Dynamic Proof of Stake vs. Proof-of-Work: Advantages/Disadvantages

I wanted to created a topic where we could discuss in detail the advantages of dPOS over POW, especially when it comes to Shelley.

Here are some obvious ones, but I would appreciate if we could gather more nuanced ones, pertinent to how shelley works here so we understand and explain them to people who ask about Shelley.

ADVANTAGES:

  1. Security/Decentralization: Unlike POW, where you can ‘rent’ mining power on hourly-basis, it is expensive to attack a network based on dPOS. The attacker would have to buy 51% of all tokens in the market to have enough power to compromise the network. Downside: This could also be a risk for smaller-cap networks, which are cheaper to compromise. The deterrent here is that the community will fork the code leaving the attacker behind, making his investment worthless.

  2. Power consumption: dPOS does not require enormous amount of power to run.This makes the network accessible to average user thereby increasing the decentralization factor compared to POW, where players with significant financial power effectively run it.

DISADVANTAGES:

  1. POW is more profitable than proof of stake for now (electricity vs. income). You can create tokens that do not yet exist and sell them at peak times.

  2. In POW, when you sell proof of work tokens, you do not sell your miner, you still own the hardware. When you sell proof of stake tokens, you are selling your miner.

  3. With POW, when a token becomes less profitable, you just dump it and point your miner at a more profitable token. Proof of work is more flexible. (at least in the context of the miner).

PS: I will be updating this list as responses come in!

2 Likes

The advantages to Proof of Work are like this:

  1. It is more profitable than proof of stake for now (electricity vs. income). You can create tokens that do not yet exist and sell them at peak times.
  2. When you sell proof of work tokens, you do not sell your miner, you still own the hardware. When you sell proof of stake tokens, you are selling your miner.
  3. With proof of work, when a token becomes less profitable, you just dump it and point your miner at a more profitable token. Proof of work is more flexible. (at least in the context of the miner).

So to summarize, I would say the advantages to Proof of Work are profit, retention of resources, and flexibility. But gosh darn the power consumption is insane.

1 Like

Thanks Ricky. Never thought about tokens as “miners”. Gives a fresh perspective.

1 Like

Changed title to “Dynamic Proof-of-Stake” at it’s mentioned in the Ouroboros paper. “Delegated Proof-of-Stake” (DPOS) is a very specific kind of a protocol, and the one that it used in the platforms like BitShares, Steem, and EOS.

4 Likes

From the perspective of risk of loss of investment - disregarding the value of the coin itself:

PoW Disadvantage:

  1. When miners with significantly higher hast rate join the network all original miners with lower hash rate are essentially reduced to junk. Owners of low hast rate miners end up as bag holders with a bunch of useless miners. The new miners with significantly higher hash rate enjoy disproportionate returns until many more miners with higher hash rate join the network. Think GPU v/s ASIC miners, there is competition within ASIC miners as well. This allows for a natural monopoly over mining to manufacturers of mining equipment. They can produce better miners and use them before selling them. This defeats the whole premise on which the coin exists - “decentralization”.

PoS Advantage:

  1. There’s no way any participant of the network can unilaterally achieve significantly higher returns with the coins they own, because the returns (staking rewards) are always proportionate to the coins they own. The coins one owns do not get reduced to junk due to participation in staking of more coin owners.
2 Likes

The attacker would have to buy 51% of all tokens in the market to have enough power to compromise the network.

With delegation, the attacker need not buy 51% of all tokens; social engineering will suffice. All they need to accomplish is convince 51% of stakeholders to delegate stake to pools that they control.

There’s not yet enough detail about the final incentive schemes of Cardano. However, if percentage of staking rewards is the primary incentive for stake pool operators it won’t be prohibitively expensive for a malicious entity or a cartel to register a large number of stake pools and offer 0% fee.

GREED prevents people from thinking rationally - and I can bet a large number of stakeholders will jump on 0% fees without questioning how the provider can operate with no fees. Additionally, the risk of loss for coin owners is minimal because your coins are safe - the loss is limited to staking reward for an epoch if the stake pool provider ends up being malicious.

However, for the network an epoch (5 days) is a really long time for any kind of disruption - even in the absence of 51% attack.

1 Like

And in what are they both different essentially from one another please?!