ uses a scarce resource “X” to ensure that not anybody can just attack and take over the network:

• Proof of work uses hash power as scarce resource. Anyone can run a miner, but if you do not invest enough hardware and energy, you are not guaranteed to produce blocks and earn rewards.
• Pure proof of stake uses the native currency of the network as scarce resource. Anyone can run a validator, but if you are not rich enough, you are not guaranteed to produce blocks and earn rewards.
• Delegated proof of stake – what Cardano does – uses delegations of the native currency by users of the system as scarce resource. Anyone can run a pool, but if you do not attract enough delegation, you are not guaranteed to produce blocks and earn rewards.

Stating that not every pool is entitled to survive is not walking back on a supposed vision. It’s just basic logic that in order to fulfil its function, it has to be hard to attract delegation, it has to be possible to fail.

Community and also CF have tried from the beginning to inform users on how to select a stake pool, to provide opportunities for pools to present themselves. There are lots of tools to view and select a stake pool. And a lot of them do manage to get enough delegation. It obviously helps if the pool is visible because it belongs to a well-known project and is known to its community, because the operator is well-known on social media or on YouTube, because it belongs to a wallet app and is more or less subtly featured in it, …

It might seem unfair that it is not enough to “just” operate a pool (which already seems so much work for some operators, especially those not well-versed in system administration … who arguably maybe just shouldn’t operate a pool?), but, no, doing the advertising and outreach is part of operating a pool. Yep, some of the things that are kind of successful are also questionable in my opinion – why should a YouTube shiller be a good SPO? – but it is how it is.

@GrahamsNumberPlus1

You defend the delegation strategy, but where’s the transparency?

In the very post we are currently discussing under? In Cardano Foundation Updates Delegation Strategy linked in that post? What more do you want? Why do you think you are entitled to more?

@GrahamsNumberPlus1

Emurgo retiring pools with 100M ADA still linked to them is a clear example of unresolved conflicts of interest. Delegators often don’t even realize they’re still tied to retired pools—this isn’t just an operational oversight, it’s a problem of communication and accountability. If founding entities don’t set a better example for decentralization, how can they expect others to?

I still honestly don’t understand what you want here. Running multiple pools is arguably bad for decentralisation, especially if it are so many as Emurgo is operating. So, retiring some of them is kind of good?

Yes, they could have announced that even more prominently (although I do remember one or more X posts and I really hope that Yoroi shows upcoming pool retirement somehow). But there simply is no way to reach delegators proactively. If they don’t follow the SPO’s announcements and don’t regularly check their wallet app because they are not that actively interested in Cardano, because it is just one among many cryptos they hold long-term, because they are dead, there is nothing Emurgo can do about it.

A good solution for this and other “sticky stake” problems would be to let delegations expire, to force people to redo and hopefully reconsider their pool delegation from time to time.

@GrahamsNumberPlus1

Similarly, CF delegations going to already large pools undermines the small operators who are critical to the network’s diversity.

Yeah, well, you know, that’s just like your opinion, man.

The design of Cardano’s reward sharing scheme “wants” it to converge to k fully saturated pools: https://iohk.io/en/research/library/papers/reward-sharing-schemes-for-stake-pools/
So, as long as a pool is not fully saturated, it should be okay to delegate to it. Could maybe even argued that it is better for this convergence to delegate to larger pools than to small pools that will likely never reach sustainability.

Is this reward sharing scheme good? I usually don’t think so. The debate about a better scheme would in my opinion be much more productive than the thousandth complaint about CF not delegating to the pools this or that SPO wants them to delegate to (mostly themselves?), the hundredth discussion about raising k in the deceitful hope that that might “fix things”, or shaming users for not doing hours of research into selecting a “good” pool, but just delegating to some huge multi-pool offering the best ROS.

What I would propose:

• As already stated above: Let delegations expire to force reconsideration, to not have delegations from lost keys, users left for good etc. pp. forever.
• Just remove “k” (stakePoolTargetNum, previously nOpt). It has been subverted by just running multiple pools from the very beginning.
• Also remove the bonus for pledge. It only really matters for fully pledged private pools, only benefits the ADA rich.
• Instead: Make pledge matter by defining a maximal leverage parameter. Saturation then is maxLeverage×pledge.
  If you want to have lots of delegation, you need to have lots of your own (or borrowed) skin in the game.
  But it kind of scales: You can start with a low pledge when attracting the first delegators and only have to raise after you have grown enough (and hopefully accumulated enough for a higher pledge in the process).
  And it incentivises transparency – large operations can still just operate one pool with all of the pledge and all of the delegation in it, there’s no advantage in running multi-pools anymore (that then have to be analysed to get a grasp on how much of a problem Binance et al. really are).
• (Alternative idea: To not have to decide on a sensible maxLeverage, we could try to let the rewards decline with higher leverage, so that delegators are incentivised to search for a pool with as low a leverage as possible, to distribute themselves over the pools. But at some point, the question arises why have delegation at all? Why not just do pure proof of stake/pledge then?)
• Remove minPoolCost (and poolCost itself?): minPoolCost never made sense since it is only paid if a block is produced at all, so it never ensured that SPOs get their operation financed. It just made small pools less competitive because a static cost is much more relevant if only few blocks are produced. It is much better to let them be competitive from the very beginning (albeit not necessarily covering the costs even with blocks produced). Maybe we don’t even need a pool margin and the rewards for the pledge are enough?
• Maybe very controversial: Introduce locking and slashing, maybe only for pledge, maybe for all delegations. If nothing is at stake, it’s not really proof of stake. Sufficiently definable bad behaviour should be punishable. If this also extends to delegators, it loses some of our selling points (“ADA are never locked! It’s totally risk-free!!!”), but it incentivises delegators really searching for a good SPO, assessing if the pool is trustworthy which they now don’t have to care about at all.

This all has little to do with CF’s delegation strategy. I still think they are free to choose what they want to prioritise and supporting outstanding contributions to the community is a very sane choice. How else should they select among the thousands of pools? Just randomly among all below 1m ADA stake?

@GrahamsNumberPlus1

Decentralization isn’t just about technical resilience; it’s also about governance, fairness, and ensuring influence isn’t concentrated in too few hands. Is this really the future we want?

This relies on your opinion that 100 pools (just an example) are “too few”. I think that can be more than enough to achieve what I want from the pools operating a cryptocurrency network. What you dismiss as “just technical resilience” is their only job. Why should I want to finance more than is necessary? Inefficiency as end in itself?

@GrahamsNumberPlus1

The idea that we can’t sustain more pools because of declining reserves ignores what mass adoption could bring. More users would mean more transactions, higher fees, and a network that could afford to sustain far more pools.

As said: Lots of moving parts.

• There cannot be that many more transactions with the current setup due to limited block space and increasing block size or reducing block time can only marginally change that. Maybe Leios fixes that. But other proposals – Hydra or other roll-up/sidechain solutions – move transactions off the main chain and, hence, lead to less transactions and fees.
• ADA’s price rising could alleviate the problem a bit, but on the other hand lead to pressure to lower the fees per transaction because users are very allergic to high fees (and we are so proud to not have Ethereum-style fees, while Ethereum is the only crypto that comes close to financing its network from transaction fees).