What is the difference between proof of work and proof of stake?
The blockchain is a distributed ledger technology in which each node on the network stores a copy of the transaction ledger, removing the need for a central authority to verify transactions. However, transactions still need to be verified by somebody. When Bitcoin was originally proposed in 2009, the chosen method was one called ‘proof of work’, which Bitcoin still uses today.
Ethereum however, has been testing a different system since January 2018. Known as ‘proof of stake’ it tackles some of the problems inherent in ‘proof of work’, but it is not without problems of its own. What’s the difference? Let’s take ‘proof of work’ first.
Proof of work
Transactions are verified in blocks, with each new one being added to a chain that stretches back to the first transaction – hence the name ‘blockchain’. For a block to be added under a ‘proof of work’ system it needs a cryptographic hash – which is like a digital fingerprint for the data in that block. Computers across the network compete to be the first to find the hash that meets the right criteria – a process known as ‘mining’.
The first computer to get an acceptable answer broadcasts it – and the proof that it is correct – to the whole network. This is known as ‘proof of work’ – the miner is effectively submitting evidence that they have done the required calculations to verify a block. If they are correct then they are rewarded with a transaction fee and some of the currency and all the miners move on to attempting to verify the next block.
There are several downsides to this system. First, the verification process gets more complex over time, which means that mining requires more computing power and therefore consumes more energy.
Second, as the cost of mining increases fewer people can do it, increasing the risk of centralisation that blockchain technologies are designed to avoid. If the major Bitcoin mining pools chose too, they could work together to control more than 51 per cent of mining, effectively giving them power to rewrite the blockchain.
Finally, there is a lot of wasted effort in a ‘proof of work’ system. The energy consumed by mining is not just the energy of the miner that successfully confirms a block – it is also the energy consumed by all the unsuccessful miners who must start all over again.
Proof of stake
‘Proof of stake’ adapts the process to attempt to fix these problems. Rather than mining, ‘proof of stake’ validation is known as ‘minting’ or ‘forging’. They are still rewarded with a transaction fee but there is a reverse incentive at play too. Forgers ‘stake’ their own coins on their validations. Validating a fraudulent transaction means they lose their stake.
A crucial difference is that validation does not create new coins like mining does. This means cryptocurrencies that use ‘proof of stake’ need to distribute coins differently, such as with an ICO (Initial Coin Offering) or some other way of allocating all the coins from the launch of the currency. It is also possible that a currency could start out with ‘proof of work’ validation and then switch to ‘proof of stake’ later.
Another difference is that, without a race to solve a mathematical problem, the forger who validates a block must be selected. Forgers can be included in a validator pool based on the amount of coins they hold. A forger will be chosen at random based on the wealth they hold or the age of their coins. Depending on the implementation, the forger might be asked to propose a new block, which other validators vote on, or they might be invited to create one, which must point to a previous block.
A large minimum stake – effectively the amount of coins you are willing to bet on validating a block – will be required to be a forger. Then, their participation will be limited in some way so that they can validate only a certain percentage of transactions.
This solves the energy problem with ‘proof of work’, since it requires much less processing power for a forger to validate a block. There is also less – or no – need to issue new coins because the cost of validating is cheaper and therefore forgers require smaller incentives to help the network. Miners, in contrast, require larger incentives because they are incurring more costs.
However, sceptics argue that it might be a cheaper system to attack. Theoretically, anyone who owns 51 per cent of the coins would be able control the entire network. Against this is the fact that buying-up 51 per cent of the coins would drive value downwards, perhaps acting as a disincentive to try. On top of that are the economic punishments that ‘proof of stake’ systems usually employ. The community can ensure that a bad actor will simply lose their coins.
There’s also a concern that, since it requires a large stake to become a forger, there is a risk that control of the system ends up in the hands of the richest. That is exacerbated by the fact that forgers get a reward, so ‘proof of stake’ could create a system in which the richest have the most power and use that to make themselves richer.
Both systems have advantages and disadvantages, so it is not clear that one is better than the other. However, because the energy demands of ‘proof of work’ will continue to increase, raising environmental concerns, there is considerable interest in ‘proof of stake’ simply because of its ability to solve this problem.
Given that cryptocurrency is a technological currency, it is possible that neither protocol is the final one. New methods of validation might be designed by existing or future cryptocurrencies that solve the problems of both ‘proof of work’ and ‘proof of stake’.