One of the most-cited benefits of web3 ecosystems is governance. Countless blockchain-based projects have touted progressive, decentralized decision-making and ownership by the community.
In some ways, this really isn’t revolutionary at all — stockholders have had voting rights in companies for centuries. The blockchain game narrative is that ownership is by the players, but in most cases the governance tokens are relatively expensive, resulting in a plutocracy of the purest form.
Some games have tried to balance this out by giving away governance tokens as rewards for playing, but typically a large percentage is reserved for the team and investors in order to avoid substantially devaluing the tokens.
As a web3 designer or investor, it’s essential to consider the implications of governance policy. Fortunately, we do have case studies on games that have been testing and exploring voting on the blockchain. In this report, we’ll be looking into Laguna Games’ Crypto Unicorns, a high-profile web3 project that is attempting to hold true to the vision of decentralized decision-making.
Overview & Updates for Crypto Unicorns
A full explanation of Crypto Unicorns is out of scope for this essay, but you can read our detailed report on the subject titled Crypto Unicorns: A Unicorn in the Making? for more in-depth details. Here we’ll just do a quick overview for the sake of context, along with identifying some of what has changed since we published our deconstruction in July 2022.
Crypto Unicorns was envisioned as an entertainment IP in which unicorn NFTs could be used across a variety of games. The project was flush with cash, raising $5 million in a seed round, 1,400 ETH in NFT sales, and $26 million in an investor token sale. The studio grew to about 60 people and began working on its roadmap. That included three internally-developed games targeting different genres: a jousting game, a racing one, and a team battle RPG. While the team hoped to get external developers to create content to support the NFT project, they also wanted to control their own destiny and deliver on their promise of interoperable tokens.
Crypto Unicorns has a relatively standard complement of tokens, fungible and otherwise.
- Unicorns: Of course, there are the titular unicorns, which unusually don’t have rarities. They are, however, made from six parts (eyes, fur, horn, hooves, tail, and mane) and some of these parts can be “mythic,” with what looks like a roughly 0.2% chance. Unicorns have a lifecycle from egg to baby to adult, requiring various resources along the progression. Unicorns can breed a maximum of eight times, though the cost increases substantially each time, meaning it’ll be unusual to see unicorns going the distance.
- Land: Land has a hard cap of 1 million plots, though it appears that only about 35,000 plots have actually been minted. Land is not required to play the unicorn games, but production gameplay like farming, crafting, and breeding is reserved for landowners.
- RBW (Rainbow Token): This governance token is capped at 1B to be distributed over 5 years. Unlike some games though the governance token also has substantial in-game utility, being required for breeding, marketplace transactions, and premium entry fees.
- UNIM (Unicorn Milk): A pure utility token, it is designed to be earned and burned in a regular cycle, with nearly all usage immediately burning the UNIM. The primary source of UNIM is gathering on Land NFTs, though it can sometimes also be earned as rewards for gameplay or events.
Deflating from Inflation
For a while unicorn NFTs were quite strong, trading with floor prices above 1 ETH and top sales as high as 20 ETH (ETH was sitting just above 2,000 USDC at the time). The prices collapsed hard in May 2022, though surprisingly not primarily because of the more general market crash. Instead, the plummet was sequela to the start of unicorn breeding and the predictable NFT inflation that came with it. The prices haven’t recovered since, with current floor at around 0.025 ETH and the highest sale in the last three months able to pull only 0.3 ETH.
News and Updates Since Our Decon
Perhaps the most exciting and positive news for Crypto Unicorns (besides of course quadratic voting) is their content pipeline. While running nine months behind with the planned Q2 2022 jousting game just now releasing, the project now has some real momentum. The team’s much-anticipated team battle RPG had a closed prototype test recently and also surprised the community with announcements of two entirely distinct genres: match-3 PvP and autochess.
In addition to that, the team has announced three casual second-party games: Unigatchi, Unicorn Party, and Rainbow Rumble. The goal is to have 4 games (presumably jousting, team battle RPG, and two of the second-party games).
A few other highlights:
- In a somewhat surprising move, Laguna announced breeding incentives to make breeding more valuable. For example, a unicorn’s max energy increases the more they breed… we’ll let you draw your own conclusions about that.
- The team has also announced an update to its staking program that includes what feels like an over-complicated “badge bonus” system.
- They delivered a State of the Union 2023 blog post to summarize 2022 and set the stage for 2023. The overview was appreciated, but it also had some concerning assertions and borderline cultish phrasing.
- For example, Laguna says, “The only thing that will save us is growth,” which of course is a worrisome statement. They also tell the readers, “Remember that it is everyone’s responsibility to evangelize,” and to “go out and spread the word of the ‘corns.”
- They wrap up with an over-the-top vision in which in the year 2030 there are 500 million unicorn holders (which would be about 6% of the world’s population) and 100 million daily active users, which would triple the DAU of Candy Crush Saga.
- Is it possible? Honestly, most likely not. But even if it was feasible, it’s irresponsible to throw around numbers like that without any concrete backing.
Beginning with the Standard System
When we wrote our deconstruction of Crypto Unicorns nine months ago, Laguna was using the common voting model in which each token (in this case sRBW, which is staked RBW) gets one vote. In that analysis, we broke down some problems with the system:
- Out of 9,900 RBW-holders at the time, the top 11 alone could meet the 33% vote quorum.
- By contrast, the lower 9,800 wallets, literally 99% of the voters, could not reach a 33% quorum on their own.
- In an election for the 11 seats on the Council, the top wallet could single-handedly appoint five of the positions.
You can see the imbalance of power in the following chart. Twenty-eight wallets make up 50% of the total voting power and the top wallet had nearly 10% of the entire vote alone:
These issues are in no way unique to Crypto Unicorn. But they were clearly due in part to Laguna involving the community into its governance process right out of the gate.
A Proposed Fix
In our deconstruction we recommended that Laguna look into quadratic voting as a method of balancing out power while still allowing the heaviest investors to have extra weight. Simply put, in quadratic voting each wallet gets voting power equal to the square root of its tokens. A wallet with 81 tokens would thus get nine votes, so still a lot stronger than a wallet with one token but not to the point that smaller wallets have basically no incentive to even bother voting.
This scheme is not novel, and Ethereon founder Vitalik Buterin explains it in detail in his blog post Quadratic Payments: A Primer from 2019. A shorter and more accessible version is Shaan Ray’s What is Quadratic Voting?, in which he describes how the Colorado government used quadratic voting to prioritize funding. It’s also worth noting that Illuvium will be using quadratic voting.
An additional feature of the system is how it shifts behavior in elections with multiple options. The following is an obviously fabricated example that is simplified to explain the mechanic clearer, but in reality this is exactly the system that Crypto Unicorns is now using when voting for council positions.
Consider a hypothetical election in which a company is picking what movie to show at their next movie night event. Alex, Blake, and Casey are each given 25 “coins” to vote with, but Alex got an extra 11 coins for having the top sales, for a total of 36 coins. The movie options are Aliens, Back to the Future, Casablanca, and Dude, Where’s My Car. Alex is a huge James Cameron fan and casts all 36 coins for Aliens, resulting in six votes. Blake doesn’t like scary movies, instead putting nine coins into Casablanca (for three votes) and 16 into Back to the Future for four votes. Casey spreads the coins out more broadly and manages to cast a total of nine votes, even though a few were probably wasted on Sean William Scott’s cinematic masterpiece. Their votes are shown below, with the number of coins in parentheses.
In the end, the company will be enjoying Back to the Future at the next movie night. There are a few interesting elements to observe here. Alex, despite having the most coins, ended up with the fewest votes by focusing them all on one option. Alex did end up with the most influence over a single choice, but the preferences of the pair of Blake and Casey ultimately won out.
The reason is that it costs a lot more to put more votes into a single choice, but voters have the option to do so with the result of fewer total votes. The Economist’s Quadratic Voting: A Square Vote has an interactive demonstration of how vote splitting like this works.
While it attempts to strike a balance between true democracy and pure plutocracy, quadratic voting is not perfect. For example, it is vulnerable to a Sybil attack (a vote-manipulation scheme that relies on the ease with which voters can create multiple identities) and it is necessary to have some protection against scenarios in which the voting may be gamed. Quadratic voting is also not as intuitive as simple linear voting. This latter point becomes clear when looking at voting platform Snapshot’s interface for quadratic voting:
The question: if a voter had 100 tokens and allocated them with the above settings, how many votes would they cast for each of their two choices? There are a couple of reasonable interpretations.
- 25 of the 100 tokens would go to PistachioSwap, giving it five votes. The other 75 tokens would go to ProofofPistachio, giving it nearly 8.66 votes.
- In order to show triple the support for ProofofPistachio, a total of 3.16 votes would be cast for PistachioSwap (using 10 tokens) and 9.48 votes cast for ProofofPistachio (using 90 tokens).
Either of these interpretations is valid and, ultimately, illustrates the failing of Snapshot’s interface that there is ambiguity present at all. The chart should explicitly show how many votes will result. In this case, it turns out the former is what Snapshot uses: showing three times the support represents three times the tokens, not three times the votes. It would be interesting to know what percentage of the Crypto Unicorns community understands this to be the case while voting.
An Escape Hatch
On December 3rd, 2022, Laguna published a blog post titled Introduction to Quadratic Voting. The post outlined some challenges with the one coin-one vote system. It also discussed the election of its 11 council seats and identified two seats that would change under quadratic voting rules. Laguna felt the quadratic result was more representative of what the community wanted and so it proposed a move to quadratic voting for multiple-option votes. In the team’s words:
- “By shifting to a quadratic voting system for multiple choice votes, we expect to open the discussion to a wider audience by allowing voters to express how strongly they support the choices, have better representation on issues, and balance the voting powers.”
Laguna decided to do this only for multiple option votes as a hedge against Sybil attacks. The idea, per what appears to be a comment from a Laguna member in the Snapshot proposal, was to give a failsafe in case a Sybil attack was detected:
- “If a quadratic voting proposal is subject to a Sybil attack, anybody from the DAO may write a counter-proposal to nullify the said proposal. These types of proposals with binary options will retain the single-choice voting system, so with the rest of the DAO, we can still collectively nullify the said questionable proposal.”
It’ll turn out later that this escape route was important, though not for the reason Laguna may have expected. As an additional note, Snapshot.org is a platform for managing web3 DAO governance voting. It handles proposal submissions, a variety of voting systems, and validating the votes. In our experience, it has become the favorite governance solution among DAOs.
The Vote for Quadratic Voting
In the Crypto Unicorns deconstruction, we predicted that convincing those with all of the voting power to voluntarily give it up would be challenging, and that did indeed turn out to be the case. The quadratic voting Snapshot proposal opened for five days of voting on December 13th, 2022, and the final vote passed by a very narrow margin.
Of the top 10 largest wallets, only two were in favor of switching to quadratic voting. It was in fact the No. 1 and No. 3 largest overall wallets, and those two wallets alone made up 21 million of the 27 million votes in favor of the resolution. Without their support, the proposal would have failed spectacularly.
Even though only two of the top 10 wallets were in favor of the switch, 88% of the rest of the voters wanted the proposal to pass. Again, this isn’t surprising since those other wallets generally had a lot more voting power to gain from the change.
In the chart above, we have further split out 70 wallets as possible “Sybil wallets.” All 70 of them had the minimum 4 sRBW to vote with, compared to the smallest other wallet at 81, and the distribution of for and against was clearly not in line with the rest of the population. Additionally all of those wallets have only participated in three votes, unlike nearly every other voting wallet. Perhaps this was an attempt from some users to caution against the risks of a Sybil attack in quadratic voting.
The proposal did of course pass, but only thanks to two of the three largest wallets deciding to voluntarily give up a tremendous amount of voting power. It was fortunate for the rest of the community that they did, but it should hardly be assumed that this would often be the case in other votes, especially if significant financial incentives are tied to the result.
With its newly-approved quadratic voting capability, Laguna kicked off the vote for the Season 2 Governance Council the very next day, and that seems to have gone off smoothly.
A Controversial Reversal
Recall that quadratic voting was only enabled for multi-choice elections in Crypto Unicorns. The given reason was to protect against Sybil attacks, which is valid: there has to be some sort of anti-Sybil measure in place when quadratic voting is used.
However, the team soon discovered that the lack of quadratic voting in yes/no proposals is just as vulnerable to large wallets swinging the results. In February 2023, with the Staking Emissions Reduction proposal, a plan was made to substantially reduce overall staking rewards to try to curb inflation. The motion was rejected by a reasonable margin of 56.5% against and 43.5% for.
Two weeks later, Staking Emissions Reduction II was put up on Snapshot with the following explanation:
- “Only 95 wallets (13.5%) voted against the proposal, while 611 wallets (85.5%) voted in favor of the proposal. Given that the majority of DAO members support reducing staking emissions, we felt it was necessary to reopen the dialogue on reducing the staking emissions.”
For only the second time in the project’s history, the team put their 200 million votes into play, completely dominating the vote and forcing the proposal to pass.
Of course, it is entirely understandable that larger wallets voted against a reduction in token emissions considering the holders would have the most to gain from such a result. The team felt that the necessity of reducing emissions for the good of the ecosystem though justified stepping in and effectively overruling the vote. This type of capability is something we see web3 projects build in from the onset, especially while projects are in their earlier years and before transitioning over to full decentralization. This is great real-world example of the conditions that led to such a scenario actually occurring.
Going forward, Laguna has continued to champion progressive decentralization. In its State of the Union 2023 post, the team wrote, “Point blank — one day our players will own the Unicorn multiverse.” The post also proclaimed that the reality of that happening is “likely less than 2 years away.” Yet Laguna already found itself overruling a DAO vote. If the project actually scales to the size and financial power Laguna envisions, the politics and stakes of such votes will only grow larger and more complicated and these problems will be further magnified.
If Laguna decides to broaden quadratic voting to include all votes, the team needs to implement a validation strategy for proof of personhood. They’re also going to have to be willing to let go of the reins — something that was easy to promise at first, but will likely only grow more difficult and worrisome over time.
The first big test is about to happen, too. Remember how Laguna released both match-3 and autochess prototypes of the team battle RPG? Now, Laguna is going to let its community vote which one to actually develop and which one to kill.
There are a few important points we recommend designers and investors consider when evaluating a governance plan:
- Voting Method: There are many ways to conduct a vote (Snapshot already supports six types of voting) and there is a strong case to be made that the most common method (“first past the post”) is one of the worst, as Nicky Case demonstrates in their fantastic interactive exploration To Build a Better Ballot.
- Voting Weight: What balance should there be between plutocracy and democracy? It might make sense in a true shareholder situation that amount of stock held directly correlates to voting power. But if the goal is to build a more equitable community, then perhaps something like quadratic voting — or even a one person-one-vote system — would make more sense.
- Contingency Plans: What happens if the community votes in a way the team believes will be detrimental to the project? Building in a contingency plan can act as an emergency brake lever, but it also signals to the community that their actual governance power is somewhat of an illusion.
In addition to those questions, the most important question is simply: does governance even make sense for this project?
- It’s a fairly well-established truth that “design by committee” is almost never a good idea, and adding the ability to buy votes is not likely to improve the situation here. It’s common in early access or beta testing for developers to ask for feedback, but our recommendation is generally to trust your game designers instead of having binding votes determine the direction of your project.
- On top of that, it remains to be seen how much governance actually matters to players in practice. We’ll make the hopefully non-controversial assertion that investors are generally buying coins for returns on investment rather than the ability to vote on game direction. So if governance doesn’t add much value to your token and creates risk and headaches, it’s quite hard to justify including the functionality at all.
As an industry of investors and analysts, we tend to spend most of our time focused on land, inflation, economies, and value flow. And for good reason — if these are broken, a blockchain project will most likely fail. But governance has also become tightly coupled with the web3 dream, and improperly-managed decision-making can undo even the best-laid tokenomics.
In an effort to court gamers and give value to their tokens, projects tend to throw around promises of governance and ownership by the players. We see leadership using terms like “digital nation” and making appeals to democratic ideals, but care must be taken in how these systems are set up in practice.
Crypto Unicorns is a great case study for some of the risks and challenges associated with actually implementing governance measures in active web3 projects, and it’ll be fascinating to watch how it progresses and what we can learn in the process.