Why Incentives Are Critical - Understanding The Evolution of Proof of Stake

Ever since first contender blockchains emerged decade ago, the consensus debate has raged on. The overall concept of proof-of-work (PoW) pre-dates Satoshi Nakamoto’s white paper by around 15 years. However, Satoshi was the trailblazer of using PoW in a decentralized peer-to-peer environment. He was the first to propose it as a means of economic incentive for preventing the double-spend problem. 

Although Satoshi unleashed a game-changer on the world, proof-of-work in blockchains nevertheless comes with its own set of problems. 

First off, PoW is extremely energy inefficient. As of mid-2019, estimates put the Bitcoin network energy consumption above that of the entire country of Switzerland. Furthermore, PoW is a limiting factor in the scalability of a blockchain network. 

Perhaps one of the biggest challenges, especially for decentralization purists, is that PoW has ended up encouraging the formation of mining pools. As mining pools grow in size, they can dominate the networks hashing power, thus increasing the potential for centralization. 

According to a Coindesk report, no single mining pool yet holds 51% of the Bitcoin network hashing power, but the five biggest control 70% of it between them. If just four of them decided to team up, they could - at least theoretically - attack the network. 

So far, the Bitcoin network has remained stable over its lifetime. Nevertheless, these challenges with PoW mean that the consensus debate continues.

The most critical challenge facing core developer teams is how to achieve consensus in a way that doesn’t leave the network vulnerable to collusion and external attacks. Dan Larimer proposed the delegated proof-of-stake (DPoS) consensus method, which is a kind of representative democracy where token holders vote for the nodes that will run the network. 

Larimer was the lead developer on Steem and EOS, which both use DPoS and achieve transaction speeds far beyond Bitcoin. Therefore, DPoS has proven to overcome the scalability challenge. It’s also more energy-efficient as it doesn’t require the computational power of PoW. 

However, fewer nodes also form the Achilles heel of DPoS. In the case of EOS, because there are only 21 block producers, the network is at risk if block producers decide to collude. This isn’t mere conjecture - it’s already been reported. 

Furthermore, between whales and exchanges, EOS tokens are unevenly distributed. The top ten EOS accounts hold 50% of all tokens, meaning that voting power is also highly centralized. 

Proof of stake (PoS) could prove to be a strong contender for overcoming the issues seen in PoW and DPoS. In the PoS method, nodes stake tokens in order to propose and vote on the inclusion of the next block. The weight of the vote depends on the value of staked tokens. 

Like PoW, PoS provides a compelling financial incentive for the network to operate effectively and reduce the chances of collusion or attack. However, unlike PoW, it doesn’t require complex computation, so it overcomes the energy efficiency challenge. 

Nevertheless, PoS has its own set of problems. There should be enough nodes to achieve true decentralization. Therefore, the costs of running a node should be sufficiently low to encourage participation, or (like EOS) the network will become dominated by whales and exchanges. Furthermore, there should be disincentives for stake-pooling in order to prevent collusion. 

Ethereum has long been in the process of implementing PoS. However, it hasn’t yet made the full move to a PoS model - in part because the core developer team has been working through these challenges. In the meantime, several other projects have proposed innovations on the PoS model, designed to ensure it’s sustainable and that incentives are aligned across the network. 

To overcome the joint challenges of PoW and PoS, a few projects have developed a kind of hybrid model. One of the most often-cited examples is Decred. In the Decred consensus model, token holders stake their tokens to buy a “ticket” which is essentially a time-limited right to vote within the network. 

When a Decred PoW miner proposes a block, five randomly selected tickets are chosen to vote on the inclusion of the block. The block must achieve at least three votes to be added to the blockchain. 

This hybrid model makes it significantly more difficult to attack the network, as PoS voters can opt to exclude a PoW miner if they think its acting against the network interest. Furthermore, PoS voters are financially incentivized to act in the interests of the network as their tokens are locked up for a fixed period. 

Another variant on PoS is Wanchain, which was originally a fork of Ethereum. Wanchain has recently launched its own innovation on the PoS protocol, called Galaxy. 

Galaxy achieves a pure PoS but with mechanisms to overcomes the challenges of stake pooling or centralization. WAN token holders can “purchase” a representation of an accounts stake in the system called Wanstake. They do this by staking their WAN into a smart contract for a defined period of time. 

The longer the WAN are locked, the larger the Wanstake in the end. This levels the playing field for those with a small number of WAN tokens, as by locking them up for longer, their tokens carry more weight. 

The process for selecting a validator node is highly technical, and described in detail in the Galaxy yellow paper. In summary, once selected, a validator node can only produce blocks and not perform any other network operations. This further decreases the risk of malicious acts among the biggest Wanstake holders. 

So far, the idea of PoS being the consensus method to rule them all has existed in principle only, and few projects have implemented it with real success. However, the Wanchain Galaxy protocol could offer the necessary features for PoS to achieve the required balance of scalability, resistance, decentralization, and sustainability. It will be intriguing to see how this plays out, and whether other projects will ultimately follow in Wanchain’s pure PoS footsteps.