I thought I would clear up some of the most common misconceptions about PoS. I'll basically make this a two piece explanatory, first I'll compare PoW and PoS and go into some details, then I'll answer some of those common misconceptions below. So you have the option to just see my answers below or read the entire thing and understand why I answered the questions like I answered them.

Understanding

What's the difference between PoW and PoS?

Both systems are intended to form consensus on the current state of the chain. Basically in a decentral project you have to come to some kind of consensus on things. Who gets to include transactions, what is the absolute order of transactions on the chain to make sure one coin is not spent double, which chain is the legitimate one and which one is the attacker chain. So these consensus frameworks lay down the rules that everyone who wants to participate have to follow.

PoW

In PoW this is done via computing power and energy. A computer that finds a solution to a very difficult mathematical problem first has the ability to create a block, include a bunch of transactions in it, secure the chain and its history. And of course profit from it in the form of a reward, some crypto currency for finding the block called "block reward" and the transaction fees for those transactions included in your block. The mathematical problem is so hard to solve, that the best known way to solve it is just guessing it (while it's hard to solve, it's super easy to verify if your solution is correct). That's what mining basically is, using a ton of electricity and hardware to guess a solution to a really hard problem that has no other use than securing the chain.

So in order to participate in PoW you need a lot of graphic cards or even better - specialized hardware called ASICs (ASIC companies is another whole new topic, I just like to mention that they usually "test" their hardware thoroughly before they send it to their customers months later. You can't really produce your own ASICs which is why there's a constant call to open source these blueprints and which is why a lot of other crypto currencies chose mining algorithms that are ASIC resistant to strengthen their projects decentralization). These things cost real world money, need electricity and lots of it, need a space to put them, produce heat and wear down or brick over time. Ideally you want to mine somewhere where electricity and cooling is cheap.

On average a block is found every 13 seconds and earns you 2 ETH + transaction fees. Now when more miners with more hardware join, blocks would be found much quicker than those 13s average, so we introduce something called the "difficulty" that adjusts the work that a computer has to perform so we are back to an average block creation of 13s. Basically if more miners start mining, it becomes more difficult for everyone else. So in order to keep your current influence on the chain and your current reward level, you'd have to reinvest in more hardware.

The protocol dictates that you always follow the chain with the most work put into it, so if you decide to attack PoW you need at least 51% of the networks computing power to pull of an "51% attack". While this attack does not allow you to spend money that isn't yours, this attack does allow you to revert the chain’s history. To give you an idea, you could sell 1000 ETH for dollars and and then just revert your transaction and have both your initial 1000 ETH as well as the dollars you earned from the sell.

So in short: Money translates to hardware and electricity. Translates to rewards and influence.

PoS

Now in PoS we have a system where you don't need hardware or specialized devices, only an ordinary computer and ETH. ETH is still paid with real world money so both systems ultimately depend on money. Instead of needing fiat money to buy mining equipment and electricity, you need fiat money to buy ETH. So the two systems are very similar in these regards, PoS just cuts out the middle man of having to buy computing power.

A stake of 32 ETH makes you a so called "validator". Instead of solving a complicated problem, in PoS a random random validator is chosen to be the block creator every 12 seconds (randomness on a decentral system is a really interesting topic, so if you are intrigued by this you can learn more here: https://eth2.incessant.ink/book/06__building-blocks/02__randomness.html#enter-randao). Note that while in PoW this was on average 13s (there could be 8 blocks in a minute and then just 1 block in a minute), PoS works like a clockwork where a new block is due exactly every 12 seconds. While the block creator earns the block reward and transaction fees like in PoW, those rewards are significantly lower. Right now you'd get just 0.02882 ETH per block instead of 2 ETH in PoW. But the validator has more duties than just blocks. 32 blocks are whats called an Epoch. Every epoch your validator is assigned a random block to attest to. Attesting basically means checking the block and their proposer for validity and signing of with a mathematical signature. You can think of it as basically vouching with your 32 ETH stake that this block of someone else is correct and the proposer followed all the rules laid out by the protocol. You earn some 0.00001 ETH ETH per attestation. Currently a validator makes around 0.1 ETH per month, which is way less of what you'd be making if you spend 32 ETH on mining hardware and start mining with it.

But since your expenses have been cut down significantly compared to PoW (remember no heavy electricity consuming devices, no need for cooling or even big physical spaces, no real wear over time except maybe your SSD, you could basically run multiple validators on one single computer that is less powerful than your phone from 4 years ago) those lower rewards actually make a lot of sense.

Since we learned that a validator has duties every epoch (and an epoch happens every 6.4 minutes), your validator has to be online a lot though. Missing one of these duties during downtime will get you penalized. Now those penalties are very low though. You roughly loose the amount of ETH you would have earned in this epoch. So if an attestation earns you 0.00001 ETH, you'd loose 0.00001 ETH if you missed one. As long as you are online for more than 50% of the time, you'll turn a profit.

Since block times on PoS are fixed at 12s and the proposer is chosen at random, the difficulty system like we saw in PoW is not really needed here. But you are still competing against other validators in two things:

1) if more validators join, your chance to be chosen for a block proposal shrinks accordingly. And with it your annual return.

2) PoS has an incentive curve. Now what this means is, if Ethereum has too few validators, the rewards for validators increase. If Ethereum has too many validators, the rewards for validators decrease. You want to incentivize enough people to stake to secure the chain, but at the same time at some point the additional security benefit of more validators does not justify a high inflation any more. For example a network that can be attacked for 500,000$ may offer 25% interest per year to incentivize more stakers to join and secure the network, while a network that takes 10 billion dollars to attack does not need to incentivize at such extreme lengths and can reduce the interest to 4% per year and cut inflation for everyone else in return.

So like in PoW, you have protocol level competition with other validators. And in order to stay competitive, you would have to reinvest your earnings and spin up more validators to keep your current influence and reward level.

While this all sounds pretty straight forward so far, it will take more to make PoS work though. The very first generation of PoS coins that came into existence around 2014 suffered from a flaw called "nothing at stake" (including the coin that pioneered PoS "Peercoin"). You see, when you decide to attack the network in PoW and build your own chain to replace the legitimate one, you need at least 51% of the networks computing power to archive this. To reverse a transaction, you'd have to do the work of the last 6 blocks again and build your new chain faster than the one that is being built by the entire rest of the network combined (remember, in PoW we follow the chain with the most work put into them).

But in PoS, since no work is attached to the creation of a block, a validator can simply replace their own last 6 blocks with 6 different blocks in less than a second. This would lower the barrier for 51% attacks extremely since even with 10% you might be the block producer for the last 6 blocks just by pure chance of luck. And if that happens, it costs you nothing to replace the last 6 blocks and attack the chain, which is why the problem is called the "nothing at stake" problem.

Modern PoS systems like Ethereum solve this by introducing two simple rules:

1) You cannot sign two conflicting attestations or blocks at the same height

2) You can not signing a block with a source or target that would surround you or get surrounded by another block you made

Since every duty you perform is signed off by you with a mathematical signature that cannot be forged, it's super easy to prove whether you violated this rule or not. You just have to listen to two duties signed off by the same entity for the same slot. That's all it takes for any validator on the chain to include this as proof on the beaconchain that you violated these rules and slash you. Slashing is basically a severe penalty where you end up losing parts or even the full 32 ETH of your stake plus you will be forcefully ejected from the network and can no longer operate as validator. Now if you try to replace the last 6 blocks you created by signing off 6 new blocks, you'll be slashed and lose your ETH and your validator role.

So in short: Money translates to stake. Translates to rewards and influence. Consumes 99,8% less energy. Bad actors get slashed, lazy actors get penalized.

Now that we understand the basics, let's dive into the most common misconceptions:

FAQ

PoS will make the rich even richer

If you read my comparison of the two systems, you are probably noticing that those two systems are not really that different from one another in this regard. In both systems, money means influence. A lot of real world money can get you a lot of mining hardware while a lot of real world money can get you a lot of ETH to stake. On PoW you buy hardware to strengthen your influence, on PoS you provide this money directly to strengthen your influence. On both, you have to reinvest your earnings to keep your position in the network or otherwise see it shrink by the rivals. Miners due to the difficulty increase, stakes due to the reward decrease and lower chance to propose a block. Both systems give incentives to participate in the consensus in the form of money.

While miners in PoW make way more money than their PoS counterparts, they also have way more running operational costs. But you'll earn less while staking than you'd be mining.

Example: 32 ETH right now (1 ETH = 1600USD) would get you about 5,1GH/s (51x 3080) of mining equipment. Using 17500W of power and at an electricity cost of 0.2USD/kwh and a pool fee of 1% would make you 1,100USD per month in net revenue. While staking those 32 ETH right now would earn you 0.1 ETH per month or 160USD per month.

Or when comparing annual ETH inflation of those two systems: PoS creates 607,500 ETH a year while PoW creates 4,414,310 ETH per year.

While the argument holds true that people with money can make more money in PoS, the same holds true for PoW. Or any other capitalist system for that matter. As we've seen with this example, PoW is more attractive to make more money and it does so even faster than PoS. PoW makes the rich richer is actually more true than the other way around.

But PoS requires 32 ETH at stake. PoS is for the rich and therefore not decentral

Another pretty common misconception is that you need 32 ETH to stake. While this is true if you want to solo stake, you can also join a staking pool with just 0.01 ETH too. This is very similar to PoW once again since solo mining is not something the average miner can do to turn a profit, so they instead form pools and share their income. This is also true for PoS.

PoS has another advantage up its alley since there is yet again another thing that is not possible on PoW, namely decentral staking pools like Rocketpool. Where ETH is provided to smart contract that spins up a validator once 32 ETH has accumulated. The duties are performed by anyone that wants to be a Rocketpool node operator.

There are also centralized staking pool solutions like they are currently with PoW. The major downside PoS has here compared to PoW is that ETH is currently locked in at these staking pools. If you decide to switch pools, you may realize that this is either not possible at your centralized staking pool or if it's a centralized staking pool with liquidity to sell your stake, you might do so at some loss (for example Lido offers a liquid staking token called stETH that currently trades 7% below ETH).

I'd encourage people that don't have 32 ETH and want to stake to look into those decentral staking solutions like Rocketpool, since you are actually strengthening the network by making it even more decentral and can profit from it even without having 32 ETH.

PoS is way less secure than PoW

That's an argument I've seen in nearly every discussion about PoS. Currently there are more than 400,000 validators active on the beacon chain with a total value locked of 13 million ether. To attack the chain you'd need more than 6.5 million ether or more than 10 billion USD at current mark prices. Attacking Ethereums PoW chain for 1h costs about 800,000 USD according to https://www.crypto51.app (to include bitcoin in this as well since it is the gold standard of PoW, 1h would cost 650,000$).

Now this comparison is not really fair, though, since many would point out that the real price is way above that since acquiring 50% of the network computing power is not an easy task. But the same holds true for PoS, acquiring 50% of the current staking ETH supply is not easy either.

Plus PoS has limits in place on how many new validators can join or exit the validator set, so even if you were to acquire 6,5 million ETH, the time it would take to convert them to validators would take around 150 days (6 new validators per block, every 12s a block).

We learned that there are slashing mechanisms for bad actors and penalties for offline validators. Another thing that improves security significantly on PoS is the concept of finality. Once 2/3 validators agree on the canonical chain, this chain cannot be reverted without resulting in massive slashes. For users and exchanges this means that once a transaction has been finalized (happens after 2 epochs or roughly 13 minutes), you can be absolutely certain that this transaction can't be reverted any more. This is a major security benefit for exchanges and merchants and is something that is not possible with PoW.

Contrary to the questions we tried to answer here, PoS offers more security than their PoW counterparts.

Ok so PoS is centralized since there are these centralized staking pools like Lido with currently 30% of all staked ETH

Lido is somewhat of a special case. While it's true that their market share is currently at 30%, not all of those validators are operated by Lido. Lido is somewhat de-central, meaning ETH is locked into smart contracts and pegs their liquid staking token called stETH. Lido has limited control over these funds and controls only a subset of these 30% of validators themselves. The node operation is outsourced to 25 different institutional node operators all across the globe, for example a few Ethereum PoS developer teams are node operators for Lido. These node operators earn 5% of those rewards and Lido has no more influence over the validators once they are operated by the node operators.

Remember that a validator cannot contradict itself and sign two blocks or attestations with the same height, so should Lido decide to intervene and start operating the validators themselves again after they have been passed on to node operators, those validators would immediately be slashed for contradicting themselves.

While I still discourage to use Lido for their market share, the argument against them is put into perspective when you realize that Lido is basically a pool of 25 different smaller pools. As long as they are committed to onboard more node operators, this won't be an issue. Compared to the current system under PoW where Ethermine controls 30% of all the hashrates and is the single node operator, a staking pool with 30% share and having 25 different node operators is definitely an improvement.

PoS is still centralised, look at all the nodes being run by Amazon’s AWS

That's true, but you may ask yourself where all those Ethereum and Bitcoin mining pools operate. They too rely on clouds like Amazon or Google.

But with PoW, Amazon and Google can't do anything since miners secure the chain

You may think that miners secure the chain, but in reality the pools are the ones that create the block. Miners solve the mathematical problem for the pool by creating this blog without knowing what's in the block or without the ability to use the solution for their own block. Simply put, the mining pool hands the miner a specific problem to solve that only the mining pool can use to create the block. So if Amazon or Google turn evil, they can attack PoW as easily as they can with PoS by just switching out the problem the miners need to solve to create an evil Amazon or Google block.

But miners can switch pool if they get wind of this while in PoS you may not

First, you have way more people solo staking from home than miners solo mining from home. Secondly, PoS entertains the thought of social consensus, basically that the people that use Ethereum have a say in what they think is canon as compared to letting algorithms dictate what the canonical chain is. In such an extreme event where maybe even more than 50% of validators have been turned bad, the community can ultimately come together and embrace a minority chain and fork those bad actors out.

Just to illustrate the point though, this is a very unlikely event and would have huge impacts on the entire crypto scene regardless if it happens to a PoW chain or a PoS chain.

PoS allows censoring of transactions like we saw with tornado cash

This has nothing to do with consensus, censoring is possible in both PoW and PoS. I know some Bitcoiners like to point that out as a flaw in Ethereum but fail to realize that this is a problem facing cryptocurrencies as a whole and could easily happen to them too. Instead of fighting each other, crypto should stand united against on chain censoring and put tribalism to side for this one.

PoS is bad, remeber what happend to Luna?

This has again nothing to do with the consensus algorithm. What happened to Luna would have happened regardless of whether they would have run on PoW or PoS. The flaw was with their stablecoin system and not with the consensus system.

Vitalik is the dictator of Ethereum, this makes Ethereum a centralized project

If Vitalik decides today to push a hard fork that would give him 10k ETH, no one in their right mind would run this update. Node operators have always the choice to reject an update by not running it. If Ethereum were centralised, there would not be an Ethereum Classic today and there won't be an EthereumPoW after the merge. People found reasons to reject the update proposed and came together to continue their version and let users ultimately decide which chain they want to be part of.

Also, contrary to for example Bitcoin, where Bitcoin Core is the only de-facto implementation of the Bitcoin protocol, Ethereum is committed to what's called client diversity. Overall there are 4 different execution clients for Ethereum and 5 consensus clients. Each client has their own dedicated team behind it. If there's a bug or foreign takeover in one client, Ethereum will continue to run thanks to these other clients. So Vitalik would have to convince four different development teams first to implement his 10k ETH fork and then convince the entire rest of the community, including all the nodes, stakers, exchanges and users to actually run it.

As with Bitcoin, anyone can participate in the development of Ethereum and propose their own changes to the protocol in the form of an EIP (Ethereum improvement proposal).

Ethereum trades security and decentralization for energy efficiency

If you made it through the entire post you already know that this is not true. PoS offers increased security, more decentralization, consumes less energy and cuts ETH inflation by a huge amount.

So no draw backs then? That's rich...

PoS is an incredible engineering feat that took over 6 years to get right. But this does not come without some drawbacks. For one, PoS is an incredibly complex system. Especially when being compared to PoW, where the simplicity of PoW is arguably the main beauty of it. Developers like to avoid complexity at all costs since it increases the possibility of errors which is why PoS took that long to develop and thoroughly test. Another drawback is that staking is a far more complicated thing to do instead of mining. There are several steps to run your own validator, there's a requirement for permanent uptime and a wrong staking configuration that could end up costing you parts or even all of your staked ETH.

The beautiful thing about PoS though is, that even though it's harder to stake, the protocol is actually designed to work together. You as individual staker earn more rewards if everyone is participating correctly then you would if someone set up their staking machine wrong. So you even have a selfish reason to help this staker to set up their device correct. Plus the people in staking are super nice and friendly, there are entire subreddits devoted to educate people about staking and helping them set up validators or offer advice on how to stake their sub 32 ETH.

The last major drawback to date though is the inability to withdraw your funds. To keep the tight merge schedule, it was decided that withdrawal will be implemented in the first update after the merge. Which is realistically still 6-12 months away.

The last one, you can not initially distribute a coin with PoS. This is actually something that PoW does incredible well, the distribution of coins at the beginning of a project.

TL/DR: PoS and PoW are actually more similar than you might have guessed. Many downsides PoS is criticized for exist in PoW as well. Some things PoS is criticized for is actually worse under PoW. If you want to learn more, read the entire thing.