2

I recently came across an argument of self-referentiality in PoS and I'm wondering what's the strongest counter-point to this idea?

In essence, the idea is that in PoS you

Similarly reduced, proof-of-stake systems work like this:

  1. The list of valid transactions determines who has coin.
  2. complexity
  3. People with coin decide which transactions are valid.

Is PoS actually self-referential as described here? If so, is this some critical flaw in the idea? Maybe PoS in general is self-referential like this but Ouroboros isn't?

3

I think the issues with this argument are actually pretty simple. He is wrong for the case of Cardano because people with coin get tickets to the lottery to decide who gets to create the next block. Therefore Cardano is not self-referential. He also is assuming a UTXO or eUTXO model for "The list of valid transactions determines who has a coin." Some blockchains use PoS and an account based accounting method, which makes the first point untrue.

For the second point, all blockchains are relatively complex when you get into the details. With that said, my anecdotal experience shows that it is easier to explain proof of stake than proof of work to someone with no blockchain knowledge.

If there was a proof of stake coin using eUTXO/UTXO and the protocol was as simple as 'People with coin decide which transactions are valid' then I would say yes it is self-referential. However, a protocol of that nature has even bigger concerns than that, which is why one doesn't exist. (Please correct me if one does exist)

I hope this is helpful!

1
  • The bolded statement doesn't solve the self-referentiality in my mind - even though you get tickets, only ticketholders get to decide which transactions get included. I also don't think the account vs utxo models make a difference in whether or not this self referentiality exists. Nov 18 '21 at 16:52
1

I came across this argument as well and thought it looks a bit constructed. In a PoW system, the people with processing power decide which transactions are correct. In a PoS system, it is those with wealth. Both come from outside of the system. The self referential nature described exists only at the beginning, when the PoS system is started. As soon as a coin distribution to unrelated parties took place, the system is no longer self referential if that distribution reflects external factors (like the financial ability to purchase the coins or to operate a stake pool).

2
  • Yeah I agree, I think there's an argument to be made for self referentiality in PoW as well - the miners get rewards which can be used to buy more processing power (which is essentially stake in PoS) which is then used to decide which transactions are included in the next block. I'm not sure that I follow your argument of distribution after genesis breaking self referentiality. It's still coin holders who decide which transactions get included, regardless of distribution. Nov 18 '21 at 16:54
  • PoS and PoW both distribute coins according to rules enforced by independent actors. My point was that the distribution of coins in PoS does not (only) come from the system. It comes from an external logic: coins sold for real money or distributed according to some other logic of the real world. If the coin distribution would follow only some internal logic of the system than it would be self referential. I know my point is a bit vague. I have here a strong intuition that it goes into the right direction but I’m not able to completely articulate it.
    – Jey
    Nov 18 '21 at 18:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.