A Proof of Stake overview

In this post we will outline the main Proof of Stake (PoS) instantiations and corresponding projects. Sometimes PoS is being used as an umbrella term for a variation of consensus algorithms. So, what is Proof of Stake? The idea of a stake based consensus algorithm is an alternative to the inefficient Proof of Work that has been around for a long time. Proof of Stake (PoS) was first introduced in an academic paper by Sunny King and Scott Nadal in 2012 and from the start was devised as an alternative to the wasteful Bitcoin protocol. Since then there have been a number of different variations of the consensus algorithm. Just like PoW, alternative consensus algorithms come with several problems which lead to adapted versions of the original protocol.

Overall, the list of PoS algorithms falls into two main categories: chain-based and Byzantine Fault Tolerant (BFT). In chain-based PoS validators are selected pseudo-randomly to create the next block. In BFT the act of suggesting the next block and creating the next block are seperated. So validators that suggest the next block are selected randomly. Then a multi-round voting mechanism determines which block gets finally added to the chain. Unlike chain-based Proof of Stake, BFT consensus does not depend on the length or size of the chain after it. Consensus on a block can come from within that same block. It is important to keep in mind that with PoS design, choices can vary greatly and that a black-and-white division of consensus algorithms is sometimes not possible because a project may be to a large extent PoS-based but have some elements of BFT (e.g. Casper or Neo).

The first generation of chain-based PoS tokens

As a response to the PoW protocol in 2012 Sunny King and Scott Nadal started a more ecologically sustainable cryptocurrency called Peercoin. Sharing much of the source code and technical implementation with Bitcoin, Peercoin differs fundamentally in the way new coins are generated. Peercoin uses a hybrid PoS/PoW system, meaning that some of the coins are created based on the holdings of individuals. Per year there will be a 1% increase in tokens for the peercoin holders. The problem with Peercoin is that it fails to give incentives to token holders to decisively opt for a single chain. Instead they can break security by voting for multiple conflicting blocks without incurring any cost to do so. This is the so-called “nothing at stake problem”. In Proof of Work mining on several blocks is prevented by the fact that a miner must split up and waste their resources to do so.

Second generation chain-based PoS: Ethereum Proof of Stake Casper FFG

Even though Ethereum currently runs on a PoW consensus algorithm, the long term strategy of the project is to transition from PoW to PoS under the name of Casper. The first version of Casper, the Friendly-Finality-Gadget (FFG) will take on the form of a hybrid PoW/ PoS system. Centrally, Casper tries to address the “nothing at stake” problem in the proof of stake consensus algorithm as described in the previous paragraph. The main problem with Peercoin is that there are only rewards for producing blocks, but no penalties. Not being penalized for signing all blocks, the token holder can simply sign every fork on the chain. This includes malicious forks which intend to reverse valid transactions. This way the person who stakes gets the reward no matter which fork wins. Casper works against this tendency of forgers to stake on all forks by penalizing actors that do so. If a validator signs on two conflicting block headers, their stake will be completely or partially slashed.

The initial version of Casper which will be a hybrid of PoW and PoS will be rolled out by the end of 2018.

Some PoS projects: Peercoin, Qtum, Ethereum

Delegated Proof of Stake

A key design difference is whether systems allow token holders to delegate their rights to other participants who can then stake tokens on behalf of smaller tokens holders. Chain-based delegated Proof of Stake (DPoS) was proposed as a solution to incentivize small token holders to continue to participate in the system and remove the burden of having to run and secure a highly available validating node. As we discussed in the blog post on centralization, due to higher profit margins for big token holders, centralization on the blockchain may occur. To enable pooling within the protocol, in DPoS users are encouraged to delegate their tokens to bigger stake holders. As a reward they get part of the staking reward from the protocol.

Some DPoS projects: Ark, Lisk, Waves, Tezos, BOScoin

BFT Proof of Stake

In order to understand BFT, it is important to know which problem it is solving: the Byzantines Generals Problem. Hereby BFT algorithms are able to achieve consensus even if 1/3 (<=33%) validators go offline or become adversarial. However, this implies that ⅔ of the network are honest. Because consensus does not depend on the length of the previous chain, but will be achieved in every single round of validating a block. Furthermore, because BFT consensus algorithms favor consistency of nodes over availability, they can manage with non-synchronous network models. Chain- based proof of stake algorithms on the contrary rely on the synchronicity of the network. There are sub-variations of BFT called modified BFT which is less researched than classical BFT. BFT algorithms can have additionally a delegated voting mechanisms built on top, which gives them a touch of DPoS.

Some BFT projects: Tendermint, Cardano (both with DPoS mechanisms)

Modified BFT: Ripple, Stellar

We hope that now the confusion about PoS fades away and is replaced by curiosity to find out more about PoS in all its variations. If you want to find out more about reward systems and forging, then read our other post on the profitability of coins to forge. Given the plurality and diversity of future Proof of Stake coins, a pool for staking is a convenient place to store and manage assets. With Pool of Stake, users can pool their different PoS coins to get maximum convenience and reward.