At the beginning of 2022, Dark Forest conducted a two-week testing round for version 0.6.5. Amidst intense competition among over 1,800 accounts, P DAO performed exceptionally well, with seven of its members securing positions in the top ten. Today, we are honored to invite the founder of P DAO, Elf, to join us in reflecting on those thrilling battles and sharing insights into the future of Dark Forest.

Introduction

ddy: To start, could you please introduce yourself briefly?

Elf: Hello everyone, I’m Elf. I first encountered the blockchain industry around 2018 and have been actively involved ever since. Currently, I work as a product manager for a wallet, and I am also a member of both P DAO and 33 DAO. In my spare time, I focus primarily on mining and have a strong interest in PoW and PoS-related topics.

ddy: What type of organization is P DAO?

Elf: Since I frequently participate in testnet activities and many people would ask me related questions, I created P DAO for us to exchange ideas and engage in various projects together. P DAO has over 30 members and is, to some extent, a very closed DAO. We haven’t had any new members join for the past two years, even though many have expressed interest in joining. However, adding new members involves a trust cost because the existing members are all very familiar with each other and have met multiple times offline. We are involved in a wide range of fields, from various types of mining to many GameFi projects. Overall, P DAO is a team of friends with a high level of trust.

ddy: Can you give some examples of the testnet activities you’ve been involved in?

Elf: We’ve been involved in various PoW and PoS-related testnets. Initially, we ran Ethereum nodes, and later we participated in testnets for projects like Mina, Casper, Marlin, Filecoin, Chia, as well as Spacemesh and Aleo.

Thrilling and Intense Battle Years

ddy: What prompted you to deeply engage with playing Dark Forest?

Elf: It was quite an intriguing journey. Initially, I noticed some friends from the community playing the game, which led me to join the player chat group. The atmosphere there was vibrant, with discussions about how enjoyable the project was and various screenshots being shared. Intrigued by the buzz, I decided to try the game myself. However, I soon found myself being overwhelmed by enemies in the universe. Frustrated, I began exploring various plugins and scripts, including GPU map mining tool, and started strategizing with friends who had spawned in the same region on how to mount a defense. This process proved to be immensely engaging, and I recognized the value of teamwork, prompting me to invite more friends to join the game.

ddy: The official testing phase of Dark Forest coincided with a bull market, and the time commitment required for highly collaborative participation was quite substantial. What motivated you to engage so deeply despite this?

Elf: The primary reason was that I was genuinely captivated by the game. The gameplay itself was incredibly enjoyable. At that time, we didn’t have well-developed plugins, so we established a 24-hour schedule. Team members on night shifts would remotely monitor the computers of others. If someone’s planet was under attack, we would call them to wake them up. Another factor was the project’s appealing background. The game had a very geeky and cool vibe, and participating in the test provided an opportunity to win Valhalla NFTs. Additionally, there was the anticipation of receiving airdrops as early testers, which added to the incentive.

ddy: P DAO is somewhat mysterious and low-key. I’m curious, at the height of your participation in Dark Forest, how many members from P DAO were involved, and what was your general strategy?

Elf: P DAO has over 30 members, and at the peak, more than 20 were actively participating Dark Forest. As for our strategy, we chose to forgo competing for the first and second places and instead focused on getting as many team members as possible into the top 63, as those in the top 63 would receive special reward NFTs. At the beginning, everyone spawned randomly in the universe, and our goal was to group together and develop in the same area. During the mid-game, some members ended up near high-level planets (level 8 or 9), and we established these high-level planets as our main base, providing protection to nearby teammates. We also coordinated scores within the team, which could be somewhat controversial. For example, I reached third place, but the gap to fourth was substantial, and moving up to second was unlikely. In such cases, the points from activating artifacts or capturing planets didn’t significantly affect my ranking. Therefore, we coordinated within the team to distribute the energy from some planets, allowing other members in the community to capture them and gain points. By the end, I recall that over 20 of our team members were in the top 63.

ddy: That’s quite impressive. I recall that after v0.6.5, there were various opinions and discussions in the community about the Gatling incident. During v0.6.5, it was possible to significantly increase energy regeneration speed by having multiple ships land on the same planet, with the effect stacking. Initially, it was expected that up to five ships could land on a single planet, but the smart contract allowed for up to ten ships, resulting in a 1024x increase in energy regeneration speed. What is your view on this issue?

Elf: Objectively, I think this is problematic, but at the time, the official team did not declare this behavior as a violation. Given that Gatling had already emerged in the universe and was deemed acceptable by the officials, we decided to participate. We used small planets with speed bonuses to gather our team’s ships and created our own Gatling setup. However, I personally think this kind of gameplay isn’t very enjoyable and, theoretically, it should have been prohibited.

ddy: As a team of around 20 players, were there ever instances where there were differing opinions or ideas about a strategic decision in the game? How did you handle strategic decision-making at the team level?

Elf: I don’t recall any such instances. Since our team members are quite familiar with each other and have established deep connections, we generally approached the game with a relaxed attitude, treating it more as a game than a high-stakes competition. This made it easier to manage strategic decisions collaboratively without major disagreements.

ddy: Dark Forest is a competitive game where only the top 63 players in each round can earn reward NFTs. Besides internal team management and coordination, P DAO must interact with other guilds or players. What kind of diplomatic strategies does P DAO typically employ during gameplay?

Elf: We prefer to focus on our own team, aiming to get as many of our members into the top 63 as possible. During the game, we do interact with other DAOs or players, and I recall there was at least one argument. In a war-themed game like this, the primary objective for a player guild is to maximize their advantage—pushing down anyone blocking our progress, as long as they aren’t part of our team. This is the ideal strategy in a competitive game. However, Dark Forest differs from traditional strategy games in that it involves practical communication. For example, if you happen to be in front of us and there are several players nearby, including one with a level 9 planet, we would typically want to eliminate you directly if we can win. But due to Twitter account verification, some players recognized us or our friends, leading to private messages asking us not to attack. In such cases, we had to cease fire, spend more time finding alternative routes, or look for other high-level planets instead.

ddy: How do you balance game objectives with personal relationships in such situations?

Elf: When we encountered this issue, we tried to anonymize all Twitter accounts as much as possible. However, some friends still knew who we were. If the communication was good and someone asked us not to attack, we might have let it go, especially since the area we occupied was quite large and the points were sufficient. The reason we pushed others out later on was that we didn’t want to overexert ourselves. Given the finite number of points available, once we had secured the points in our surrounding area, there was no need to engage in further battles.

Discussion on Game Features

ddy: Dark Forest is a fully onchain game compatible with EVM, where each player can only submit a few transactions per minute. Do you think this will affect the enjoyment of the game?

Elf: I don’t think so. For example, in Dark Forest 0.6.5, the game lasted for two weeks, and after the second day, each move operation took a long time—sometimes several hours. So, a slower transaction rate doesn’t matter much. However, sometimes onchain transactions can be lost, but switching RPCs usually improves the situation. Frequent transaction losses can significantly impact the gaming experience.

ddy: From a player’s perspective, what are the major differences between playing a fully onchain game that utilizes zero-knowledge proofs and other games?

Elf: Personally, I don’t think there’s a significant difference. If Dark Forest had a centralized version with its backend hosted on centralized servers, I don’t think it would be a major issue. The integration with blockchain is certainly cool, but it feels more like a neat feature rather than an essential necessity.

ddy: Given that the use of zero-knowledge proofs to shape the fog of war enables some players to use GPUs or even higher-performance hardware to uncover large portions of the map, do you think this creates an unfair advantage for those players?

Elf: I don’t consider it unfair. I see it as one of the core highlights of the game. Dark Forest is a war game, and just as some players excel in social interactions and team-building, utilizing GPUs to reveal the map or writing scripts for automated attacks is part of how players can leverage their advantages within the existing rules. This is a strategic choice for participating in the game. The rules are public, so it’s not that only some players are secretly using GPUs to reveal the map. If others are using them, you can do the same. For us, we can reveal the entire map within the first 1-2 days of each game round by renting dozens of servers. Essentially, we start with a full map view.

ddy: Are you concerned that new players might perceive the game as too difficult and decide not to participate?

Elf: Yes, I do have some concerns. From a purely gaming perspective, it boils down to the difference between paying players and regular players. Paying players can rent GPUs or hire developers to create specific plugins, which is a normal phenomenon and not a core blocking issue. The real barrier is whether new players are interested in researching or participating in the game. Given the current state of the blockchain industry, it’s somewhat pessimistic to say that only projects with significant airdrop potential or returns can attract players. If there’s an expectation of such airdrops, no matter how complex the game is, there will always be a group of people eager to dive in.

ddy: Regarding plugins, Dark Forest’s client has implemented open interfaces that allow for the creation of various functionalities. Some players have even developed large-scale semi-automated expansion plugins using these interfaces. What impact do you think plugins have on the game’s ecosystem?

Elf: Overall, I think plugins are a positive development. Without plugins, the game would be unplayable for many players who might be managing hundreds or thousands of planets. Automation plugins provide significant assistance in this regard. Unless the entire interaction model of the game changes, plugins are likely to remain necessary. On the other hand, with the advancement of AI technology, I wonder if we might see AI similar to AlphaGo emerge in the Dark Forest ecosystem, though this would likely require substantial costs.

Future Visions and Possibilities

ddy: Compared to Web2 or Web2.5 games, what do you think are the prominent advantages and disadvantages of Dark Forest?

Elf: The standout advantage of Dark Forest is its playability. I’m a big fan of strategy games and have played many, but I believe Dark Forest is the best strategy game I’ve ever played. What really impresses me is the role of small planets in combat. When attacking an enemy, directly using cannons to target high-level planets can easily lead to a counterattack. Thus, using small planets to advance is often necessary. I send frontline troops to capture a small part of the enemy’s planets, then hold those planets. This involves a lot of strategic decisions between small planets, which I find very interesting.

The disadvantage of Dark Forest is the poor user guidance and interaction experience. Compared to the detailed tutorials in Web2 games, Dark Forest’s tutorial is quite rudimentary. For instance, while the game mentions that there are many open-source plugins available, it lacks details on how to use them, their effects, potential issues, and troubleshooting steps. New players have limited resources to resolve these issues, as there is little technical material available online. Usually, members of the community help adjust things, but new players who aren’t part of these groups may find it difficult to solve their problems.

ddy: From a player’s perspective, have you considered how to improve Dark Forest?

Elf: I think introducing fatigue mechanics similar to Web2 games could be beneficial. For example, limiting each account to a maximum of 1000 actions per day would make players more mindful of their action points. In this scenario, players might spend 1-2 hours each day to complete their strategic deployments. During conflicts with other players, they would also think more carefully about each action’s impact on the war situation, treating each move with greater consideration.

ddy: Dark Forest has previously not considered any economic aspects. If you were to add an economic layer to a game like Dark Forest, what kind of economic model would be most suitable?

Elf: We’ve considered a few preliminary ideas. The first is to implement a model similar to a paid Web2.5 game, introducing some centralized services and using Dark Forest’s token as a payment medium. For example, moving the fog of war to a centralized server and making planetary information a service. Of course, this involves trade-offs, potentially removing some of the zero-knowledge proof aspects. I believe Dark Forest’s core game mechanics are excellent and are a major draw for players.

The second idea is to tokenize in-game resources. For example, the energy and silver mines previously transmitted between planets could be abstracted as ERC20 tokens. We could design mechanisms for their release and consumption, implement taxes at various liquidity points, or introduce mining mechanisms.

ddy: Could the economic aspects of Dark Forest conflict with its gameplay, and how should they be balanced?

Elf: I think this issue could be informed by the successful experiences of Web2 games. Frankly, Dark Forest isn’t a game where spending money translates directly into significant advantages; it primarily relies on players spending time to engage in intensive operations. In games involving economic models, it’s inevitable that large studios might come in, hiring many players to seek optimal arbitrage solutions. This is a normal situation in both Web2 and Web3 environments, as long as it remains balanced.

Ideally, you would design and create a complex web of interests where different players choose various strategies to compete or collaborate, ensuring that the game remains balanced.