Michiel Mulders

A Deep Dive Into DPoS Platforms’ Game Theory & Incentives

Delegated Proof of Stake (DPoS) networks are considered to be the next stage in the evolution of blockchain network’s consensus mechanisms. The ability to return to ancient greek democracy in a decentralized, cryptographically secure and global manner is being tested in the past two years, and will continue to be tested over the next few years at the very least.
In this research piece, I will take a deep dive into the leading DPoS platform’s network incentives and rewards, to try and assess which network has its incentives more aligned with long term collaboration and participation, rather than attacking the network for self benefits.
However, before we jump into it, let’s quickly explain what are PoS & DPoS and what was the research’s methodology.

What is Proof of Stake?

PoS has grown out of the growing concerns regarding the sustainability of Proof of Work (PoW). As it requires much energy to secure the network and verify its cryptography, some see it as a feature that hurts the environment and our future. In addition, the token economics of PoW-based networks are one-sided and aim to incentivize the miners alone, whereas PoS networks incentivize all network participants. It does so by rewarding stakers on a daily/weekly/monthly basis, for contributing to the network.
PoS has gone through the next step in its evolution when Delegated Proof of Stake (DPoS) was introduced. The DPoS consensus mechanism grew from the understanding that a democratic governance model is more suitable for a scalable network.

What is Delegated Proof of Stake?

A “schoolbook” DPoS consensus works with a set amount of delegates. A delegate is responsible for receiving blocks, validating them, and adding them to the chain in a round-robin model. The users of the ecosystem can stake their tokens which gives them voting power to elect delegates. When selecting the top X number of delegates, the aggregated vote weight is used for ordering them.
Delegates are incentivized to contribute to the community to maintain their position. Whenever a delegate acts malicious or stops contributing to the community, voters can change their vote to another more trustworthy delegate, or other delegators can even penalize him. Therefore, a delegate needs to make active contributions to the community in order to remain in the delegate list.

About Myself

After finishing my academic degree in software engineering, I was active as a software engineer within a consultancy company gaining expertise in several programming languages like Nodejs, Golang, and Kotlin.
Early on in 2017, I started my internship with a blockchain consultancy called TheLedger as I personally believe blockchains will be widely adopted as a technology that empowers people and redistributes value, and have chosen to take part in the industry mostly because of that.
I’ve worked with BigchainDB, Stellar, Hyperledger Fabric and Ethereum to build prototypes.  Also, TheLedger allowed me to learn about token economics and financial systems in general. 
From this curiosity, I’ve decided to write this research piece. In this piece, I will be looking into EOS, TRON, ICON, NEO and Cardano rewards systems. These platforms rely on staking as a crucial element of their consensus and governance, and I felt it was important to fully grasp their token economics and their incentives mechanism, to hopefully understand which will last for the long term.
It is also important for me to mention I have decided not to include Lisk DPoS in this research, as I’m employed there and it felt unobjective.

Research Methodology

I have tried to answer several questions I see as important using every public material that can be found on the platforms’ website, technical documents, project blog or other external research conducted by 3rd parties. These are the questions and data points I have looked into:
  • What is the platform’s consensus protocol and how is it built?
  • What is the token economics (supply, monetary inflation rate, secondary tokens)?
  • What are the network’s incentives based on? Who is benefiting from them?
In the end, my purpose is to score each project for the following criteria:
  1. Expected future decentralization
  2. Incentives
  3. Rewards
  4. Community contribution

DPoS Platforms and Their Incentives Mechanism

An Overview of EOS

What

EOS, launched on June 10th, 2018, is a blockchain platform for smart contracts and the development of decentralised applications (dApps).
EOS’s smart contract platform is powered by WebAssembly which supports multiple languages widely used in application development. EOS also provides a set of services and functions that closely resembles what an operating system would offer for dApps. All these features enable businesses to develop and deploy high-performance dApps.

Consensus

EOS implements Delegated Proof of Stake (DPoS) consensus.
It defines “delegators” as the democratically elected block validators of the blockchain. EOS opted for a group of 21 delegates who act as the master nodes in the network. Their "job" consists of validating transactions, signing them, and adding the transactions to the blockchain. Only delegators receive the right to add data to the blockchain.
These delegates are voted into “office” by the stakeholders of EOS tokens. The more EOS tokens a stakeholder holds, the greater their voting power.
The reason Daniel Larimer, the founder of EOS, chose to appoint just 21 delegates in EOS is because any more would be detrimental to stakeholder attention, thereby causing voters to make poor decisions.
“You require a ⅔ majority to have an honest system. Originally BitShares started with 100. There’s not enough oversight of who those 100 people are because there’s not enough bandwidth of voters’ attention to decide. Bringing it down to 21 reduces the cost of the system. The network has to pay each person that runs a full node.” — Daniel Larimer
In short, EOS uses token voting to achieve decentralization, and the more EOS tokens a stakeholder owns, the greater their voting power. However, this can be seen as a potential risk as whales can upvote specific individuals and create cartels or not share rewards with the ecosystem.
EOS Constitution
EOS also introduces a novelty governance system, with a Constitution being an integral part of it, making EOS a governed blockchain. The original constitution guidelines can be found on the EOS Github page.
A governance system is fundamentally challenging to implement inside a blockchain ecosystem because everyone is trying to figure out the rules. Some people want to replicate existing legal structures, others want to regulate all manner of behavior, while others want to retain Code-is-Law.
The single most compelling feature of Code-is-Law is the removal of any room for dispute. All contractual terms are expressed in the code, which will execute predictably for all parties in the absence of bugs or extraordinary situations.
EOS, however, was created with the recognition that bugs/extraordinary events will happen and that the community needs a process to establish the intent of the smart contracts in order to quickly resolve issues transparently and predictably, when they do occur.
To enable such interventions, EOS Constitution has been created. Why these kinds of edge cases can be handled with EOS? The EOS community combines the best aspects of crypto-contracts, human contracts, and human dispute resolution. This makes EOS the first smart Ricardian Contract blockchain.
Cartel Formation Issues
Specific research has been conducted to prove that cartels can be formed and take over the network. This is an excerpt from the research conducted by the Whiteblock team that looks into blacklisting other EOS holders to remain elected. The report was written by Zak Cole, Brent Xu, Dhruv Luthra, and Nate Blakely.
"This type of attack can be executed by block producers simply accessing the explorer to discern the EOS holders with the largest voting powers and colluding to blacklist the accounts that threaten the authority of their existence as block producers. This type of vulnerability in the system is important to distinguish as there is no algorithmic rule preventing block producers from behaving in this manner. There is no proper protocol that is setup to prevent block producers from colluding to maintain their role as block producers. This further proves the high level of centralization that exists in the EOS network and the tremendous power these block producers possess."
The report proved that this kind of attack is actually possible. And so, also proving that block producers get too much power inside the EOS ecosystem. An actual EOS cartel took this one step further and executed such attack. This triggered the Block Producers to vote for a change in the constitution laws to prevent such behavior or change something about the token economics which are likely misaligned. 
The Block Producers move to quickly change the constitution also presents another decentralization issue, and that is the ability to make big and impactful decisions in a rapid and unsafe manner - simply because they believed it was the best solution at hand. However, proper consensus requires more research and understanding whether it’s the right decision for the network over the long term.

Token Economics

According to the staking calculator for EOS, the current annual yield is 1.84%. The current circulating supply is 912,836,253 EOS (accessible funds) of which 439,388,026 EOS participate in consensus (stake ratio of 48%).
The inflation rate for the network is set to 5%. 21 Block Producer share 0.25% annual inflation. All Standby Producers share 0.75% Inflation.
EOS Commodities (Tokens)
The commodities used to build and interact with finished products on the EOS network are RAM, CPU, and NET (bandwidth). While RAM is utilized exclusively by developers building decentralised applications, CPU and NET are needed by ordinary users wishing to interact with applications on the blockchain.
Exploring CPU and NET
-> CPU is a time-denominated resource which measures the amount of time an EOS BP should dedicate to transactions from your account.
-> NET is a space-denominated resource measuring what share of a blocks’ network representation can be used to store your transactions as they transmit on the P2P layer.
CPU and NET are regenerable resources that can be acquired by staking EOS tokens. Staking is the process of locking up your tokens in order to allocate yourself CPU and NET. Your staked tokens are also the amount of voting power you have to elect the Block Producers of your choice. Depleted resources a renewed eventually, allowing you to further stake and transact.
Exploring RAM
In traditional computer science terminology, Random Access Memory (RAM), temporarily stores all the data relevant to current operations, serving as the computer’s ‘working’ memory.
Similarly, RAM on the EOS blockchain is responsible for holding data such as keys, balances and contract state for a short period. dApp builders use RAM’s quick read and write capabilities to store the state balances of users interacting with their applications.
Insufficient RAM will hamper dApp performance. Unlike CPU and NET which are regenerable, RAM is a scarce resource which can be bought.
RAM uses a market-driven allocation model that prices RAM according to demand and supply. A smart contract holds reserves of both EOS and RAM, using the ratio of available RAM to EOS tokens in the contract to calculate the price of RAM.
This market-driven allocation model guarantees that users will interchange their RAM for EOS tokens or the other way round. It also allows EOS to remove the financial aspect from the RAM token. In short, a RAM token is a scarce resource which entitles you the right to store data.
Token Economics of Block Producers
The following snippet also comes from the Whiteblock report, talking about the misaligned token economics for block producers.
"Block producers are making a substantial amount of EOS daily. This means that these block producers can simply create new accounts and/or use proxies and reallocate their EOS to those accounts to vote for themselves or other block producers with which they may choose to align forming something of a cartel of block producers. This way, these block producers can stay a block producer, even if it is as a different node/account, and their allies remain in control of the network."
It is clear that incentives are misaligned. Block producers are not incentivized to contribute to the network and there is also nothing that prevents them to buy a lot of EOS tokens to become a block producer.

Conclusion

Decentralization: 1.5/5
Cartel formation possible and many technical issues with the chain itself that undermine the true decentralization of the network in my opinion.
Rewards: 1/5
The staking yield for EOS is very low with its 1.84% and can be a possible risk for delegates to act maliciously. Especially for crypto, 1.84% does not cover a lot of the crypto fluctuations making staking in EOS a more dangerous investment for small gains. Besides that, the EOS inflation rate consists of 5% in total which seems very contradicting with the low staking reward. This means stakers lose 3.16% yearly if they participate in staking, a clear misalignment.
Incentives: 1/5
I don’t see any incentives in the EOS network besides the monetary aspect for EOS block producers. The Whiteblock research clearly shows the flaws of the EOS ecosystem as block producers can use their rewards to again vote for themselves and so strengthen their positions in the ecosystem.
Community contribution: 2.5/5
Besides the monetary aspect, there aren’t many factors that move people to actually contribute to the ecosystem. DPoS is a good incentivization mechanism for encouraging delegates to contribute, therefore EOS gets a 2.5 score.

An Overview of TRON

What

Tron was founded in September 2017 by a Singapore based non-profit organization called the Tron Foundation. It is headed by CEO Justin Sun and has a dedicated in-house development team. TRON is a blockchain ecosystem for creating decentralized applications. It’s been created as an alternative to Ethereum and it claims to be more scalable and cost-effective than other solutions.
Just like Ethereum, TRON is based on the programming language Solidity and its virtual machine is very much similar to Ethereum’s as well.

Consensus

TRON is based on DPoS and works with delegates named Super Representatives (supernodes), within the TRON ecosystem. Everyone can become a node by acquiring 9,999 TRX tokens.
Next, 27 supernodes are elected out of a pool of many candidates on a rotating basis every 6 hours. A usual node can go to the upper level by either acquiring 110 million TRX tokens more or by gaining the same amount of votes from the community.
Every candidate can post its proposal on Tronscan.org where you can find his details. For example, SkyPeople, offer an 80% reward return if you delegate your power to them. It is a common practice to list your contributions to the network so users can determine your validity and potentially vote for you.

Tron Vulnerabilities

It turned out that just one computer could have brought TRON’s entire blockchain to a screeching halt. This bug which has been fixed by now can be found in a Hackerone bug report.
Until quite recently, bad actors were reportedly able to maliciously consume the CPU power of the network with Distributed Denial-of-Service (DDoS) attacks.
“Using a single machine, an attacker could send a DDoS attack to all or 51 percent of the [Super Representative] nodes and render TRON network unusable, or make it unavailable,” reads the report, labeled high severity.
The bug was fixed privately before disclosing the report to the public. However, it is quite worrying that the network could have been taken down with just a single machine.

Token Economics

Tron offers a yearly staking yield of 4.13%. The available supply is 99,456,931,298 TRX. The current voted balance contains 7,494,931,010 TRX which is 12.11%. This is quite a low staking number for users that participate in the consensus.
Voting happens by freezing your tokens in the TRON ecosystem. As a reward, you will receive TRON Power. TRON Power is not a token but gives you the right to submit transactions to the TRON network for free, if you have acquired sufficient Tron Power. 
Supernodes can monetize their vote rewards for their own profits or distribute them among their supporters to give an incentive to vote for them again.

Conclusion

Decentralization: 2/5
If we want to review this, it becomes clear that users with a lot of financial capabilities can buy a high number of TRX tokens to become a supernode more often and earn more rewards. The risk is real that this wealthy node owner will not share rewards.
On the other side, TRON knew this "cartel formation" can happen and that's why they implemented the rotating supernodes system in combination with the voting system to balance out the system. Still, this is a clear threat to the decentralization aspect of the TRON ecosystem.
Rewards: 4/5
A yearly staking yield of 4.13% is quite a good return yearly. Also, TRON burns tokens in order to decrease the supply and so increase the token price. That’s why TRON scores well on the reward criteria.
Community contribution: 2.5/5
For contribution to the community, I have mixed feelings. By having a rotating supernodes system, more delegates can forge and possibly contribute to the community by funding projects or sharing rewards. However, there is still a strong focus on the monetary aspect of being a supernode, especially if you take the cartel formation into account. Typically, 80% or more is returned to the users that have voted for a delegate according to the TRON voting tool.
Incentives: 2/5
I couldn’t find many efforts to contribute to the ecosystem. Only Super Reps and Super Reps Candidates are incentivized to contribute to the ecosystem in order to become a Super Rep. In addition, a staking percentage of only 12% (out of the token supply) is very low and can harm the decentralization and the future of the network. 

An Overview of ICON

What

ICON is a decentralised blockchain network, purpose-built to facilitate interoperability between communities of independent chains. The ultimate aim of ICON is to serve as the framework for a web of chains comprising enterprises and organizations from industries like banking, health care, and education. Each can formulate their governance system while remaining connected.

Consensus

ICON hosts a new type of consensus algorithm which they call Delegated Proof of Contribution (DPoC). Again, this is a delegated system which is very closely related to regular DPoS. From the ICON yellow paper, we can read the following about DPoC:
"The central philosophy of the ICON Network reward distribution is fair compensation based on relative contribution. Each participant can demonstrate their contribution through the ICON Network’s unique contribution evaluation system. In the end, contribution is the most important value shared within the ICON Network, and therefore will be the sole standard in the network. Delegated Proof of Contribution (DPoC), as described herein, is the sole justification for electing representatives.”
In short, DPoC is designed to provide:
  1. A clear governance structure.
  2. Better throughput.
  3. More efficiency.
  4. Less centralization than some other DPoS and PoW mechanisms (cartel formation).

ICON Vulnerabilities

As reported in Korea Economy (Hangook Gyungjeh) on the 22nd of January, a potentially serious security vulnerability has been discovered in the ICON blockchain code by Korean R&D group Adevt. If exploited, the vulnerability may allow hackers to take control of the blockchain (and Loopchain), and even manipulate price fluctuations of ICON tokens. The vulnerability was quickly dispatched without damage being caused.

Token Economics

Issuance
ICX can be issued up to 20% of total volume annually with the consensus of ICON Republic, taking into consideration the ICX trading volume, DEX trading volume, and freezing volume.
However, ICX is not directly issued but by providing nodes with the right to issuance. These rights are granted to the nodes on ICON Republic, and each node is entitled to receive ICX from the ICON Foundation by exercising its rights.
This means that ICX could create new ICX tokens – up to 20% of the total volume per year – if needed and approved by the ICON Republic. These tokens would come on top of the 800 million total supply. No need to panic. It might take years before this might happen. Maybe it will never happen as it has to be approved by the ICON Republic.
At last, on the 24th of July, ICON has released version 2.0 of the IISS Yellow Paper. This change includes that rewards are first paid from transaction fees and only then new coins are minted.
As an example, assume that 10 ICX would be necessary to reward all 4 parties. If there were 3 ICX worth of transaction fees in the block, only 7 ICX would be minted. If there were 15 ICX worth of transaction fees in the block, then 5 ICX would be burnt and no ICX would be minted. This creates a deflationary environment in the situation of high network utilization.
Delegation and Voting
An ICON stakeholder will stake his tokens to a candidate that will manage the network on behalf of others. In ICON’s design, the top 22 delegates will be elected (also called Public Representatives - P-Reps). A P-Rep is tasked with validating and confirming transactions, maintaining the network and implementing upgrades. For this effort, P-Reps will receive a portion of the transaction fees or newly minted coins as block rewards.
The token model that ICON has implemented is unique as ICON stakeholders can receive rewards for staking their tokens but also receive rewards for contributing to the network.
A token holder can vote on existing DApps and 'Ecosystem Expansion Projects' (EEP). It does not matter how technical you are, anyone can vote on the projects they think are a good fit for the ICON ecosystem.
DApps and EEPs will receive rewards reflective of the delegation they receive. This is an essential difference vs. any other platform. Instead of the foundation providing grants to whomever they see fit (e.g. how it currently works at Ethereum), the community itself will be able to decide who receives the grants (in the form of block rewards) by delegating their respective ICX to projects or improvement proposals they feel best improve the network.
This way, ICON stakeholders remain voting right in the ecosystem and feel valuable (can contribute to the future of the network).
Rewards
Image source: ICON Foundation
Below are a few key observations of the ICX reward rate:
  • As % of network staked increases, the reward rate decreases.
  • As % of network staked decreases, the reward rate increases.
  • The reward rate per category ranges between 2% and 12% per year. Staking to all three categories results in a rate that ranges between 6% and 36%.
  • Once % of network staked crosses above the 70% mark, the reward rate becomes fixed at 2% per category.
Currently, when staking in all three categories (most simple use case - Representative, DApp, and EEP delegation) you will receive a return annually of 9% according to the calculator on the ICON website.
The reward rate used by ICON makes perfect sense. If for some reason users decide to stop staking their tokens, the reward rate will increase. This will again incentivize other users to join the ICON economy and stake their tokens in order to reap the rewards. This reward rate helps to maintain a stable amount of staked tokens in the network. This also means that representatives in the ICON ecosystem are selected more carefully.
Also, by allowing token holders to receive rewards for delegating power to DApps, EEPs, and delegates they are actually in charge of deciding how the ICON ecosystem should progress. By rewarding them for picking the best projects ideas and DApps, you create an ecosystem where the token holders hold most of the power.
ICON P-Rep Elections
The first elections for becoming a P-Rep are about to begin in September. In terms of entry barriers, anyone can apply for this role - which comes with benefits as do all DPoS/C platforms.
First of all, there is a monetary reward a P-Rep receives for producing blocks. However, according to ICON Foundation, a cost of more or less $2,500 is expected for running a node per month. With, additional costs like HR and maintenance engineers, the budget can run up to an extra $10,000 per month.
Once annual representative awards are collected by P-Reps, their production costs could be more than compensated: according to one of the ICON papers, that source of income could bring as much as 720,000 ICX per year (around $262,500 as of June), although the exact amount is not certain until representatives are elected and all variables are decided upon.

Conclusion

Decentralization: 4/5
The step for becoming a delegate (P-Rep) inside the ICON ecosystem is higher than other projects. It might involve running a high-cost node and potentially hiring HR and maintenance personnel while setting up a company on some occasions. However, this brings more trustworthy and reliable node operators to the ICON ecosystem which are properly rewarded for all these efforts.
Also, Delegated Proof of Contribution is a very good alternative for DPoS as it encourages people to actively take part in consensus and get rewarded for it.
At last, in my opinion, is decentralization of the network maintained by the flexible reward rate. The rate goes up as fewer people stake tokens, incentivizing others to stake their tokens as the reward is higher now. This mechanism helps to maintain an active number of users that contribute and watch the ICON ecosystem.
Rewards: 4/5
Although the high costs of being a P-Rep, node operators are properly rewarded for their efforts. On the other side, stakers are rewarded for voting for EEPs, DApps, and P-Reps. This means that voters are properly rewarded financially but also delegates are properly rewarded for all their efforts towards running their node and maintaining them.
Community contribution: 3.5/5
Both parties (P-Reps and ICONists) are rewarded for actively taking part in the ICON ecosystem. This means that both parties are satisfied and incentivized to contribute to the ecosystem to grow it.
Incentives: 4.5/5
Users are rewarded for actively taking part in consensus by voting for P-Reps but also for selecting improvement projects for the ecosystem and voting for existing DApps to be maintained/supported.
ICON chose to build an ecosystem that incentivizes both the delegates but also the token holders to act honestly and build a decentralized community. In combination with the reward rate that goes up when fewer people stake money, they accomplished in building a proper game economy for ICON. Whenever less money is staked, new stakers will pop up and contribute to the ICON ecosystem. At last, ICON token holders are rewarded for spending time reviewing Ecosystem Expansion Projects and the current DApps. This guarantees people will choose for proper projects that contribute even more value to the ecosystem.


An Overview of NEO

What

NEO is an open source project driven by the community and was founded in 2014 by Da HongFei and Erik Zhang. It utilizes blockchain technology and digital identities to digitize assets and automate the management of digital assets using smart contracts. Using a distributed network, it aims to create a "Smart Economy".

Consensus

As a "Smart" or "Digital Economy" relies on fast single block finality, they introduced Delegated Byzantine Fault Tolerance (dBFT). The protocol is very similar to the concepts of DPoS while implementing Byzantine Fault Tolerance.
Neo’s dBFT requires approximately 2/3 of the validator nodes’ signatures in order to come to a consensus. For dBFT, where F is the number of allowed failed nodes, the following equations give the number of required validator nodes and the number of required signatures (source: NEO dBFT 2.0).
  • F = number of allowable failed nodes
  • Validator Nodes = 3F+1
  • Signatures Required = 2F+1
As you can see, this is the standard behavior for a BFT implementation. As we are speaking about a delegated system, BFT only applies to the elected delegates in the NEO network.

NEO Vulnerabilities

NEO has 7 master nodes that are deployed to verify transactions. A bug has been detected that could be exploited by users which caused all master nodes to crash. Furthermore, it will cause a denial of service across the entire NEO network. 
“The NEO Smart Contract Platform provides a contract with a system call (System.Runtime.Serialize) to a certain object on the serialized virtual machine stack. This call processes the contract request without considering the nesting of the array, which will cause a crash of the smart contract system platform.”
More technical information a360security blog.
Another incident reported that Zhang and Hongfei (founders of NEO) used funds from the ICO to create independent entities housed under the foundation, including Neo Global Capital (NGC) and the NEO Eco Fund (which is managed by NEO Global Development (NGD)). These funds were invested in external startups like Chromaway, Ontology, StarkWare and Oasis Labs. This shows that the NEO founders are still making all decisions and invested the NGO funds in other startups for profit, instead of for the benefits of the network itself.

Token Economics

NEO has two native tokens, NEO and GAS. As NEO wants to support a smart economy and solve on-chain governance issues. They do this by using the NEO token for voting on major issues with the blockchain but it also gives you the right to claim dividends in the form of GAS. GAS also serves as the fuel token for operating smart contracts on the NEO network, while NEO’s utility stands as a voting mechanism alone, and not as a transactional crypto-currency.
When a block gets issued, 8 GAS tokens are created. Both NEO and GAS are capped at 100 million total supply, with the rate of production is reduced by 1 token for every 2 million blocks generated. Sometime around 2039, GAS circulation will reach 100 million and production will cease. Unlike NEO, GAS can be divided into decimals.
GAS token is an "interest-bearing" token. Interest is earned by holding onto NEO token. Currently, the annual yield for holding NEO is 3.02%.
With GAS as a reward, token holders are incentivized to build a trust filed network. Token holders and nodes incentives are aligned with NEO’s long term impacting decision making.
NEO’s 100 million tokens are divided into two portions. The first portion is 50 million tokens distributed proportionally to supporters of NEO during the crowdfunding. This portion has been distributed. The second portion is 50 million NEO managed by the NEO Council to support NEO’s long-term development, operation, maintenance, and ecosystem.
It is worth noting that due to the unlocking plan, NEO Foundation will have access to a considerable number of votes in the next few years. This can be a blocker for full decentralization of the network as they own a large amount of power in the ecosystem and can still make decisions themselves.
Considering that the NEO founders invest the money in other projects for profit shows how much they believe in their product. NEO holders get rewards in the form of GAS tokens. It doesn’t matter if they stake their tokens or hold it on an exchange. This means they are not incentivized to stake their tokens and actively participate in the network. This means they are not incentivized to vote but only for “future benefits” and nor for financial rewards.
Also, the cap for GAS tokens is 100 million. After that, NEO holders won’t receive any GAS tokens anymore.

Conclusion

Decentralization: 1/5
A very centralized system in my opinion as 7 master nodes are deployed by the NEO foundation themselves. Important to note that users can apply for becoming a consensus node. Currently, there are 8 other consensus nodes active. The status can be found on the cityofzion.io monitor. Next, the found DDoS vulnerability that crashes all nodes gives me even less confidence in the decentralization of the project. Also, the NEO foundation holds a large number of tokens which is in my opinion also a form of centralization.
Rewards: 2/5
The GAS token is an interest bearing token. You get GAS token by holding NEO tokens. However, the supply of GAS is capped at 100 million tokens. When the 100 million cap is reached, no more new GAS will be distributed.
Also, the transaction fee for using the blockchain is set to zero. It’s free. Consensus nodes have the power to introduce a transaction fee in the future. This transaction fee will then be distributed among consensus nodes only, and not to all NEO holders.
Community contribution: 2/5
The network is mostly focused on the monetary aspect and few contributions are made.
Incentives: 1.5/5
The incentives in the NEO network are completely misaligned. Users are not incentivized for staking NEO tokens in the network as they can also receive the rewards when just leaving their NEO tokens on an exchange (not all exchanges return the GAS tokens to the owners).

An Overview of Cardano

What

Cardano is a blockchain project founded by Charles Hoskinson, co-founder of Ethereum, to “provide a more balanced and sustainable ecosystem” for cryptocurrencies. According to its website, ADA is the only coin with a “scientific philosophy and research-driven approach.” In practical terms, this means that its open-source blockchain undergoes a rigorous peer-review process by scientists and programmers in academia. 
Specifically, Cardano aims to solve problems related to scalability, interoperability, and sustainability on cryptocurrency platforms. 

Consensus

The platform uses a form of DPoS algorithm called Ouroboros, which allows stakeholders to choose the nodes who will produce blocks and verify the network consensus.
These nodes perform a special form of task, they confirm the correctness of transaction blocks that are integrated into the blockchain. Every stakeholder has the possibility to operate a consensus node himself if he is selected. In addition, each stakeholder can delegate other stakeholders.
Both stakeholders, whether the voting stakeholder or the elected stakeholder, have an advantage. Stakeholders who delegate another stakeholder will receive a reward if their delegated nodes operate a consensus node. Each consensus node receives a higher reward. This not only provides incentives to operate a consensus node but also to vote for consensus nodes.
Cardano’s Ouroboros consensus protocol works with so-called time epochs. In each epoch, there are different time intervals, so-called slots, which represent potential blocks. An algorithm randomly selects a consensus node that can select a slot and fill it with transaction information. After a block has been created, the block is forwarded to other nodes for verification. In each period, all newly created tokens are collected in a pool and then distributed to individual nodes, depending on the participation.
The trick of the Delegated Proof-of-Stake algorithm is that Consensus Nodes can process multiple slots on many different blockchains at the same time. This solves the scaling problem. The platform uses the Sharding concept, which is currently being implemented at Ethereum.

Cardano Vulnerabilities

The following vulnerability is more of misbehavior by the foundation. The Cardano Foundation hired the FP Complete firm to audit Cardano’s codebase, which costs $600.000 per year. 
The issue with this report is that they never addressed the issues raised in this report. Besides that, it turned out that the firm had been reviewing a proof of concept that had been deployed by another company even though the code was not intended to be used long-term. 
The community was quite upset that the Cardano Foundation spent/wasted community money on this useless report.

Token Economics

Cardano sold ~26 billion ADA tokens (called 'vouchers') in an initial crowd sale, with an additional 20% of that total (~5.1 Billion tokens) created and issued to three key entities in the Cardano ecosystem. Additional tokens will be issued as block rewards over approximately 24 years, with a capped supply of 45 billion tokens.
When users delegate their stake, they pass their right to protocol participation on to a stake pool operator, and a delegation certificate is created on-chain to record that they have delegated their ADA. It’s important to note, that whilst users are delegating their right to protocol participation on to a pool operator, they are still granted full monetary rights over their ADA holdings — as ADA is a functional means of exchange in its capacity as a functioning cryptocurrency.
So, for instance, Amy may have 5,000 ADA held in her wallet which she wishes to stake with the ‘Hephaestus’ staking pool (one of the first pools created for delegating stake). While she delegates her stake, and therefore her protocol participation, to Hephaestus, she is still free to spend her ADA from her wallet address at any time.
The Cardano developers have achieved this through allocating multiple addresses associated with different cryptographic key pairs, including one to delegate stake, and another to spend ADA as a functional unit of currency.
For users delegating their stake to a pool, they receive rewards proportional to their stake whenever their pool is chosen to validate a new block, except they pay a fee to stake pool operators to perform this service on their behalf.

Conclusion

Decentralization: 4/5
I like the fact that users can run their verification node in order to double-check the validity of block producers. In my opinion, this adds up to the decentralization of the network as more users are watching it.
Rewards: 4/5
Both stakeholders, whether the voting stakeholder or the elected stakeholder, have an advantage. Stakeholders whom delegate another stakeholder will receive a reward if their delegated nodes operate a consensus node. Each consensus node receives a higher reward. This not only provides incentives to operate a consensus node but also to vote for consensus nodes.
Community contribution: -/5
I cannot find much information about this topic. As both parties are rewarded, I assume there are more contributions to the network as there is less focus on the monetary aspect because both stakers as validators are satisfied.
Incentives: 4/5
Like the ICON project, both stakers and block producers are rewarded within the Cardano ecosystem for actively participating in the network. I like this approach as both parties are satisfied and this stimulates people to also contribute to the network because all of them are part of it and thrive from a well-functioning ecosystem. 

The Bottom Line

As this research tries to determine which DPoS platform is more game theory-oriented in terms of incentives, we can find two projects that score well: ICON and Cardano.
In general, DPoS seems to be a very attractive consensus algorithm as it is cheap to run but also aligns the incentives between community members and delegate node operators. However, NEO’s implementation of DPoS is pretty poor as token holders are not incentivized to stake in the network or participate in voting.
Personally, I would pick ICON as the winner as it rewards (like Cardano) both stakers and delegates. However, ICON has an edge over Cardano by providing the reward rate that incentivizes stakers to stay part of the network. Even though stakers decide to “unstake” their tokens, the reward goes up and it benefits other users in the network. This helps the ICON network to maintain an active number of stakers that care about the network as they are rewarded for voting on delegates but also for voting on DApps and EEPs.
Additionally, unlike TRON and EOS, which employ somewhat similar models, the ICON network suggests benefits for both sides: P-Reps choose their own perks which then are reviewed by ICX holders during the election stage, while the voters are also rewarded.
Based on game theory-oriented incentives, ICON seems to be the most proper candidate for a large, self-sustaining decentralized network.
Disclaimer
This research consists of the many sources used for compiling this piece. It acts as a reference piece when trying to compare game economics and incentivizes between DPoS system. I'm not affiliated with any of the projects (independent report).

Addition Resources Used:

Tags

Comments

August 20th, 2019

This was seriously a very deep dive. More like a research paper I’d say.

Honestly, didn’t read all of it, but did so in parts. And it’s really detailed and put things into perspective.

August 20th, 2019

Really deep dive into the technology and a nice general overview of the blockchains. I was wondering why you left Steem and bitshares out which are also very traditional DPoS platforms. Especially Steem has interesting proposals in terms of rewards (Even though it doesn’t play out as well in practice).

More by Michiel Mulders

Topics of interest