Blockchain Governance: An Overview and Prediction of Optimal Strategies using Nash Equilibrium


Blockchain governance can be formally defined as a subset of IT governance. The researchers use the Nash equilibrium to predict the outcome of three governance scenarios, and from the results generate formulae for choosing the best strategy to avoid a hard fork.


  • Khan, Nida & Ahmad, Tabrez & Patel, Anass & State, Radu. (2020) Blockchain Governance: An Overview and Prediction of Optimal Strategies using Nash Equilibrium. arXiv:2003.09241.


Core Research Question

  • How do you define a formal framework for assessing blockchain governance?


  • Blockchain governance exists on a spectrum. On one end exist platforms that are controlled, funded, and developed opaquely in siloed environments by centralized teams, then released to users. On the other exist platforms that are governed entirely by Decentralized Autonomous Organizations (DAOs), where formalized Improvement Proposals are organized and posited by community members. Development and funding are all controlled by token holder votes.
  • A Hard Fork can occur in any blockchain platform, and can be accidental or purposeful, adopted by nearly all validators/miners or contested. A hard fork can occur for a number of reasons, including significant upgrades to the software underlying a blockchain or to mitigate the effects of catastrophic events, as in the DAO Hack. Hard forks ultimately split the community between two or more canonical chains, and represent a threat to the mechanical reliability of its protocol.


  • Currently no formal method of defining blockchain governance exists.
  • This paper attempts to give a formal structure to blockchain governance by consolidating its position within the hierarchy of institutional governance, specifically as a subset of IT governance.
  • They provide an overview of the interactions between disparate and unaligned agents in the context of a network upgrade and a possible hard fork.
  • The Nash equilibrium was used to predict the outcome of three different scenarios: no governance, off-chain governance, and on-chain governance.
  • A payoff matrix describing the outcomes of those different scenarios was calculated and used to create mathematical formulae for analyzing the best strategy for avoiding a hard fork.
  • The proposed formulae were validated using real data from the 2016 DAO hack.


  • The main components of a blockchain network are Validators, Users, Consensus Protocol and the Governance Mechanism.
  • The Nash equilibrium is described as “a set of strategies, one for each of the n players of a game, that has the property that each player’s choice is his best response to the choices of the n-1 other players”
  • Blockchain governance is unique in that agents can simply fork the original protocol as an opt-out mechanism.
  • Blockchain Governance is assessed via the Nash equilibrium and framed as a two player game, with k number of Voters (V) involved in voting on a proposal, which affects n number of entities in the Community (C).
  • A payoff matrix (Fig. 2) was constructed, with β representing proportion of V who voted Yes on a proposal. γ represents the proportion of C that moved to the upgraded chain.

  • The Gambit project’s software tools for game theory, version 15.1.1 were used to compute the Nash equilibria for varying values of β and γ. Nine simulations were computed, with varying relative proportions of β and γ.
  • No governance, off-chain governance, and on-chain governance are modeled, compared, and contrasted. Real world data from the Ethereum DAO hack was compared to the predictions.


  • The payoff in the original and upgraded chain in scenarios where there is no formal external governance mechanism can be modeled in the upgraded chain by
    SU = γnSC ≤ nSC and in the original chain by SO = (1 - γ)nSC ≤ nSC
  • In off-chain governance scenarios, the payoff for an individual Community member is greater to follow the majority decision, reducing the possibility of a hard-fork. In on-chain governance, the addition of a testnet phase further reduces the risk of a hard fork splitting the community.

  • The DAO Hack was viewed within the framework the Nash equilibrium, and the researchers analysis, which corresponds to Fig. 4 row 2. The reality of the greater payoff in supporting the upgraded chain (ETH vs ETC) was viewed to validate the predictions offered by equations 8, 9 and 10.

Key Takeaways

  • On-chain governance, as modeled here with a testnet phase of any new proposal, provides the ability to increase the number of community members who will adopt a given upgraded chain, and as such is considered a superior model of governance.

Implications & Follow-ups

  • This paper, via analysis of the constructed payoff matrices, provided mathematical formulae that can be used to predict the occurrence and adoption of a hard fork in the face of a new proposal to a blockchain.

It looks like governance has been creeping back to the forefront of people’s minds lately. I think that’s pretty positive for a lot of projects! This summary seems particularly important with that in mind, as it is proposing a method to be able to evaluate and rank approaches to governance. I would be interested in knowing if anyone has seen attempts to use this approach, but making predictions about what is about to happen as opposed to analyzing what has already happened?

There are a series of other questions we might as of this article and governance approaches in general, so maybe that’s something that a discussion post might take advantage of?


Agreed, it would be really interesting to see if anyone is using this type of method to model interactions vs backward-looking analysis. I also wonder if someone is looking at a base model that assumes a multi-person interaction, not just a 2 person dynamic.

A project like Govbase is a useful tool for being aware of governance-related projects. Having more open-source projects breaking down governance approaches and incorporating game-theoretic dynamics as part of that would be fascinating and, I imagine, of use to the industry.