A few specific thoughts, which are not fully formed nor validated, about fully on-chain games and their future.
- Fully on-chain games (FOCG), as a standalone category, are vastly more different from play-to-earn (P2E) games like Axie than P2E games from traditional, web2 games. This is because neither converting all ownable objects into freely tradable NFTs, nor launching a game token representing the ecosystem and participation value, impose real constraints on the game system/logic itself. This is not true for FOCG. Logging all game states and transitions on-chain comes with significant tradeoffs in design space (especially with incomplete information games like Poker), real-time experience, content creation, flexibility, and scalability, just to name a few. The Dark Forest founders summarized it very well in a video. These tradeoffs originate not just from the blockchain’s computational weakness, but also from unsolved structural issues in the blockchain ecosystem, such as that of implementing ZKPs smoothly. I would not be surprised if a traditional game studio sees FOCGs as imposing silly constraints on the game which makes them harder for players to enjoy and harder for developers to earn money. I almost think that, because we’ have re at this time window when folks are just starting to get excited about ZKPs, FOCGs are enjoying much interest from novelty itself rather than how fun the game actually is. In the long run, the major benefit of trustless interoperability may bring FOCGs toward mainstream or even give rise to entirely new categories of games. I simply wish to point out that whether the tradeoffs are worthwhile is by no means certain - we all need to be painfully aware of it as we move forward.
- As technology continues to fuel automation in the near future, people will spend ever more attention in virtual worlds. guiltygyoza raised the notion of games as the basis for human-AI adversarial advancement of civilization, and Max Wolff talked about increasing acceptance of cyborg interaction in games - both very intriguing north stars of the future. There are a couple of things we need to achieve to reach that world. For one, blockchain needs to support much heavier computation at much less cost than today - there are many possible routes which Vitalik elaborates, but it is a must. On top of the enhanced efficiency, privacy computation for AI models should be extremely convenient that it becomes default (or at least among the choices for most researchers). Meanwhile, FOCGs need to have reached a stage where cyborgs can be meaningfully incorporated for content creation and gameplay. Furthermore, there need to be more real and verifiable parallels between the physical world and virtual worlds in a way which does not compromise fun, so that adversarial gains may be translated procedurally to the real world towards some collective defined objectives. (Regarding how those objectives are made, there may be some vital changes in how we as societies organize ourselves, but that might be too far to know.)
- A potential challenge in shifting people’s attachment of value from physical to virtual is the concept of analog. Analog objects are characterized by their (1) novel irregularity, (2) multisensory experience, and (3) embodiment of effort, which people still love owning and attaching special meanings to. Take the example of vinyl vs. digital music. Vinyl records, as physical objects, are inherently multisensory and irregular in their visual and auditory aesthetics compared to digital albums, and acquiring them required much great effort. In some ways, hardware such as VR is solving (2) and blockchain is solving (3), but the solutions are nowhere near perfect as of now and we need to think more.
- One dichotomy in game design which I’m observing lately is that between flexible composability and immersive storytelling. If this was a spectrum, then Minecraft and Roblox falling closer to composability, and MUDs and Legends of Zelda closer to storytelling. The former, often thanks to its multiplayer attribute, feels like reality unfolding, while the latter feels like exploring in a rich, legendary world. The former is much more decentralized than the latter. It is surprising how the two rarely intersect one another; in fact, allowing composable evolution in a game seems to decrease shared common experience among players to some extent. I am curious to what extent this observation is true, and if not, the exceptions which allow both goodies to exist in a virtual world.