What are Proof of Work (PoW), Proof of Stake (PoS) and Pure Proof of Stake (PPoS)?
After forming an understanding of blockchain fundamentals, it is important to understand the differences in how each chain achieves consensus. Think of consensus like an agreement between the community of users to ‘approve’ the transactions proposed in the current block. There are a few different ‘validation systems’ the different blockchains employ that we will take you through in this article.
The PoW system was the first to be used by a running blockchain, originally deployed by Bitcoin. In PoW, nodes (computers that make up the blockchain network) can earn the right to validate a block while simultaneously recording transactions onto the ledger. Strong incentives are typically given to the node that is awarded the status of ‘winner’. The winner is granted the authority to validate the block. For reference, the incentive for this winner is currently a bounty of 6.25 bitcoins (at a current market value of $46.3k USD per bitcoin [Aug 17, 2021]). This authority is earned when a user is the first to solve a complex cryptographic problem and, in doing so, obtains the missing key (i.e. ‘hash function’) to finalise that block. This process is commonly known as mining.
You may have heard that the mining process has been heavily scrutinised for its detrimental effect on the environment. Essentially, it requires enormous levels of computing power to be able to run random inputs into the cryptographic problem and achieve the correct result. When you have multiple mining farms (i.e. networks of mining computers) all competing for the one bounty, there will inevitably be a huge amount of wasted energy. These farms are comprised of large facilities, with rows of very expensive, highly specialised computers that consume HUGE quantities of energy. For reference, Bitcoin produces around 70.35 Mt of CO2 annually, which is comparable to the emissions of developed nations like Greece.
As a result, only large entities with sufficient resources to operate these mining farms have any hope of participating in block validation (creating a centralisation of power). For example, only a handful of mining pools in large part currently control the Bitcoin blockchain. Their collective power due to a large controlling interest in the currency creates underlying concerns for everyday crypto enthusiasts. This possibility has been called a 51% attack. It is believed that in a 51% attack, the entity with the majority stake in the blockchain would be able to reverse historic transactions and block new transactions from going through. At the same time, an entity that controls 51% or more of the network would harm the value of their own assets by acting unethically. Nevertheless, the fact that such an instance is possible is cause for concern.
Another curious feature of PoW blockchains is that they regularly experience ‘forks’ in the chain. A fork can either be a ‘soft fork’ or a ‘hard fork’, both with implications for its users. A soft fork can occur when two miners simultaneously solve the cryptographic problem and create a split in the blockchain (with effectively two correct chains operating at the same time). One of the forks will eventually be pruned, even though both miners acted honestly. The fork that ultimately prevails will be determined by which one has more work added to it through further mining. This means that only the miner who worked on the prevailing branch will receive the reward.
Soft forks also occur when a dishonest actor has contributed to the blockchain without completing the necessary work. In such cases, the network will collectively reject this addition and the data will not be verified.
Whereas soft forks occur regularly and are quickly pruned, hard forks have more permanent implications. Hard forks are more radical changes to a blockchain and often result in a split in the network into two forks: one with the radical change implemented and one that continues as is. For example, there are a number of blockchains with similar names to Bitcoin such as Bitcoin XT, Bitcoin Unlimited, Bitcoin Cash and Bitcoin Gold that are actually hard forks of the original Bitcoin with some alterations.
Another example is the infamous decentralised autonomous organisation (DAO) hack of 2016 in which a node stole 3.6 million Ether from the DAO blockchain. A hard fork was made in which the funds were never stolen and the original fork where the funds had been stolen also continued.
Because of forking, even nodes that have correctly mined the next block must wait for their work to be fully verified and enjoy the associated coin rewards. This greatly reduces a PoW blockchain’s capacity to transact at speed because the network is constantly having to check and re-check submissions before it can confidently proceed. In Bitcoin, for example, it takes ten minutes for new blocks to be finalised (i.e. ten minutes for any transaction to be processed). This means Bitcoin can only process 4.6 transactions per second. This is very slow compared to the 1,700 transactions VISA completes in the same period. At such speeds, Bitcoin cannot sufficiently serve the needs of a large network of users. This, combined with the massive levels of carbon emissions via mining activity, means that PoW is unsustainable and ineffective in creating a truly decentralised blockchain.
That is a lot to take in, so let us review the main characteristics of PoW:
● A node earns the right to create a new block by solving a cryptographic problem (the work).
● The first node to solve this problem can create the next block.
● This node will broadcast the new block to the network.
● The network will add this block to its blockchain.
● As a reward for its work, the node is given newly generated coins (i.e. mining).
● If two nodes solve the problem at the same time, a soft fork is created.
● Soft forks can also be created if a node proposes a block without doing the required work.
● In both cases, soft forks will eventually be wiped out.
● Hard forks are permanent splits in the blockchain that create separate blockchains.
● Forking causes huge delays in the block validation process, meaning that transactions take ten minutes to be processed.
● The mining process is hugely expensive and requires massive quantities of energy to complete.
● Mining farms have caused mining power to be centralised within a few large corporations.
The PoS protocol was conceived to solve the inherent flaws of the PoW. Because it was initially proposed in 2011, there have been many variations of PoS the different blockchains have used.
Delegated PoS, one of the earliest iterations, involves a fixed number of entities being elected by the network to a committee of ‘validators’. This committee validates new blocks on a rotational basis, effectively replacing the mining process from PoW. Validators are elected by its users to this committee and allocated votes proportional to the number of coins they hold (commonly called their ‘stake’).
This overcomes many of the problems inherent in the PoW protocol, including a reduction of energy consumption, increased speed of block validation, and greater scalability. This system fails to solve the decentralisation issue because meaningful participation resides largely with the committee. Further questions are raised around the vulnerability of this system. Corruption may exist within the committee, or users who uncover the identities of committee members may apply pressure on them.
The EOS blockchain, for instance, comes under scrutiny for only having 21 ‘block producers’ who hold such power.
So let's recap the key features of delegated PoS:
● Mining is replaced by a committee of users who are responsible for validating new blocks.
● Committee members validate new blocks on a rotational basis.
● This means that the computational power used is minimal.
● This process allows transactions to be processed rapidly, making it scalable.
● Power is centralised within the elected committee, making the blockchain vulnerable.
Bonded PoS was conceived to overcome the centralisation issues with delegated PoS. The key difference is that bonded PoS allows the entire network to participate in block validation rather than a selected committee. Users set aside a number of their coins in a bond (their stake) and are allowed to participate in the process, proportional to their stake (i.e. the more coins a user stakes, the more votes they have). All eligible stakers are entered into a lottery to determine who gets to validate the upcoming block. If a user dishonestly validates a block, the coins they staked will be forfeited as a punishment for dishonest behaviour.
Although this method does provide a more decentralised solution, imperfections remain. Because power is effectively based on how many coins a user is willing to stake, wealthy users could pose a power imbalance. These powerful users, if incentivised, may elect not to uphold the interests of the wider community. Additionally, because users are required to lock up their coins in bonds, they are inhibited from having the freedom to use their coins elsewhere.
Let us go over the main features of Bonded PoS again:
● Like in delegated PoS, the power to validate new blocks is vested in different members of the network.
● Any user can participate in block validation by staking coins in a bond.
● The more coins they stake, the more likely they are to be selected in the random lottery to validate the next block.
● Users who validate blocks dishonestly will lose the coins they staked.
● This system vests more power in wealthy users with an abundance of coins to stake.
● The requirement to stake coins to participate inhibits the free use of coins.
Algorand founder Silvio Micali conceived PoS. It has been implemented on the Algorand blockchain. Micali is not just the recipient of the Turing Award (in computer science), he has also received the Gödel Prize (in theoretical computer science) and the RSA prize (in cryptography), and he is the Ford Professor at MIT. Building on preceding PoS systems, PPoS allows every user in the network to potentially participate in new block validation.
Unlike in bonded PoS, PPoS does not require users to stake their coins in bonds but instead employs a random lottery system whereby any coin a user holds is selected at random to validate the next block. The likelihood of a user being selected to validate the next block is proportional to the number of coins they possess, which still provides an accurate representation of their stake. Users are also given the option to only transact and not validate if they wish, providing further flexibility. Users do need to stake their tokens to be able to participate in Algorand governance, in exchange for rewards (more Algorand tokens), but no such stake is required for block propagation. The application of the bondless lottery ensures that PPoS provides a genuinely distributed and decentralised model of consensus.
Here is a look back at the key features of PPoS:
● Every user is automatically added to the lottery by holding coins.
● If selected, a user will validate the next block.
● No bond stake is required to participate.
● The more coins a user holds, the more likely they are to be selected in the lottery.
● Users can opt in or out of validating blocks.
● This means the power to validate blocks is genuinely distributed across the network.
PPoS’s Solution to the Blockchain Trilemma
The blockchain trilemma posits that a blockchain can only succeed in fulfilling two of a blockchain’s three key aims: security, scalability and decentralisation. As we have established, PoW compromises scalability through slow transaction speeds, delegated PoS compromises decentralisation by vesting too much power in elected committees and bonded PoS compromises security by restricting the freedoms of users who stake their coins to participate in block validation. PPoS, by contrast, provides security through a robust lottery-based validation model, scalability through rapid transaction speed with only modest computing requirements and decentralisation through a legitimately distributed network that offers equal opportunities to participate in validation.