Overview
We propose a novel token weighted voting mechanism for making governance decisions. We call this “combat voting” because it works like a battle. During a vote, users stake tokens on their preferred option, after some time, the vote closes and the combat algorithm runs. Tokens are slashed symmetrically from all options until only a single option is left with any tokens. This is analogous to armies annihilating each other in battle to control a location. The option with the most tokens wins and any unslashed tokens are returned proportionally to the winners. We believe combat voting has properties and incentives that improve on traditional voting when making consequential decisions.
Why Combat
Create Consensus
Traditional voting systems measure consensus, they do not create consensus. Combat voting, as far as we know, is the only system which creates consensus by punishing dissent at the token level. The creation of consensus is important to achieving “finality” of a decision. In traditional voting, one can debate and vote on the same contentious decision, over and over again. With combat voting, the ability to dissent in a vote is burned out with the first vote, all participants lose their tokens, in some cases making it impossible to dissent in repeat votes, which makes the vote results final.
Create Debate
Not only does combat voting create consensus after the vote, it incentivizes voters to coordinate with each other before committing to a vote. Rational actors, wanting to maintain token holdings, will understand that consensus in the vote is key to retaining voting power over time, therefore some kind of pre-vote consensus process should take place. This is unlike traditional voting where, while debate may exist and be spirited, voters are not really incentivized to be organized around creating consensus to avoid risk of token loss.
Remove Quorum
Token weighted voting systems require an arbitrary participation threshold, or “quorum”, in order to be considered valid. This can be problematic, as votes risk failing due to lack of participation. Combat voting has no such requirement: any non zero amount of tokens participating in the vote make the vote outcome valid. Unlike traditional voting where it’s free to participate, in combat voting participation is disincentivized, since there is risk of loss. Only participants who really care about the outcome more than the potential loss of tokens will participate. It is assumed non voters abstain, which means a minority of token holders can pass a vote.
Inform Voters
Traditional voting systems do not require voters to have an understanding of the issue being voted on. This is an age-old problem in voting. Voters need only form an opinion based on superficial information and their vote is just as significant as a subject matter expert. Token weighted voting improves this by potentially giving more tokens to higher quality voters, but this problem still remains. This issue goes away with combat voting because voters must risk their tokens to get the outcome they want. This natural risk/reward incentivizes understanding and education of the issue before participation, and if the issue is not relevant, abstention is the best option. Not only does education about the issue at hand become important, but so does the need to understand the general sentiment of other voters, creating more depth to voter participation.
Reduce Bribes
In typical voting systems, token holders with little to no care in the vote outcome can be “bribed” outside of the protocol to use their tokens to vote for a particular outcome. Many organizations embrace this, and use it as a feature, but the ability to bribe for votes is controversial. Bribing a combat vote requires convincing token holders to risk loss, which is still possible, but would be a more expensive and complicated proposition. Similar to bribing, the use of token loaning services cannot be used in a combat vote due to staking and the risk of insolvency.
Shift Power
One of the big problems with token weighted voting is that entrenched power structures stay entrenched, effectively calcifying the state of decision making. While not necessarily bad in itself, there is no mechanism to challenge this. Combat votes create that mechanism. In a combat vote, dissent creates a token burn, and this burn applies to all dissenting voters, not just the minority voter. An unpopular, but entrenched power structure can take real losses if they continually fail to convince smaller token holders of their position. Combine this with an inflationary token distribution model, and you have potential for a “populist uprising”, and dynamic shifts in power over time.
Redistribute Tokens
While tokens can be deleted in a combat vote, they don’t have to be. By default the system burns tokens to effectively reward all token holders who abstained from the vote. Ultimately though, it’s up to the organization to decide what the best use is for these tokens. Organizations can use the tokens in creative ways to achieve their mission. For example, it may make sense to use these tokens to reward the author of the vote, to send the tokens to a treasury, or to send them to the core team or founders of the organization.
Specification
-
Combat votes assumes the existence of tokens or some kind of point system, distributed among potential participants, that can be transferred to or controlled by the voting system.
-
Combat votes must have at least 2 options that can be voted on.
-
When voting, a voter must specify an option and an amount of tokens to stake on that option.
-
When voting, control of tokens specified by the voter must transfer to the voting system.
-
Combat votes must have an open and closed phase.
-
During the open phase, voters must be allowed to stake tokens on their preferred option.
-
During the open phase, voters may be allowed to un-stake tokens on their preferred option.
-
Token stakes during the open phase may be public or private.
-
Open phase must remain open for an appropriate amount of time to allow all potential token holders to participate
-
Once the time limit is reached, the open phase closes and vote enters the closed phase.
-
In the closed phase, voters must no longer be able to change votes.
-
Once closed, outcome must be calculated (see calculations):
-
Winning option - Which option won, or nothing if a tie or no participation happens.
-
Tokens burned - How many tokens were burned.
-
Tokens returned - Who receives tokens back.
-
Once closed and calculated, voters must have some way to receive remaining tokens back from the vote, if they are owed any.
-
Once closed and calculated, tokens earmarked for burning may be appropriated for uses other than destruction.
-
Once closed and calculated, if a winning option exists, it may be executed.
Equations
12.a Winning Option
12.b Tokens Burned
12.c Tokens Returned
Discussion
Technical Feasibility
Combat voting is not a fully featured governance model, but aims to improve voting for existing governance systems that use token voting. We propose replacing token weighted voting with combat voting. This attempts to be a drop in replacement to an on-chain voting system, though deeper analysis would need to be done on a case by case basis to determine technical feasibility. A system like this would probably not work with popular off chain voting systems like Snapshot, without additional on-chain infrastructure.
Token Design
We feel combat voting works best when tokens can be fairly and continuously distributed across many stakeholders. Continuous token inflation is preferred in order to offset potential deflation due to token burning. Tokens should be used primarily for governance and have little to no monetary value. Tokens with significant value may create psychological barriers to voter participation. Even if these conditions are not met, combat voting would still technically work as long as there is engagement with the system.
Voting Strategy
Abstain
It’s worth illustrating how strategy changes for the average voter when combat voting is introduced. Every vote has the potential to increase your relative influence if you abstain from voting, due to the deflationary effects of combat. The more contentious a vote is, the higher the incentive is to stay out of it, which is nearly opposite of traditional voting, where you want maximum participation in contentious votes. Voters must now carefully consider how much they need the outcome versus their token holdings.
Build Consensus
Voters would also be wise to build coalitions of support for important votes before combat plays out, since unanimity means zero token losses. Ideally this leads to consensus being reached before a vote happens. In this world, the off-chain debate and negotiation process becomes the primary battleground and the combat system becomes a mere formality.
Bluff
In the current specification for combat voting, it is possible to “bluff” voting conviction by staking more tokens than one would commit to, then unstaking them right before voting closes. Bluffing could lead people to believe that there is more support for a particular option than there really is, dissuading others from dissenting. It’s not clear if this behavior would actually be problematic, but variations of combat voting could be used to reduce or eliminate this strategy. It’s possible within the current spec to enable a private voting system (more technical), or disallow changing a vote once committed (less flexible), though perhaps better solutions could be found.
Other Variations
Softening Combat
Combat voting may seem particularly harsh because of the token burn, fortunately there is a way to dial that harshness down, by creating a variable that reduces the burn rate. At 100% burn (default), all tokens that conflict get burned. We can reduce the amount of tokens to burn proportionally by lowering the burn rate, which enables voters to receive some tokens back. At a burn rate of 0% combat voting becomes very similar to regular voting, where no tokens get burned, and everything gets returned to voters.
Reward Voters
There is a way to incentivize participation in a combat vote by offering a token reward for participation. The reward would be paid out proportionally to all voters based on their tokens staked, regardless of their position in the vote. Typically a reward like this should not be needed if the outcome of the vote has value, but in case it is needed, this is an option. The size of the reward must be carefully considered; if too large, it can nullify the advantages created by combat.
Conclusion
Combat voting is a novel approach to voting inspired by the real world. It’s rare to discover such a simple solution to long standing problems around how we make consequential decisions. The goal of this proposal is to illustrate how combat voting works and the theoretical advantages it could bring to governance models. While still quite new and experimental, we have worked on an offchain implementation of combat voting (and other governance mechanisms) which are free to use at https://makerealms.com.
Credits
From the team at
Authors
David Adams (daywiss)
david@adams.software
Bananachain
nanner@blabsresearch.com
SenselessCEO
ceosenseless@gmail.com
NeonDaemon
neondaemon@protonmail.com