To start, we should visit what the actual purpose of BTC is today. Whereas many once it was thought to be a cash replacement, the limitations on transactions per second (currently around ~10) meant it could never compete with Visa (around ~40,000). As of this writing, the narrative for Bitcoin has instead solidified around the idea of the “store of value.” Gold has historically held this role in society because it is both appropriately scarce and shiny. You use a store of value when you think that the market is over-valuing stocks and that a crash will be met with quantitative easing. Think of it this way: if the market is going to collapse, you should hold cash. Except! The government will print cash to help bolster a collapsing economy, which means the cash you are holding is also dropping in value. Bitcoin offers you an exit. Instead of buying stocks or holding cash, you hold Bitcoin or a commodity like gold. 

Most people immediately counter here with the fact that gold is “useful.” It has some inherent properties that make people want to incorporate it into their jewelry or computer circuits. The problem though is that these use cases are actually annoying. While you were trying to store your value, let’s say the global demand for computer chips and jewelry goes down, then your fundamental use case drags gold down with it. Bitcoin on the other hand, is fundamentally useless outside of a store of value. This in turn makes it easier to reason about.

But there is more. Bitcoin is 10x more useful than gold is as a store of value. For one, it is easy to send 100m$ of it around the world in 10 minutes. That task costs about 1$ in fees. Moving the same amount of gold is way harder obviously. It is also easier to seize gold. It is also easier to fake gold (in 2020, over a billion dollars of it was discovered to be gilded copper). Finally, it is easier to stop the sale of gold. Because Bitcoin operates on a decentralized and game-theoretically sound immutable ledger, its movements cannot be inhibited. 

At this point, the Etherean will stop and argue that Ether can do all of those 10x better things, but also more. It is a store of value that can operate in a decentralized financial market. And even if that gives our store of value a use case, which means it is once again harder to reason about its value (e.g. what if for instance the use cases go down is users?), the odds of the use cases going down is very low. After all, it is a Turing-complete money.

And the thing is, the Etherean is almost right. But while I hold Ether for exactly this reason (i.e. I believe the demand to use Turing-complete money will go up), I also hold BTC. Recall my earlier statement: “Bitcoin operates on a decentralized and game-theoretically sound immutable ledger”. That was a lie. We actually cannot know that Bitcoin is game-theoretically sound because it is too complicated to think about holistically. However, we have way more certainty of its soundness relative to Ethereum. This is because Ether is on a Turing-complete chain. The amount of things that can go wrong are unbounded.

And I am not talking about the billions of dollars lost to bugs last year alone. Those can be patched over time and it was not Ethereum’s fault, it was poorly managed smart contracts. No, the worry I have is that Ethereum will destroy itself. To explain this, I am going to turn to Basketball as an analogy.

You may not know this, but Basketball has been broken many times. In 1944, the extremely tall 

George Mikan (6ft10in, which is actually not that tall by today’s standards) began what is now known as “goaltending”. Basically, Mikan would stop the ball from going in the basket by jumping high enough to cover the net. A new rule had to be introduced to prevent the defense from touching a ball on “its downward arc” or while it was in the “imaginary cylinder” above the net. Basketball broke once more in 1947 when it was determined that zone defense was too powerful a strategy. This mode of play was clogging up the area in front of the net, ruining the pleasure of watching the game. And before 1954, there was no shot clock, meaning you could hold onto the ball as long as you wanted. To force a turn-over, you would need to physically strip the ball from your opponent. If you were in the lead, your team could simply form a fortress of bodies around your possessed ball. The league had to introduce the shot clock (forcing change of possession after 24 seconds). Nearly a half-dozen times, the dominant strategy of Basketball deteriorated spectatorship, by forcing the game into becoming dull to watch. 

Long story short, it took decades to iron out the kinks in Basketball and the rules continue to be altered today as players reimagine how to play it optimally. And the thing is, we might one day face an irreparable Basketball game. What makes Basketball fun to watch may one day be destroyed by some new dominant strategy that we cannot nerf without lowering the overall excitement of the game. (Soccer for instance is dangerously close to this point with its fake injuries.)

Bitcoin, like basketball, is a protocol. It is played for money, you follow the rules, and if you are good at it, you make money. Whereas Basketball’s protocol is meant to generate a spectacle, Bitcoin’s rules generate an uncensorable ledger that can process 10 transactions per second.

There have been papers that argue the rules of Bitcoin may also fail one day. Pool Cannibalism was a favorite of mine at one time. Basically, large pools make more money than small ones because they can save on operational expenses. Except, in a situation where there are two or three very large pools, they become incentivized to send their miners to their competing pools. Those miners provide valid hashes, but withhold hashes that meet the protocol’s target difficulty. The math shakes out where anyone who isn’t in a pool or who is in a private pool makes money. But it seems that it ends in an uncertain Nash equilibrium that doesn’t seem to destroy Bitcoin (but it could have!).

Of the other prospective problems, I can name two more fun ones. As you may have already heard, Bitcoin will one day stop emitting the coinbase reward and the incentives to protect the network will be entirely dependent on the transaction fees. Most people stop there and go womp womp the chain will die because there isn’t enough incentive to mine it. That’s smol brain complaining. The gigabrain’s complaint is that even if these fees are sufficiently high, there may be instances where miners would prefer to re-organize the chain so that they can cash in on a juicy transaction fee rather than mine a new low-fee block. And yeah, that would be pretty bad.

Another bad outcome results from “Selfish Mining”. It is possible that given a large enough stake in the network (maybe 5-20%) a miner could purposefully withhold blocks, creating a private chain. If they are lucky, they may mine three blocks in a row and create an awkward situation where they invalidate the work of 80% of the network by leaking two blocks at a time. This gives them a weirdly unfair advantage and makes them more profitable. This is bad. If a miner is more profitable than any other, over time it will gain increased hashing power relative to its competitors. In that instance, it will reach a point where it controls 51% of the chain and can monopolize all future block rewards entirely. That is very bad.

Those are almost all the bad things that we think can happen to Bitcoin. This is because it is an incredibly simple system to think about. And what is more, these bad things are either very unlikely to happen or far away in time.

The problem with Ethereum is just like 1930s basketball was unable to imagine a 1940s 6’10’’ Mikan, we are unable to imagine the next pool cannibals or selfish miners. That risk alone gives BTC a fighting chance over Ethereum. So, if you are conservatively trying to store the value of your capital and are not interested in speculating on the future demand of a “world computer,” Bitcoin is your obvious pick. It is just a de-risked Ether.

Let me give you one example, which I hope to write about in a later post. On Ethereum, I can create a contract that pays miners to censor users. Because there are lots of ways to compete over money in DeFi, it is possible that miner extracted value (MEV) can get A LOT worse.

Finally, the Etherean asks: okay, but why Bitcoin and not Bitcoin Cash or Bitcoin SV or Litcoin or Dogecoin… There are lots of Turing-incomplete blockchains out there. To that the answer is pretty simple. Bitcoin was the first. Switching from Bitcoin to something else breaks the spell of “store of value.” Building a Schelling point around a store of value is very hard. Changing a Schelling point is even harder (see the move from gold to Bitcoin as an example). It’s like why you should avoid looking for a partner amongst married people. Their willingness to leave is a sign that they may leave you too. And because Bitcoin was the first, it attracted some of the smarter builders and developers in the world, who understood the power of it as a Schelling point, and dedicated their lives to improving it cautiously and conservatively. 

Ultimately, what this means is that BTC appears to have a monopoly and little chance of failure. Ethereum is quite a bit more useful, but has a lot of competition and an unknowable amount of obstacles to overcome. Both deserve a lot of your respect. 

Tokenizing Protocols for Governance

For the past few years, blockchain protocols have been issuing “governance tokens.” These are usually transferrable tokens that can be used to cast votes, similar to shares in a corporation. Often these tokens are awarded to the core team of developers, investors, users (through airdrops) and eventually the public. Examples of protocols with governance tokens can be readily found on Snapshot, which is the primary service used to craft proposals, track who has tokens, allow their owners to vote with them and ultimately tally the results.

The vast majority of governance tokens are focused on DeFi protocols. Token holders generally end up becoming a DAO (decentralized autonomous organization), which has alignment with the protocol’s success and thus votes to better it (at least in theory). Examples of non-DeFi projects include: the Ethereum Naming Service, which acts as a censorship resistant alternative to DNS; Friends With Benefits, a web3 Soho House; Axie: Infinity, an NFT-powered video game.

Governed Games

In this series, my focus is on what I will call “Governed Games”. I will define these as games with two components:

1. The rules are, at least in part, enforced by smart contracts on an immutable ledger.

2. The smart contracts are designed to be upgradeable or alterable based on votes cast.

I have explicitly omitted the need for tokens, because there are other ways to tally votes. I’ve already written about governance in Dark Forest and dfdao’s use of Snapshot to vote on rules changes. In that instance, we used points scored in the previous round and weighted them by taking their log().

I’d like to present a second example today, a project I have been working on with Henry Caron on Axie: Infinity governance.

MMR/ELO for Votes

In standard protocol governance, the people who own the most tokens get the most votes. The idea here is that the people whom governance most affects should have the biggest say. The assumption is that these individuals will be knowledgeable or know knowledgeable people and as such vote intelligently and prudently.

Because $AXS, the governance token for Axie: Infinity, has not be widely distributed to players yet (although that is immanent on the roadmap), the current holders are not those with expertise regarding game balance. So while $AXS holders may be better governors when it comes to making budgeting decisions, they are not ideal candidates for micro-managing the competitiveness of each Axie card.

For this reason, we decided to base votes on match-making ranking (MMR), otherwise known as Elo (after Arpad Elo who invented this zero-sum ranking system). In Axie: Infinity, each player gains or loses MMR based on their performance against their opponents. Over time, players are sorted from incredible with >2000 ratings to terrible <800 (these players are so bad as to likely be bots). In this system, a pro player with 2000 MMR has twice as much say as a bad player with 1000 MMR.

Here the weighting system is arbitrary, it could easily be MMR^2 so that a pro with 2000 MMR is actual 4 times as powerful as someone with 1000 MMR. There could also be a threshold so that someone with <1600 MMR is actually not allowed to vote. All of these parameters themselves could be governed.

The obvious problems with this kind of governance is that players will want to collude to perform insider trading. Imagine that a cartel of pros secretly corner the market on a type of Axie card. And then Imagine that at the voting deadline, the cartel of pros vote to buff their cornered card beyond the realm of reason. All of a sudden the value of their assets skyrocket as players try to shift their holdings to include this newly powerful asset. This is a terrible outcome.

To prevent this kind of collusion, we proposed in our prototype that cards could only shift in power by +-20%. But I would further recommend AXS holders cast votes to ratify the changes or not. This way we have checks and balances from both the pro player side and from the stakeholders.

So anyway, this is what I am thinking about these days and what I will be writing on over the course of the next month.

Next Time

How to vampire attack Chess and create a governing body that can change the rules for the better.

Custodians of Governance

As of today, we are between rounds v0.6.4 and v0.6.5, and the Dark Forest team working very closely with its three established guilds: MarrowDAO; dfdao; and ordenGG. Their goal is to have each take the reins of a community round and run Dark Forest in their own way. The protocol is decentralizing itself somewhat radically as the core team teaches us how to maintain them game ourselves going forward. MarrowDAO has already completed its round with a focus on collecting specially named planets and incorporating sponsorships into the game. Dfdao is next, launching its community round Tomorrow. I expect ordenGG will follow shortly after.

Although we haven’t launched the round, the preparations have been remarkable. For one, we started by incorporating Snapshot into Dark Forest for the first time. For those who aren’t aware, Snapshot is a voting tool, where different addresses can cast ballots if they meet some threshold according to some on-chain data. Usually, this is the balance of some governance token, which one day (if not already) threatens to capture value from the protocol. In the case of Dark Forest, no token exists. Instead, we used the previous round’s score.

I think this is really cool because players are now “playing-to-govern”TM. It’s hard for me the write this blog post because the possibility space has been literally exploding in my mind. I want to tell you about how this can apply to all forms of games from Chess to Basketball. I want to tell you how the incentives are so well aligned because the best players are usually the most knowledgable about the game (as opposed to the most impacted or richest). I want to talk about how this is Sybil resistant governance because of how score is exponential in Dark Forest. But most of all, I want to tell you what the community voted on.

We asked our Discord server to propose rule changes for our allocated round. In total, 16 proposal were created. These proposals were casually vetted by the community’s reactions, with dfdao’s core team pushing the top three to Snapshot.

Now, because Dark Forest has exponential scoring, it would have been unfair to the majority of players to count each point as one vote (ordenGG would have had 777,777,777 votes while the median score was 270,000). However, it would have also been unfair to treat all players equally, especially because many did not play much at all. Here is where we used quadratic voting. Everyone got log2(Score) votes. That means if you scored 270,000 points in the previous round, you got 18 votes on the rules for the next. If you scored 777,777,777 points, you only got 9 additional votes for a total of 29 votes.

By the end of voting, my own proposal “Death of the Universe,” won by a narrow margin over “Fight of the Giants”. The gist of the proposal was to take further inspiration from Cixin Liu’s Dark Forest (the novel). Without spoiling it too much, we learn that planets are destroyable, something that is not currently possible in the canonical implementation of Dark Forest (the game). I recommended that an attack twice as strong as the maximum defence of a planet should simply delete that planet from the map, rather than take it over. And, to insure a continuously brutal battle, recommended a shrinking universe.

I think of all the things I accomplished in 2021, none seem so poetic as this. I’ll remind the reader that this insatiable quest to get exposure to Dark Forest, first by trying to get a job in the core team, then by starting a DAO of strangers to collect jpegs, has now become all too real. Although it is not a “canonical round” hosted by the core team, I got to re-write the rules to Dark Forest and now get to see how these really bold (and possible horrendous) decisions play out. My hat is off Gubsheep and Ivan, as they have stayed true to custodying their protocol, while also progressively decentralizing it those who would cherish it.

In a pretty robust write-up of how play-2-earn economies can bootstrap web3 communal development, Peter Pan describes the phenomenon of what I went through. I am paraphrasing, but essentially he motions towards game protocols rewarding players with ownership of the protocol because it is a strong path towards decentralization to a robust set of governors. Peter does not seem to have Dark Forest as a ready example, for he is focused and rightfully worried about a less composable set of games. What Peter is also missing, is that the simple threat of a token is enough to bootstrap the project. I would also like to clarify here that the majority of Dark Forest players and builders are probably not that motivated by rewards from play. The game offers a lot of intrinsic value both by existing at the bleeding-edge of technological achievement and by having great vibes.

If you are keen to play in dfdao’s community round tomorrow, feel free to DM me on Twitter.

Additional Notes on Vibes:

Dark Forest’s great vibes come from many different tasteful decisions made by its leadership.

1. It is entirely funded through donations that are both private and through Gitcoin.

2. It is entirely open-source.

3. It is really into decentralization. For instance, it selected xDAI (now Gnosis Chain) because it felt like it was the most decentralized scaling solution available at the time, given its tie to Ethereum. (Please note that Polygon was not truly tied to Ethereum, just well-bridged). This kind of decentralization maximalism virtuously signals to crypto OGs that this project is legitimate.

4. The game has an enormous amount of friction, which would normally be bad, but it forces strangers to work with each other.

Open-Source Savvy Players

Round v0.6.2 has started and almost the entirety of our newly formed seven-person team, dfdao is primed to win. The rules are slightly different, as the discovery of artifacts now contributes to score. Doing some paper napkin calculations, all other things being equal, they stand to represent half of a winner player’s points. While people have generally hand-waived at how artifact rarity is determined, I actually bother to go deep into the Solidity contracts to understand their randomization. This is actually pretty darn cool in itself, as you can imagine almost all web2 games will hide some facets of their design from players in their proprietary code. Already we are lesson 1: web3 games will necessarily incentivize players to learn to read code, even if they won’t write any.

Manan works overtime to pre-mine the map so that we can try to spawn near level 8 and 9 foundries (which are where I discovered those mythic artifacts spawn with certainty). Building on some public tools and using several of the whitelisted keys we received for participating in the previous round, we start empires in about four locations, with almost all of us piloting. By half-way through the round, our strategy appears to be working, we’ve discovered three mythic artifacts and have just barely overtaken some stranger named @ghst_gg. Our lead is short lived however, as my back-of-the-napkin calculations did not anticipate a sharp increase in leading players’ score efficiency between rounds. It looks like scoring clearly favors extracting silver (which was the only scoring mechanism of Round v0.6.1. We only have a few more mythic artifacts to chase, so our empire is on track to take second place at best.

Ownership Means Ability to Sell

Desperate to win, I start making “deals” for score. Something like: “give me your empire and I will give you a mythic artifact NFT”. After all, we’ve shifted focus from getting mythics as collectibles to seeking out Valhalla planets, which are the NFT rewards for winning rounds. Luckily, I am brilliant salesman, land a couple juicy empires and we begin to edge closer to the lead! In order to seal the deal, I try to buy the third place player, ghst_gg’s, empire. Herein lies lesson 2: in web3 games, you can own everything, which means you can sell it too. Victory can be purchased.

Unfortunately, through some miscommunication and miscalculation, I fail to capture ghst_gg’s empire. In fact, I give him the idea to start buying empires himself, but not before he launches a smear campaign against us for trying to do it ourselves. I have upset the mores of the community. And later Ghost has begun to pay more for points than I am willing to win. Meanwhile, I have run out of mythic artifacts to trade and have no leverage to acquire more points.

Memes Win Wars

And then it hits me, we should play off people’s hope, loneliness and greed. I begin telling people that they can give us their empires in exchange for “being remembered”. We create a Discord channel called “The Remembered” and let strangers into a newly formed public part of our community. When people ask what it means to “be remembered”, I tell them that we have grand, but undetermined, plans and that we will remember them for the help they offered. It is clear that players are intrigued about joining something bigger than themselves. They are also curious about potentially reaping significant rewards (after all, we are in the age of reverse air drops where people are unexpectedly rewarded for having barely done anything).

Right away, the demand to “be remembered” is overwhelming. A few people give us their empires, but the client and our computers are having a hard time keeping up. The size of our empire is becoming too unwieldy to handle and only David can still pilot. We end up creating a plugin so that they just give us their valuable planets (turns out beggars can be choosers). We also create a plugin to help them better share their maps (because we cannot pilots empires in the regions of space we cannot see). Once again, we are on pace to win! With Lindsay ramping herself up to head a web3 community, we are able to corral everyone into our Discord to build more hype.

The smear campaign escalates. We are called whiny beggars. Seventh and third place players stop scoring points and ghst_gg begins to grow at three times its previous speed. While we were rustling up average and below average players, anger amongst the better players had them coalesce into a galactic superpower. We watched in despair as our lead was eroded. We scrambled to recruit more members, but it felt like bailing water out of a sinking boat. With only a couple hours left in the round, I go to bed exhausted as the rest of the team throw a hail Mary’s into a loss.

Here are “The Remembered”: @bitcoin4bernie, @clemensWittmann, @jacksondame, @eichenkundinger, @algoriddims, @ner0nzz, @TheVelorum, @nationalbank_, @spencecoin, @scotato, @retinoic, @stuneberg.

Looking back, we accomplished so much. We were the first DAO to ever reach first place in Dark Forest. Ghst_GG, it turns out, was a guild that was previously focused on Aavegotchi (another composable web3 game with cute ghosts). They’ve since renamed themselves to orden_gg and now focus on both of those games. I feel like we came for the king and missed. Not only did we not get first place, we caused such outcry that our opponents doubled down, and by making the round such an incredible nail-biter with people working overtime in concert to win, the perceived value of the stakes escalated. Our opponents going forward, were to become full-time Dark Forest players, while each of our DAO members had real lives and jobs to return to. Dreams of ever winning were fading fast.

The only thing we had going for us was our creativity. And in keeping with Cixin Liu’s axioms of cosmic sociology, the opportunity to one day make some kind of technological leap. From my perspective, our only chance at beating Orden was to make the “remembered” meme more powerful and to scaling our tooling to take in more player contributions. Anthony begins working on improving our plugins to create an automated system of receiving and tracking contributions for the following round. Yitong works on creating a commemorative NFT for “the remembered” depending on how much they sacrificed to us. Lesson 3: all talents are useful in web3 gaming because playing here looks nothing like playing in web2.

Bad Luck

As our plugins for donations are improving, the prospects of winning Round v0.6.3 are looking somewhat reasonable. Except, when the rules are released, a new objective is announced (closest to the centre of the universe wins) and the “gifting” of planets is removed from the game. Essentially, all of the tools we built are useless because they all depended on other players donating their planets to us. It is probably for the best, as I am on a beach vacation at the time.

Interestingly, the DAO is taking on a life of its own without me. The core team decide to recruit Scott and Velorum from “the remembered.” They take time during the round to build something new, an artifact market which lets players sell artifacts they have found during the round. This is a pretty astounding achievement, because through third-party development on a composable game, we organically turn Dark Forest into a play-to-earn opportunity. Whereas previously players could only safely trade artifacts through trusted middle-men, the market not only removed counterparty risk, but 100x improved discovery of buyers and sellers. Lesson 4: in composable game design, you only need a small core team to kickstart the game loop, the community can build the rest.

Permissionless Scaling

Although I think removing gifting is an inspired game design choice, I let Gubsheep know that we built cool things we cannot use anymore. He ends up giving us some assurance from the core team that round 4 will bring back gifting, enabling us to try out our new tools. With that great news, Anthony and Scott end up connecting with Ansgar to improve the security of the gifting plugins by creating a smart contract player. Basically, we whitelist the smart contract to act as a player in Dark Forest Round v0.6.4 and call it “The Astral Colossus”. The contract itself cannot play the game except for execute the point-scoring function. We end up creating a plugin whereby players gift the Astral Colossus their score-able planets, it claims the points and returns to the donors.

The Astral Colossus ends up coming 34th in the round. While that is a far cry from second place, the technological achievement is remarkable. We believe this might be the first time a smart contract played any web3 game, let alone created a permissionless method for over 50 people to participate. The Astral Colossus also kept track of people’s point donations in order to later reward them pro-rata. You’ll notice that the sixth greatest contributor was our sworn nemesis, @orden_gg, who ultimately came in first place with over 10 times our score. Interestingly, Orden won the round by over 150m points over second place. Had they donated that excess to the Astral Colossus, they would have gotten 2/3 of 9th place. Meanwhile, Orden has already thanked us for our service and plans on deploying the smart contract itself in the future. We are really good at taking missed shots at this king.

OMFG. We made a smart contract, as a public good, that coordinates players (who would never be able to win alone) to win permissionlessly together. While it might seem trivial in the context of a videogame, there is no reason to believe that this kind of complex human-motivating protocol cannot exist in the real world.

Quick recap on our collectivism:

1. Round v0.6.1: We form a dao of doxxed people that are hand-picked for their trustworthiness and ability to contribute.

2. Round v0.6.2: We partially open up the dao to people we meet online to discuss handing over empires for potential NFTs.

3. Round v0.6.3: Dark Forest team ruins our plans accidentally.

4. Round v0.6.4: We launch a tool to let total strangers hand over their empires to guarantee a score that might lead to them eventually getting potential NFTs.

The next steps towards decentralization are in removing counter-party risk where the points scored in the smart contract are the votes required to use assets that smart contract wins. Lesson 5: decentralization is a progressive process, as each step is complicated to predict, understand and build.

Additional Notes:

1. I still think I can meme our second place planet reward as the most important NFT rewarded in all of web3 gaming history, because I think our DAO made history that round.

2. We are friends with Orden_gg now.

3. Orden_gg disputes having bought any empires. They could be “just that good,” or psyops.

Permissionless Play Automation

When you can play for external rewards, the normal incentive structures of games change. For example, if I play Chess against a friend, I will not us an AI tool to cheat. The fun of trying my best against a friend outweighs the benefits of a valueless win. On the other hand, if we are playing for a prize with important value, we are both incentivized to cheat. In fact, you can say that we are playing a new game called How Good Is Your Chess AI?, in which players compete to build better and better software to win at Chess. The scope of the game then is dramatically expanded to something akin to running a software company.

Playing permissionlessly in web3, looks a lot like the latter case. In this space, with real prizes, open-source code and a permissionless backend we call blockchain, you need to build to win. We are then presented with an enormously impactful constraint, which is that web3 games need to still be fun in this context. In this blog post, I explore how Dark Forest can remain an interesting game in this context.

Some of the Code is Law

It is May 2021, Dark Forest v0.6.1 has been released and Ansgar Dietrichs, a protocol researcher, decides to direct his knowledge of web3 and Dark Forest’s codebase towards winning the round. While he takes several steps to give himself an edge, perhaps the most shocking is that he alters the game client.

For those of you who aren’t familiar with what this means, allow me to expand. Online games can be thought of as having two parts, the client (where people craft their moves) and server (where moves are received and the game-state is changed). In web2, the client is always closed-source and permissioned to communicate to the server. By design, it should be impossible to alter it, as doing so would enable certain unfair moves. When web2 hackers succeed in undermining the integrity of a game, it is often because the client has been designed to do some of the rules policing and that policing has been circumvented. In Dark Forest, because the game-state is on a permissionless blockchain, the integrity of the the client cannot be enforced. In other words, anyone can communicate anything they like to the smart contracts on chain.

Ultimately, what Ansgar did was not rocket science. Once he noted that the game waited to finish making a move before communicating the next one, all he had to do was alter a few lines of code. All of a sudden, Ansgar was able to parallelize his moves and accomplish 10x what anyone else could. You do not need to be a brilliant general when you can issue 10x more instructions than your opponents. And in this case, because Ansgar was able to deploy a bunch of publicly released scripts (which he had admittedly also spruced up), he was able to automate his moves at about 50% the efficiency of a dedicated human pilot. Comparatively, Ansgar could be 5x more capable, while never worrying about sleep*.

I asked him if he thought this was cheating. He responded:

I 100% don't consider that cheating whatsoever. I think the beauty of an open-source game is that the base rules are on chain for everyone to see, and as long as you follow them, fair game. The vanilla interface is really just a "base suggestion" of how to interact with the game, and I think the Dark Forest team would agree that they very much hope more people will write their own custom clients in the future. They even actively help with that, by e.g. publishing npm packages one can use for building df tools…

Of course, cheating is a complicated concept, with a lot of grey area and social construction. Fun fact: one of my academic supervisors, Mia Consalvo, actually wrote the book on it. In some communities, what Ansgar did would have been reprehensible, but the web3 ethos here is pretty clearly on his side.

Sybil Attacks in the Absence of Official Clients

Sybil attacks are bad things hackers can do when they can pretend to be lots of different people. It is generally bad for a competitive online game when one player can pretend to be many players.

Because Dark Forest cannot enforce a specific client on its players, it cannot preventing Sybil attacks in the usual manner. Often, for online games, the client will check against a database to see that the player has a unique email address, or even cellphone number, on top of mitigating botting in a variety of ways. A permissionless game like Dark Forest cannot*. This is a property of blockchains where individuals can create unlimited numbers of accounts. In other words, one player can pretend to be 20 players.

Interestingly, Dark Forest does not have to work very hard to prevent Sybil Attacks because players grow exponentially in power. If you have the money, computational power and bandwidth to make 10,000 moves, you are probably better off spreading them across 5 accounts than 50. While this is more than piloting one account and thus not optimal from a web2 game design standard, the problem is at least bounded.

A good model for thinking about how this works is quadratic voting, but backwards. In quadratic voting, you vote with the Log() of your tokens. Having one person with 100 tokens is better than one person with 10, but much worse than 10 people with 10, despite both groups having the same number of tokens. The quadratic voting mechanism is designed to give more weight to people who hold more tokens, while still allowing the majority of people to have some power too. Quadratic voting is a mechanism which incentivizes Sybil attack. You would ideally take your 100 tokens and put each one in a unique address to maximize their value (Log(100) < 100⋅Log(1)). The converse then, which we find in *Dark Forest*, is exponential scoring, which disincentivizes Sybil attack (100^2 > 100⋅1^2). Pretty cool. In fact, as we discovered at dfdao, 7 people playing under one account is quite a good idea!

Incentivized Sharing

In order to disincentivize players from establishing permanently dominant positions, Dark Forest’s core team rewards players for open-sourcing their plug-ins and altered clients. Here the hacker’s dilemma is to either:

1. hide the code in hopes that it can confer an advantage across multiple future rounds to win several points,

2. or reveal the code and win immediately a prize for transparency.

While this might be a tough nut to crack, there is a meta-layer of incentives motivating the playerbase. Many players believe that revealing the code of their own plugins is good for the growth and future excitement around the game. If the game grows, then it is safe to assume value of the historic prizes will grow (as collectibles or maybe even as governance tokens).

Here then is a radically unexpected property of giving hackers ownership of their hacked objects. All of a sudden they want the game to stop being unfair. Their immediate incentive is to tell everyone how they hacked the game so that it can either be patched (or so everyone can use the hack equally). This in turn creates a virtuous cycle of players building tools to better play the game for other players to eventually use.

I think this is so neat that I want to say it a few times differently.

1. When you give hackers governance tokens, they are incentivized to remediate the situation.

2. If hackers can only steal governance tokens, then at a certain point they want to stop stealing and ensure that what they have stolen retains value.

Wow. I can’t even.

And I can speak from experience because in round v0.6.2, we, and everyone else who made it to the top of the leaderboard, used Ansgar’s client.

Additional Notes

1. Ansgar’s victory was not quite as easy as I made it out. You see, to broadcast a move to the blockchain, you need a node on the network that will listen. The public nodes for xDai were throttling everyone, so Ansgar also needed to spin up his own node to push his moves in without hindrance.

2. Dark Forest currently has a whitelist to limit players in order to prevent the xDAI chain from being overloaded. That said, one player can use many whitelisted keys as the keys are not strictly tied to specific people.

FOMO on Friction

It’s February 2021 and I stand no chance at winning Dark Forest v0.5. My recent acquaintance (and now dear friend) Omar Bohsali offers me a whitelisted key to play mid-way through the round, explaining that he hasn’t the time. Curious to learn more about the game I peruse the blog.zkga.me and find myself only asking more questions. I boot up the in-browser client and connect my Metamask wallet to xDAI for the first time. The devs are kind enough to automate the funding of my newly created public/private key pair with just enough xDAI to play a couple moves. But when I run out, the barriers to entry are remarkable. For one, I have to use Uniswap to trade some $ETH for some $DAI, then I need to use the xDai bridge to bring it over to the side-chain. I persist because I have previously found that overcoming friction in crypto is extremely profitable (but I do mistakenly only bring 5$ over paying something like 40$ in gas).

Right away it is clear that I am playing a slower, uglier, friction-laden, MMO version of Galcon. You start with a planet, it creates ships. You can send the ships to take over nearby planets to increase your overall ship production. You expand exponentially until encountering resistance in the form of other players looking to take your planets. One significant difference in the gameplay however is that the player can capture artifacts on the various planets and that these are withdraw-able NFTs.

Now, if you recall the NFT market is booming, with CryptoPunks and the likes climbing in value. I myself have been middle-curving myself out of the entire bull-market and am desperate to ape into something I can understand. These Dark Forest artifacts appear to be my alpha, and so I decide I am going to commit to collecting them.

Annoyingly, there are some red flags, for the core team is saying that the artifacts have no value and will almost certainly not be reusable in future rounds. My take after reading the blogs however, is that these ultra-goody-two-shoes devs are going to want to decentralize their gaming protocol and they are going to have to do that based on on-chain records (they still haven’t to this day). The artifacts are the perfect target for reverse airdrops. Even better, I am hearing that they won’t accept capital from investors (they still haven’t to this day), claiming that their grants are sufficient to keep them afloat and that they really want to be “community first”. I just read that as “governance tokens incoming” and double down on my conviction.

The problems to overcome in getting said artefacts are as follows:

1. I am bad at the game.

2. I do not have enough time to play the game.

3. The game is fully on-chain and the number of moves to make are exponential in number, meaning that I need to craft plug-ins to script my moves better.

4. I am bad at making plug-ins and do not have time.

5. My computer is too weak to hash out the map. (For context, the game is played in zero-knowledge with a deterministic map that is unknown to anyone until they compute a hash for a given coordinate and then use the protocol to decode the resulting hash into either an astral body of some kind or just empty space).

But as stated before, the more friction, the more conviction. To solve these problems, it is obvious to me that I need to recruit a team of both players and engineers.

Assembling the Avengers

Omar is an obvious first target, he is a brilliant 100x engineer who had recently left Coinbase and was the entrepreneur in residence at Paradigm. Although he is pretty busy, he is in “an exploratory mode“ and agrees to practice using Terraform to provision AWS servers as needed to hash out the map. He also helps here and there with modifying plugins to give me a small edge. But before we can really take off, round v0.5 ends and I have basically a bunch of low grade artifacts that must be worth garbage. Had we started sooner, we may have done half-decently, a fact which keeps me excited.

Meanwhile, the devs invent a newer, higher, rarer tier of artifact called Mythics and award the first of eight to Jordan Spence for having contributed enormously to the management of the community. The other seven go to those who open-sourced their top-tier plugins for the community. This is obviously super annoying because I am a shitty dev and will never make something worth a Mythic artifact. It is also incredibly brilliant and makes me even more bullish on Dark Forest, because now gigabrains are being paid in Monopoly money to build out what must be the biggest permissionless nerdsnipe-of-a-project in history. I chat with Omar and we agree to take the next round more seriously and start on time.

Over a month later, Round v0.6.1 starts. There are some new rules created by new smart contracts, which break most of the pre-existing plugins, and Mythic artifacts are now available for discovery in the universe. I have just left my forensics gig and am starting a brand new job at DeFi Alliance in a week, which means I have an inordinate amount of time to try to win. However, with the plugins broken, I am not able to automate play, meanwhile actual devs are making the minor changes needed to keep on trucking. Omar is too busy to help full-time, he is building another composable piece of open-source software, the canonical Uniswap V3 Staker. Me, desperate to take first place, look to the leaderboard to start recruiting.

Interestingly, because Dark Forest has some diplomacy requirements, people are incentivized to dox their public keys with their Twitter handles. When I decide to recruit, someone named @Manan19 is winning by a solid margin. I notice there is a Manan in the Discord server, so I reach out. I decide that my strategy to breaking the ice is to flatter and ask for strategic help.

Through our conversations, I learn that Manan is currently managing engineers at Coinbase, which bodes well for his domain expertise and ability to contribute (but also now I know where he works so, if he fucks me, I can seek retribution). It takes a little while to sway him, but he eventually agrees to partner up. I offer him Omar’s AWS servers and Manan provides the plugins.

The problem is that I need to take care of my kids once in a while and even sleep sometimes. But Dark Forest never sleeps during a round, so I need to find a solution. I start working my way down the leaderboard and find the best doxxed players I can. Using a similar strategy of flattery and tit-for-tat offering, I partially recruit Zha Yitong who is writing a blog about captaining an empire in Dark Forest. It turns out he is also at Coinbase, doing UX/UI (this will be helpful later). I also manage to coax in wunderkind, David Zhou, who is building learning materials. Luckily, he is still in college, meaning he can play 10 hours a day. He offers to co-pilot the empire.

Despite my efforts, our collectively piloted empire is lagging in the top 30s. Given that we still aren’t enough, I decide to turn to my extended family. I start on my nephew Anthony, who has just graduated in CS and Poli-Sci and is looking for career advice. Easy, I tell him to quit all his internships and prospective jobs and work full-time on Dark Forest. Poor kid hasn’t even create a Metamask wallet (do they teach CS grads anything?) and here I am altering the course of his life for jpegs. I remember the conversation well, my kid is sleeping in the stroller and so I have 90 minutes to red-pill the heck out of him. It was thoroughly exhausting, but ultimately he concedes to at least help in the immediate. Spoiler alert, he ends up not only giving up all his normie web2 prospects, he becomes the only full-time dfdao member, effectively retraining himself as a Solidity dev and ZK nerd.

Finally, realizing that the Dark Forest team and my dao are going to need a media arm and some gender diversity, I recruit my sister-in-law, Lindsay. Earlier that year, I had forced her to read Cixin Liu’s trilogy, of which the second book, Dark Forest, is the inspiration for the game. She, like me, had become obsessed with the novel (which is reasonable given its the best piece of science fiction ever written). At the time, she was taking a break from a PhD in literature, and my hope was that she could manage our Discord communications and social media. My secret hope was that playing this game would red-pill her into crypto too and gainfully employer her in the long run. Spoiler alert, she’s was hired by Zapper only weeks later to help lead their own content strategy. Today she is the glue that keeps the operations and socials of dfdao running.

The Denouement

Together, the seven of us begin meeting regularly and start to feel like a team. David, Manan, Yitong and the collective empire I started, ended up placing in the top 63. We never find a Mythic artifact, but using The Graph, I find someone who did and pay him 500 xDAI for it. It’s name is Ek’thun and he stood as the dao’s pfp for a short while. At the time, artifacts were connected to a GPT3 AI that was fed a few pages of lore. Lindsay and I decided to stream an interrogation of it on Twitch.

It was at the end of round v0.6.1 that the core team introduces NFT rewards on Ethereum called Valhalla Planets. Everyone in top 63 gets a planet jpegs based on how well they did. Plugin creators are also reward. Interestingly, the core team thought our interview with Ek’thun was awesome enough to reward us with another mid-tier planet, validating yet another thesis that they would be wise enough to recognize not only victories and plugins, but communal memes as well.

Now that a new Schelling point for NFT collection emerges, I revise my thesis that Mythic artifacts hold value. It also makes double sense to reject them, given that new artifacts are capable of being “used up” a.k.a. destroyed for in-game effects. (You see, it is bad game design to make in-game objects you’d rather not use and keep instead for collecting.) I feel a little sad about my 500$ Mythic artifact purchase, but given that it enabled the interview, and the subsequent planet reward, it’s really fine.

Okay, now let’s take a step back. Because I wanted some collectibles to avoid being left out of the NFT boom, I reached out to four strangers and two family members to join me in creating what amounts to a gaming a web3 gaming guild. I had not even heard of NFT farming DAOs like Yield Guild, and just assumed I should call us a DAO and not what is obviously a web3 gaming guild. And, because no one else had created a guild or DAO to play Dark Forest, we hubristically called it dfdao (or @d_fdao because of a bug in Twitter handle naming). The lesson here is that a strong financial incentive created from an absurdly under-legitimized conviction can lead people to independently do amazing things with a web3 game where players can threaten to value-capture.

Worshipping Protocols

But more than just refocusing our DAO on a different collection of jpegs, I see an opening for a more ambitious goal. In wrapping up the round, I host a meeting pitching to everyone in the group why we should team up going forward. I share my entire thesis and explain that we are particularly well situated to capture Dark Forest’s value. From my perspective, the game is like the Ethereum protocol, which the core team will maintain through a a not-for-profit body or Foundation. Our DAO, on the other hand, is closer to Parity, a for-profit organization building sellable products on top of the apolitical underlying protocol layer. If the Dark Forest team doesn’t want to sell out, that is their prerogative, we can do it for them and share the wealth as a group.

Said differently, Dark Forest, isn’t a videogame, but the rules that underly it. In the same way that a given Chess set or Chess app is not Chess itself. The inventor of Chess does not capture Chess’s value. Rather, it is the ecosystem that has gathered around this rule-set that has. This is the way of composable games. And, done right, they can suck in players to perpetuate themselves forever. Chess is what I like to call a “ludo-god”, a protocol that continuously memes itself from mind to mind, thanks to those who worship at its altars. Dark Forest threatens to do the same.

We aren’t really a DAO or a guild then, but something closer to a Lovecraftian cult. Each of us is a cultist working to summon this new ludo-god into existence.

Additional Notes:

1. As of today, we’ve grown by two members, Scott (who now works at 0xHorizonGames) and Velorum (who has taken on a more full-time role in the dao). Omar has since taken a step back to focus on building the Google search for NFTs at Center.app. For about 20 minutes Paradigm’s CTO, Georgios Konstantopoulos, was counted amongst our membership.

2. In the early days, we did not trust each other, but were also too lazy to use the multi-sig we created. My own threat modelling made me feel sufficiently safe however because everyone was doxxed internally with much too much reputation to lose in screwing over the rest of the dao. Only as of a week ago have we started using a generously configured multi-sig (1 of N), mostly for bookkeeping purposes.

3. In my next post, I’ll talk about the emergence of competing DAOs and how we managed to lose our lead, coming second in the round.

Permissionless and Composable: A New Value Proposition

Web3 introduces novel value-capture opportunities through smart contracts and their tokens. One resulting side-effect is that builders are incentivized to open-source their code. This is because web3 code isn’t copy-paste-able, its authentic instantiation lives at some Schelling point on a distributed ledger. Duplicating the code does not duplicate the value that it is escrowing (see Bitcoin’s thousands of forks for examples). With the code open-sourced and behaving permissionlessly, allowing anyone to interact with it, future builders can build their projects atop it. And if the superimposed protocols are designed a certain way, both will capture the upside of any added value. The property of being able to build on top of other projects is called composability.

DeFi Summer was an exemplary period of permissionless composability. MakerDAO produced a smart contract that could take in price feeds to mint a USD stablecoin called DAI, securely backed by ETH. Uniswap created a smart contract that could swap that DAI for ETH and vice versa. Aave created a smart contract that could lend out ETH if it was over collateralized by DAI. From these primitives, users could suddenly take short or leveraged positions on ETH/USD. From nothing, these three protocols created the first killer blockchain app, a near-frictionless financial market.

Some people called these contracts DeFi Legos, detractors called them DeFi Jenga (given how many times the stack collapsed due to unforeseen bugs). Regardless, the power of composability in DeFi has been made clear by that glorious/terrible time. Today, we are witnessing GameFi summer, where in-game assets are able to plug-and-play with DeFi. Axie: Infinity for instance has players grind out $SLP, a token required to breed more Axies (cute fighting Axolotls). That $SLP can be borrowed, traded, staked, etc. However, the buck stops there. Axie: Infinity is a closed source game that is too centralized to build on. And because of Axie’s success, competitors have little impetus to make their own games composable.

In the following series of essays, I hope to chart my journey into a particularly notable exception. A truly composable web3 game called Dark Forest. Launched in September of 2020, Dark Forest created a persistent “4X” space game built on zero-knowledge proofs. Imagine a giant Chess board with hundreds of players playing on squares you cannot see. Each player submits a proof with each move that their piece was truly theirs and that it moved according to the protocol’s rules. Only once you visit the relevant squares can you come to make sense of the previous proofs provided and learn the local state of the game.

The entire game’s state is on-chain and the smart contracts controlling it are publicly known. This means that the game is front-end agnostic. You could re-skin the game to be about pollinating flowers, so long as the moves generated that by your UI were in accordance with the protocol. It also means it is player-agnostic, enabling AIs, bots other smart contracts to play it. To the point where it is literally impossible today to play the game yourself. I am reminded of The Simpsons episode “The Secret War of Lisa Simpson” where the commandant played by Willem Dafoe offers a commencement speech to the cadets:

“The wars of the future will not be fought on the battlefield or at sea. They will be fought in space, or possibly on top of a very tall mountain. In either case, most of the actual fighting will be done by small robots. And as you go forth today remember always your duty is clear: To build and maintain those robots.

When I first played Dark Forest, almost one year ago today, I knew deep in my bones that this was the most important game I had ever played. And let me tell you, this is a surprising feeling for someone like me. I have been gaming basically full-time since I was 5. I taught game history and design for seven years, and my obsession was such that I was even motivated to complete a doctorate on the study of protocol design in games. (I do not mean to brag, I just want to impress upon you that there is a tectonic shift at work here.)

At the time, I was working as an expert in blockchain risk management, essentially creating reports for financial auditors on the cryptocurrency holdings of various companies. and foundations It was an interesting challenge, but Dark Forest felt like such an opportunity that I decided I would try to join the core team. In my mind, it was obvious that they would hire me, 3 years of full-time experience in bear-market-crypto, certification in Solidity development, and of course my previously mentioned gaming background.

I was so wrong. Instead of getting a job, Alan told me to work on his project for free. At the time, my self-entitlement made me feel quite frustrated. Of course, this was before I understood DAO life and the web3 world of just building first and reaping rewards later. But from my perspective, this was the best project in the world and I would be damned if I was going to be sidelined. The joke was on me because since then, I’ve worked part-time on Dark Forest for free. Actually, worse than free, I’ve had to spend probably a few hundred dollars for the privilege.

Here are the forthcoming chapters in the story:

“Forming a DAO”: Given my determination, I decided to recruit an elite team of players in order to outcompete the players who were all going at it solo. These were a mix of 6 technologists, gamers, and writers (from Coinbase, Zapper and Rainbow) who would go on to build a vibrant community of 225 members.

“Creating Plug-ins and Altering the Client”: Because the rules are only enforced on chain, players are incentivized to automate their actions through custom code and re-writings of the game client. We did that to great effect, coming second place in the round.

“Designing a Smart-Contract Player”: Extending on the first idea of winning as a group, we created a protocol for a collective win that engaged 75 players to pool points permissionlessly. I believe this was the first time that a smart contract was deployed to coordinate players around sharing a victory.

“Governing Rule Changes”: As Dark Forest decentralized, it gave our DAO the blessing of deploying the smart contracts for the first round of 2022. Applying my training in game design, I proposed a series of rule changes and over 100 community members voted to approve them.

Coming soon^TM