Proof of stake versus Proof of work

In a short sentence, proof of stake means that by keeping coins in your wallet you gain coins.

Proof of work means that by supporting a network and keeping track of the blockchain you get coins when you solve a difficult puzzle.

What is proof of work?

Now that you know the short versions we will go a little deeper into the matter.

In 1993, the early days of the Internet, there was quite a bit of spamming going on, just like today. The problem was that networks could be brought down by distributed denial-of-service attacks (DDoS). A DDoS attack means that you send so many requests to a server that it can no longer provide service and goes down.

Cynthia Dwork and Moni Naor had found a solution to this. They required someone requesting a service to do "some work" to get the service as well. This bit of work consisted of computer processing time. This simple method made it costly to attack networks.

The terminology "Proof of Work" was invented by Ari Juels and Markus Jakobsson in 1999.

In 2008, Bitcoin's white paper was written by Satoshi Nakamoto, describing the concept of proof of work and its consensus mechanism and how it is used by the Bitcoin network.

Proof of work was the first method in the world of cryptocurrency to acquire new coins. You got these coins by being the first to solve a mathematical sum in a network of connected computers. The actual work you did was validating transactions and adding new blocks on a blockchain.

A blockchain is a chain of blocks that expands as time goes on by 1 block at a time per period. You can visualize it by a chain of blocks in a row with a link between them.

In Bitcoin, a block is added to the blockchain every 10 minutes by validators. A block contains all transactions and new coins from the network. These validators are called miners. They try to be the first to solve a difficult mathematical puzzle and thus acquire the right to add a new block to the chain and become the owner of the newly coined coins.

It uses Hashcash, a proof of work system that was actually designed to limit spam and resist DDoS attacks. This system dates back to 1997 and was invented by Adam Back. This algorithm ensures that the difficulty of this puzzle keeps changing based on the computing power of the entire network.

At least 51% of the miners or validators must agree that the new block is correct. This is called a consensus algorithm.

The cryptographic protocol used in this process is zero knowledge proof. This means that a node, or validator, proves to another node that it knows the value of a statement without giving away any other information. Thus, a node must prove that it knows the state of the network without sharing or getting any further information.

Let's take a small trip to the beginning of the Bitcoin network. If you are the only miner, you get all the new coins every 10 minutes. When new miners arrive, the puzzle to solve is still simple, for example guess a number between 1 and 1 million within ten minutes. The first computer to guess it gets to make the new block and gets the new coins.

At one point, there are a thousand computers working on solving this puzzle. Then the solution is found very quickly and too many new blocks are created. The solution Satoshi Nakamoto, the creator of Bitcoin, proposed for this was to make the sum harder and harder, so that a new block is always added every 10 minutes. And so this is how it has worked for more than a decade, completely automatically.

As the price of Bitcoin rises, new miners keep joining and the sum becomes so difficult that many miners can no longer solve it first and give up. It is also possible that the costs are higher than the returns.

Mining with proof of work has become very difficult and expensive for Bitcoin, making only the very best computers and the lowest power rates profitable. These computers are only affordable by people with a lot of money.

Mining Bitcoin today already consumes as much electricity as all of Sweden consumes in a year. This is beginning to attract some criticism.

By the way, the reverse can also occur. If the puzzle becomes so difficult that it is solved less often than once every 10 minutes, it must become easier. After all, if the solution is not found, then no new block will be added. And if no new block is added, no transactions are completed.

Another reason for making it easier to find the solution to the puzzle is if there are fewer miners. This can occur if the price of Bitcoin drops dramatically or if the price of equipment and power makes mining unprofitable.

How does proof of work actually work?

Proof of work provides a secure and decentralized blockchain. It is so secure because so many miners keep up with the network. It is decentralized because anyone can keep track of the network if they want to. Anyone with a computer can choose to do so and is then a node. So there is no central authority, everything is automatic. You also don't need permission from an authority to join.

Once a block is added to the blockchain, it cannot be changed. This fixed state of a blockchain is recorded in a ledger and can be viewed by anyone. This is called a distributed ledger. All miners and nodes have access to the state of this network and ledger. This decentralized storage method is therefore a good security against attacks on the network or corruption.

Since everyone is allowed to participate, censorship is also impossible. Although governments can decide to ban the network or the coin itself.

When adding a new block, Byzantine Fault Tolerance is used. This means that the connected computers do not have to trust each other, but agree that the new block is correct because they are counting on the rules they are applying to be reliable. If 51% of the computers agree, then the new block is approved and added to the ledger.

Each block has a hash, which is a code, that is unique to the block. This hash is added to a new block and the hash of the previous block is also written into the new block, making the chain impossible to forge. This hash is a code calculated by an algorithm.

Proof of work takes so much computer power that it is usually too costly to attack the network. That is what the protocol was invented for.

Variants of proof of work

1. Proof of elapsed time.

   Miners are drawn by lottery to add a new block and thus get the new coins.

   This was invented by Intel Corporation in 2016 and developed jointly with IBM and Linux. Some features of this:

   • KYC (know your customer)

   • Random time that a miner or node does nothing

   • Permissioned consensus instead of permissionless

   • The winner of the lottery is the miner who had the shortest random time

     An example of a cryptocurrency that works with this is Hyperledger Sawtooth.

2. Delayed proof of work

   Smaller cryptocurrency networks have a low hashrate, or easier puzzles to solve due to lower computing power in the network. This makes them vulnerable to 51% attacks, where attackers try to get coins they are not entitled to by double spending or want to take down the entire network.

   In delayed proof of work, the hashrate of a strong network, for example Bitcoin, is used to protect the smaller network from attacks. Moreover, an attacker now needs 2/3rds of the hashing power to attack the network, instead of the usual 51%.

   A snapshot of the network is regularly taken as a backup against an attack. If an attack then follows, the original network can always be recovered from this backup.

   Example: Komodo, a fork of ZCash.

3. Proof of access

   This variant asks a miner if it can access older blocks in addition to the previous block. If he can cough this up then there is consensus and he is allowed to add a new block to the blockchain.

   An example of this is Arweave.

What is proof of stake?

Proof of stake (deposit) is a consensus algorithm used by blockchain networks to create new blocks and validate transactions.

Validators keep track of the state of the blockchain. A validator may add new coins to the network and validate transactions as long as he has staked his stake. In fact, he must usually have a fairly substantial amount of coins to be a validator.

These coins are at stake. If a validator does not perform his duties properly, he can be punished with a fine. This is called slashing. Some of his coins at stake are taken so that this validator knows not to do this again.

For example, if a validator is offline too often, other validators apparently have to do all the work. For that, a slash is handed out. If incorrect transactions are approved, you can also get a penalty. If you go wrong too often you can even be banned from the network.

Of course, if you are staking coins on an exchange, you are not a validator. That is the exchange. They have a lot of coins at stake and you get to hitch a ride with them for a usually soft price. There are small differences between how much validators give you for your stake, so some research can get you some extra coins.

The first cryptocurrency to work with proof of stake was the one and only Peercoin in 2012. This was followed by NXT in 2013. In later years, proof of stake networks came in until big brother Ethereum even started working with it.

Other well-known proof of stake networks include Tezos, Cardano and Polkadot.

Proof of stake could be called a distribution system for the rich. The more coins you have, the more coins you get. The less wealthy among us may join these wealthy in their shadows and hope that these rich validators give them a fair price.

From the other side, it is a fairer system than proof of work in many ways. You don't have to become a validator and yet you can still get extra coins even if you only have a handful of coins.

One factor that comes into play is how long you are a validator. The longer you do this, the more likely you are to be allowed to add a new block and get the transaction fee. Networks try to get reliable validators this way.

Proof of stake also works by randomly assigning validators to add a new block, otherwise the longest sitting would make all the blocks.

A nice aspect of proof of stake is that a 51% attack is useless. Since you get slashed for making mistakes, you would lose huge amounts of money that way.

What types of proof of stake are there?

1. Delegated proof of stake

   In this form of proof of stake, validators are delegated to add blocks to the network. Network participants vote on who the validators will be, like representative representation.

   Because there are few validators, the protocol is fast and scalable.

   A validator will do everything possible to keep his "constituents" happy so that he remains a delegate. After all, this earns him his income. He does this by keeping his promises, but also by offering commercially attractive rewards over the rest of the delegates.

   One drawback to this system is that it can involve nepotism. Some delegates are then chosen not on the basis of what they do for the network, but because of their extensive network. In extreme cases, such a clique of friends can even blow up the entire network, as they only serve their own interests or even attack the network together.

   An example of a coin with this system is EOS.

2. Effective proof of stake

   The purpose of this consensus mechanism is to prevent centralisation of stacked coins. They do this through a combination of early proof of stake (whoever has them the longest is often the most reliable), these strikers get the highest rewards, in addition to a lower reward for later investors that is still economically interesting.

   Example: Harmony.

3. Bonded proof of stake

   In this protocol, validators tie up their coins for a certain time (bonding), giving them a fee for validating new blocks.

   The more coins you have at stake, the more voting power you have.A disadvantage of this system is that validators are at great risk if the network or price collapses.

   Example: Cosmos.

4. Pure proof of stake

   The more coins you own, the more likely you are to be allowed to create a new block and get the coins.

   Stakers do get randomised and secretly chosen to validate blocks, but if you have more coins at stake you get chosen faster. It's like a lottery where you win more often if you have more lots. Actually, that makes a lot of sense.

   It is a very decentralised system that makes typical tricks like creating many accounts or favouritism of delegates impossible.

   Example: Algorand.

5. Proof of stake time

   The longer you have a coin at stake, you get more returns. This gives you an increasingly reliable network, as strikers leave their coins at stake more often.

   A disadvantage of this system is that coins become more and more in the hands of the few. Moreover, there is then virtually no trading in this cryptocurrency, so that does not bode well for the price.

   Example: Vericoin.

6. Nominated proof of stake

   Under this consensus protocol, validators will generate new blocks. Nominators nominate validators they support. The more support validators receive in the form of height of stake, the more likely they are to be allowed to create a new block.

   Both validators and nominators can be scrapped for implementation errors.

   Example: Polkadot.

7. Anonymous proof of stake

   Here you can anonymously stake and receive rewards without it being known in the outside world.

   Example: PIVX.

8. Leased proof of stake

   A validator needs a stake of a good number of coins from a network to be able to call itself one. Nodes can then lease coins to create blocks and share the proceeds with the validator you have leased from.

   Each validator sets its own conditions, so if you are going to lease with whom, you have to be careful. The leaser can always switch to another validator, so validators should also make sure they give good terms.

   Example: Waves.

9. Hybrid proof of stake

   Tries to get the best out of the combination proof of work and proof of stake.

   Example: DASH.

10. Secure proof of stake

    In this protocol, validators are chosen randomly in a way that cannot be influenced.

    It works with the number of coins at stake, but also with a rating of the striker himself. As soon as a validator makes a mistake, your rating will drop and you will not be chosen so easily.

    Example: Elrond.

11. Delegated proof of contribution

    Some networks reward people who do something for the network. People with coins may vote for people who have helped the network, with both voters and nominees receiving compensation.

    Example: ICON.

12. Roll delegated proof of stake

    Roll here stands for rolling a die, in other words an RNG (random number generator).

    This variant of delegated proof of stake tries to avoid the shortcomings of that protocol by having delegates validate a new block on a random basis.

    Example: IoTeX.

13. Tresholded proof of stake

    Your coins are locked at stake for a certain time, after which the rewards are handed out and your coins are unlocked again. There is a certain threshold (treshold) below which you don't get rewards, depending on the number of participants and the coins that are at stake.

    Example: NEAR.

Differences between proof of work and proof of stake

You can already learn a lot from the above text, but we will review the differences once again so that they are clear.

Both networks use a consensus mechanism to coin new coins, but proof of work consumes a lot of power in the process.

Proof of stake is a system where the rich get richer, but at least you can still join if you have very little, whereas with proof of work, the miner alone is too expensive for most of us.

With proof of stake, the security of the network is guaranteed by the fact that a rogue or bad validator gets slashed and thus loses coins at stake. Because of this, validators do ensure that theirs does not.

With proof of work, the security of the network is guaranteed by the fact that the cost of an attack is so high that it becomes unprofitable over time.

With proof of work, you need quite a bit of technical knowledge to participate. With proof of stake, you mainly need money.

Proof of work networks tend to be slower than proof of stake, making it more common for traffic congestion to occur and people to sit and wait for their transaction to be validated.

With proof of stake, the problem can arise that validators get so many coins together that they can start to influence the price of the cryptocurrency.

With proof of work, there is a danger that some networks will accumulate so much computing power that they can mount a 51% attack. With Bitcoin, for example, there is a pool that owns more than 20% of the computing power of the entire network. When pools get too close to 50%, they are often asked to force a split to avoid compromising the network.

Transaction costs for proof of work are usually much higher than for proof of stake.

Proof of work coins can be sold at any time. If your coins are at stake you cannot. This can be a problem in a crash or bear market. Stop losses are not possible if your coins are at stake.

Smaller networks operating with proof of work can be attacked faster if the computing power required for an attack is cheaper than its cost.

With proof of stake, this is virtually impossible because a 51% attack requires a huge number of coins that are at stake and where the attacker can suffer large losses when his behavior becomes evident. On top of that, the price of his coin drops significantly if he does so. Thus, the proof of stake mechanism is based on an obvious economic concept.