5 ms·

One of Ethereum's core goals is decentralisation. How does Ethereum plan to deal Proof of Stake naturally monopolizing block creation and the Ether supply? In

One of Ethereum's core goals is decentralisation. How does Ethereum plan to deal Proof of Stake naturally monopolizing block creation and the Ether supply? In my estimation there are many compounding factors such as MEV and liquid staking with a massive economy of scale for first movers that combined with staking interest might make the top staking provider eventually hold the vast majority of Ether. If the company that runs Lido is responsible for validating 99% of the blocks and the US Treasury Department comes knocking with a list of bad actors to blacklist, what happens next?

rcxdude 4y ago
It's the same forces which push proof of work to the same conclusion, so it's not made much difference to the problem.
- Salgat 4y ago
  The difference is that if a miner does nothing, eventually they stop making money regardless of how many coins they have. If a staker does nothing, they continue to collect and collect even more indefinitely with no additional investment beyond running a very cheap node (or just letting someone else host it). Staking is just trivial in comparison, it's apples and oranges.
  - yokem55 4y ago
    The flip side of this though is that stakers can have their capital burned because of misbehavior. If a staker gets too entrenched and the other folks in the network don't want them around anymore, their stake can be slashed. Proof of work cannot destroy the equipment of a hostile miner, only wipe everyone's equipment out with an algorithm change.
    - jimmydorry 4y ago
      If a staker gets too entrenched, by what mechanism will the minority slash the majority staker's stake?
    - doomroot 4y ago
      There is no such thing as a hostile miner, what do you think proof of work means?
  - ruuda 4y ago
    > If a staker does nothing, they continue to collect and collect They do, but the supply of the token also goes up, so their share of the supply doesn’t grow as much, and relative to other stakers, it doesn’t grow in proportion at all. It is true that inflation rewards effectively move value from non-stakers to stakers, but nothing prevents most of the supply from being staked, especially with liquid staking derivatives. > Staking is just trivial in comparison I agree that staking is, but operating the node is not! Operating a validator is actually more demanding than operating a miner. For a miner, there is no disadvantage in being offline except for the opportunity cost. But a validator has a duty, and will be penalized for downtime. (Granted, the requirements on Ethereum are quite lax — on purpose, to make it easier for enthusiasts and other non-professionals to operate a node.) Operating a validator requires monitoring and maintenance like any other software. Your server will run out of disk space, your node will disconnect if you don’t update the node software in time, etc. Most validators in most of the PoS networks are operated by professional companies with dedicated SRE teams because of this.
    - Salgat 4y ago
      For a major staker, are you saying that running a validator is equivalent in work/cost to running a warehouse full of miners? Remember, we're talking about major players, not just some random guy doing this in his closet.
PretzelPirate 4y ago
I don’t have a source handy, but eventually block production will centralize and the protocol will allow a “block proposer” role which will require lower resources. The proposer will be able to force a producer to include specific transactions and if they aren’t included, the producer will be punished (via their stake being slashed). If a centralized entity is told to include transactions by a proposer and they refuse for any reason, they will eventually lose their stake and fall out of the producer role.
redox99 4y ago
The problem of pools getting too big (Lido, CeXes, etc) is worrying. I think once withdrawals are enabled (shangai fork), liquid staking won't be that attractive, so the situation will improve.
sophrocyne 4y ago
Your note on Lido being responsible for validating 99% may have been hyperbolic, but just for reference - https://beaconcha.in/pools https://beaconcha.in/pools There's an argument to be made that there is a centralizing force in the role of the consensus/security layer for the chain, because the asset being earned (ETH) can be staked and earn further returns - However, this role and phenomenon is mirrored in the PoW world. The difference is that the centralization happens one step removed from the on-chain asset - PoW miners consolidate profit into further mining investments, such that an increasing amount of the hash rate is owned by the largest miners, who can acquire improved access to electricity and/or equipment relative to smaller miners. One could argue that a PoS system actually has less room for exploitation, however, since you can't restrict a solo-stakers access to ETH, while you could restrict (or provide severe barriers to entry) on the competition of electricity/equipment. The concern of the "top staking provider eventually holding the vast majority of Ether" is highly unlikely. Given that the asset will soon have ETH issuance cut by 90+% as part of the merge, and the fact that there is no mechanism by which top staking providers are incentivized more than the small solo-stakers, this would soon enter the realms of the purely theoretical, and seemingly take lifetimes to happen if one could even envision it happening at all without a reallocation into other investments. MEV is being democratized as well (whether a good thing or not, will let you be the judge), with even solo-stakers being able to use an MEV client alongside validation clients in order to benefit from additional income on block proposals (see Flashbots mev-boost client, releasing alongside the merge). Would love to better understand whether that better informs your perspectives on the subject, or what other concerns you still have.
- game-of-throws 4y ago
  In PoW, mining has real-world ongoing costs (electricity, hardware) and so miners need to continually liquidate their earnings to stay afloat. This counterbalances the effect of wealth centralization. If you try to remove real-world costs from that equation, you lose the counterbalancing effect.
  - eterps 4y ago
    > This counterbalances the effect of wealth centralization Yes, this ensures a limited supply with PoW, right? > If you try to remove real-world costs from that equation, you lose the counterbalancing effect. I assume you can't remove real-world costs in the case of PoS, but yes I can see how the real-world costs will be much, much lower. Are you saying that there is a good reason to keep those real-world costs higher? (pardon my ignorance, I haven't looked into the principle you describe here before).
ruuda 4y ago
> If the company that runs Lido is responsible for validating 99% of the blocks and the US Treasury Department comes knocking with a list of bad actors to blacklist, what happens next? The company that employs most of the Lido contributors is based in a region where the US has little influence. But aside from that, the Lido node operators (who operate the validators) consist of more than a dozen companies, registered in various countries, using servers in different locations. Diversity of jurisdictions is an explicit criterion for node operator selection. Possibly some of them could be forced to censor transactions, but I think it’s not the lowest hanging fruit (going after the client software developers might be a more impactful avenue, there are fewer of those than validators, though fortunately Ethereum has multiple client implementations). But suppose somehow 99% of the blocks are produced by validators who censor certain actors. Then it depends a bit on how far-reaching the censorship is. If validators refuse to include certain transactions in their blocks, but still accept other’s blocks for consensus, then I expect that blacklisted actors will find it more difficult and more expensive to get their transactions included, but it would still be possible. If they offer a juicy transaction fee, they will have to wait on average 50 blocks to get their transaction in, but then a validator will be more than happy to include it. If validators refuse to include certain transactions in their blocks, and also refuse to vote on blocks that do for consensus, then they will be able to enforce their censorship. One way to think about this, is that the censorship has been built into the protocol, and the 1% of validators still willing to include the blacklisted transactions, are producing what the others consider invalid blocks. Consensus can tolerate 1/3 of the stake misbehaving. 99% is an extreme case though; if the censorship is limited to less than 1/3 of the stake, then the opposite happens. Validators who refuse to vote on certain blocks don’t fulfil their duty of voting, which incurs a penalty. If the validator is not allowed to vote on even a descendant of a block that contains a blacklisted transaction, then effectively it will be prevented from ever voting again, and its stake will slowly evaporate due to penalties. An interesting case arises when censoring nodes control more than 1/3 of the stake, but less than 2/3. This could result in a liveness failure, where no blocks get finalized (because neither the censoring nor the non-censoring nodes can get the required 2/3 majority). This triggers an “inactivity leak”, where the stake of the misbehaving nodes gets destroyed, until there is less than 1/3 of it left, and consensus can be achieved again. Of course, who is “misbehaving” here depends on your point of view, and the situation is symmetric, so this would lead to the chain forking into one where the censors have a supermajority, and one where the non-censors have a supermajority. The side which had most stake will be the first to achieve consensus again. (Because of this risk, there is an ongoing discussion in Lido about whether it should self-impose a limit to not control more than 1/3 of the stake. But it’s a tough situation, because the way it looks right now is that that stake would instead be controlled by centralized exchanges who offer staking services.)
pa7x1 4y ago
What pushes towards centralization are two factors; economies of scale and barriers of entry. For those mathematically inclined you can imagine plotting economic reward (y-axis) vs economic input (x-axis). Economies of scale determine the shape and growth of the curve while barriers of entry determine the cut with the y-axis. Obviously, the greater the barriers of entry or the higher the reward as your economic scale grows the bigger the push towards centralization. In the first case because you forbid small actors to participate in the game, in the second case because you reward bigger actors more generously. So how does this graph look for Ethereum? Pretty simple, if you have more than 32 ETH it's basically flat. You get the same APY irrespectively of your size. And if you have less than 32 ETH? Well, you can then stake with RocketPool (a decentralized staking pool) in which case your APY is 0.85 the full APY. So the graph for Ethereum is: - 0.85 * APY between [0 ETH, 32 ETH) - APY between [32 ETH, infty ETH) Where APY is the yield returned by the network which depends on total amount staked in the network and network fee revenues. This is a remarkably flat curve, which highlights that there are almost non-existent economies of scale in PoS as designed in Ethereum. If you do the same analysis for PoW you will find it requires significant investment in specialized HW (either top of the line GPU or ASICS), and there are significant economies of scale in the form of access to cheap or unusable sources of energy.
- anonymousDan 4y ago
  Interesting. Do you have any references to resources describing the theory of how economies of scale/barriers to entry interact?
- ruuda 4y ago
  > You get the same APY irrespectively of your size On-chain, yes. But there is an off-chain cost to operating the validators, and there is economy of scale there. You can run many 32-ETH validators on a single machine using almost the same resources as running a single one, so the amortized cost of the hardware goes down. And when you do need to expand to multiple machines, the same applies; you don’t need 10× the people to manage 10× the machines. Of course, you can pay somebody to operate the validator for you, and those parties benefit from economy of scale.
Yizahi 4y ago
The core goal is neo-feudalism and infinite rent seeking. Which totally in line with all token developments we see today.

rcxdude 4y ago

It's the same forces which push proof of work to the same conclusion, so it's not made much difference to the problem.

Salgat 4y ago
The difference is that if a miner does nothing, eventually they stop making money regardless of how many coins they have. If a staker does nothing, they continue to collect and collect even more indefinitely with no additional investment beyond running a very cheap node (or just letting someone else host it). Staking is just trivial in comparison, it's apples and oranges.
- yokem55 4y ago
  The flip side of this though is that stakers can have their capital burned because of misbehavior. If a staker gets too entrenched and the other folks in the network don't want them around anymore, their stake can be slashed. Proof of work cannot destroy the equipment of a hostile miner, only wipe everyone's equipment out with an algorithm change.
  - jimmydorry 4y ago
    If a staker gets too entrenched, by what mechanism will the minority slash the majority staker's stake?
  - doomroot 4y ago
    There is no such thing as a hostile miner, what do you think proof of work means?
- ruuda 4y ago
  > If a staker does nothing, they continue to collect and collect They do, but the supply of the token also goes up, so their share of the supply doesn’t grow as much, and relative to other stakers, it doesn’t grow in proportion at all. It is true that inflation rewards effectively move value from non-stakers to stakers, but nothing prevents most of the supply from being staked, especially with liquid staking derivatives. > Staking is just trivial in comparison I agree that staking is, but operating the node is not! Operating a validator is actually more demanding than operating a miner. For a miner, there is no disadvantage in being offline except for the opportunity cost. But a validator has a duty, and will be penalized for downtime. (Granted, the requirements on Ethereum are quite lax — on purpose, to make it easier for enthusiasts and other non-professionals to operate a node.) Operating a validator requires monitoring and maintenance like any other software. Your server will run out of disk space, your node will disconnect if you don’t update the node software in time, etc. Most validators in most of the PoS networks are operated by professional companies with dedicated SRE teams because of this.
  - Salgat 4y ago
    For a major staker, are you saying that running a validator is equivalent in work/cost to running a warehouse full of miners? Remember, we're talking about major players, not just some random guy doing this in his closet.

PretzelPirate 4y ago

I don’t have a source handy, but eventually block production will centralize and the protocol will allow a “block proposer” role which will require lower resources. The proposer will be able to force a producer to include specific transactions and if they aren’t included, the producer will be punished (via their stake being slashed). If a centralized entity is told to include transactions by a proposer and they refuse for any reason, they will eventually lose their stake and fall out of the producer role.

redox99 4y ago

The problem of pools getting too big (Lido, CeXes, etc) is worrying. I think once withdrawals are enabled (shangai fork), liquid staking won't be that attractive, so the situation will improve.

sophrocyne 4y ago

Your note on Lido being responsible for validating 99% may have been hyperbolic, but just for reference - https://beaconcha.in/pools https://beaconcha.in/pools There's an argument to be made that there is a centralizing force in the role of the consensus/security layer for the chain, because the asset being earned (ETH) can be staked and earn further returns - However, this role and phenomenon is mirrored in the PoW world. The difference is that the centralization happens one step removed from the on-chain asset - PoW miners consolidate profit into further mining investments, such that an increasing amount of the hash rate is owned by the largest miners, who can acquire improved access to electricity and/or equipment relative to smaller miners. One could argue that a PoS system actually has less room for exploitation, however, since you can't restrict a solo-stakers access to ETH, while you could restrict (or provide severe barriers to entry) on the competition of electricity/equipment. The concern of the "top staking provider eventually holding the vast majority of Ether" is highly unlikely. Given that the asset will soon have ETH issuance cut by 90+% as part of the merge, and the fact that there is no mechanism by which top staking providers are incentivized more than the small solo-stakers, this would soon enter the realms of the purely theoretical, and seemingly take lifetimes to happen if one could even envision it happening at all without a reallocation into other investments. MEV is being democratized as well (whether a good thing or not, will let you be the judge), with even solo-stakers being able to use an MEV client alongside validation clients in order to benefit from additional income on block proposals (see Flashbots mev-boost client, releasing alongside the merge). Would love to better understand whether that better informs your perspectives on the subject, or what other concerns you still have.

game-of-throws 4y ago
In PoW, mining has real-world ongoing costs (electricity, hardware) and so miners need to continually liquidate their earnings to stay afloat. This counterbalances the effect of wealth centralization. If you try to remove real-world costs from that equation, you lose the counterbalancing effect.
- eterps 4y ago
  > This counterbalances the effect of wealth centralization Yes, this ensures a limited supply with PoW, right? > If you try to remove real-world costs from that equation, you lose the counterbalancing effect. I assume you can't remove real-world costs in the case of PoS, but yes I can see how the real-world costs will be much, much lower. Are you saying that there is a good reason to keep those real-world costs higher? (pardon my ignorance, I haven't looked into the principle you describe here before).

ruuda 4y ago

> If the company that runs Lido is responsible for validating 99% of the blocks and the US Treasury Department comes knocking with a list of bad actors to blacklist, what happens next? The company that employs most of the Lido contributors is based in a region where the US has little influence. But aside from that, the Lido node operators (who operate the validators) consist of more than a dozen companies, registered in various countries, using servers in different locations. Diversity of jurisdictions is an explicit criterion for node operator selection. Possibly some of them could be forced to censor transactions, but I think it’s not the lowest hanging fruit (going after the client software developers might be a more impactful avenue, there are fewer of those than validators, though fortunately Ethereum has multiple client implementations). But suppose somehow 99% of the blocks are produced by validators who censor certain actors. Then it depends a bit on how far-reaching the censorship is. If validators refuse to include certain transactions in their blocks, but still accept other’s blocks for consensus, then I expect that blacklisted actors will find it more difficult and more expensive to get their transactions included, but it would still be possible. If they offer a juicy transaction fee, they will have to wait on average 50 blocks to get their transaction in, but then a validator will be more than happy to include it. If validators refuse to include certain transactions in their blocks, and also refuse to vote on blocks that do for consensus, then they will be able to enforce their censorship. One way to think about this, is that the censorship has been built into the protocol, and the 1% of validators still willing to include the blacklisted transactions, are producing what the others consider invalid blocks. Consensus can tolerate 1/3 of the stake misbehaving. 99% is an extreme case though; if the censorship is limited to less than 1/3 of the stake, then the opposite happens. Validators who refuse to vote on certain blocks don’t fulfil their duty of voting, which incurs a penalty. If the validator is not allowed to vote on even a descendant of a block that contains a blacklisted transaction, then effectively it will be prevented from ever voting again, and its stake will slowly evaporate due to penalties. An interesting case arises when censoring nodes control more than 1/3 of the stake, but less than 2/3. This could result in a liveness failure, where no blocks get finalized (because neither the censoring nor the non-censoring nodes can get the required 2/3 majority). This triggers an “inactivity leak”, where the stake of the misbehaving nodes gets destroyed, until there is less than 1/3 of it left, and consensus can be achieved again. Of course, who is “misbehaving” here depends on your point of view, and the situation is symmetric, so this would lead to the chain forking into one where the censors have a supermajority, and one where the non-censors have a supermajority. The side which had most stake will be the first to achieve consensus again. (Because of this risk, there is an ongoing discussion in Lido about whether it should self-impose a limit to not control more than 1/3 of the stake. But it’s a tough situation, because the way it looks right now is that that stake would instead be controlled by centralized exchanges who offer staking services.)

pa7x1 4y ago

What pushes towards centralization are two factors; economies of scale and barriers of entry. For those mathematically inclined you can imagine plotting economic reward (y-axis) vs economic input (x-axis). Economies of scale determine the shape and growth of the curve while barriers of entry determine the cut with the y-axis. Obviously, the greater the barriers of entry or the higher the reward as your economic scale grows the bigger the push towards centralization. In the first case because you forbid small actors to participate in the game, in the second case because you reward bigger actors more generously. So how does this graph look for Ethereum? Pretty simple, if you have more than 32 ETH it's basically flat. You get the same APY irrespectively of your size. And if you have less than 32 ETH? Well, you can then stake with RocketPool (a decentralized staking pool) in which case your APY is 0.85 the full APY. So the graph for Ethereum is: - 0.85 * APY between [0 ETH, 32 ETH) - APY between [32 ETH, infty ETH) Where APY is the yield returned by the network which depends on total amount staked in the network and network fee revenues. This is a remarkably flat curve, which highlights that there are almost non-existent economies of scale in PoS as designed in Ethereum. If you do the same analysis for PoW you will find it requires significant investment in specialized HW (either top of the line GPU or ASICS), and there are significant economies of scale in the form of access to cheap or unusable sources of energy.

anonymousDan 4y ago
Interesting. Do you have any references to resources describing the theory of how economies of scale/barriers to entry interact?
ruuda 4y ago
> You get the same APY irrespectively of your size On-chain, yes. But there is an off-chain cost to operating the validators, and there is economy of scale there. You can run many 32-ETH validators on a single machine using almost the same resources as running a single one, so the amortized cost of the hardware goes down. And when you do need to expand to multiple machines, the same applies; you don’t need 10× the people to manage 10× the machines. Of course, you can pay somebody to operate the validator for you, and those parties benefit from economy of scale.

Yizahi 4y ago

The core goal is neo-feudalism and infinite rent seeking. Which totally in line with all token developments we see today.