AL #64: My 7 Best Ideas for Web3 Games
Want to build an Autonomous World?
I’ve been asked a number of times if there is a secret behind maintaining a weekly research letter.
When it comes to idea generation especially, there is no fancy system.
A lot of times I don’t know what I’m writing about until a few days before the post goes out.
My only principle is to write about things I’m genuinely curious about or want to explore that week.
That way I can keep going forever.
This week, I’m sharing 7 ideas for web3 games I would love to see built.
I don’t know whether my mind is drifting towards gaming due to a lack of free time or perhaps inspired by reading Eye of the World (first book in the Wheel of Time series).
Either way, let’s jump in.
Web3 gaming – a confluence of narratives
As excited as I am about web3 gaming, the industry has had a confusing start so far.
The contrasts of significant fund sizes but low block space revenues / monetary premiums.
People arguing for fun games vs financially driven games with valuable assets.
The debates between whether web3 games should go mainstream or cater to the narrow but highly engaged group of crypto-natives.
Questions about whether web3 gaming is about the convergence of development and play enabled by smart contract composability or whether it’s about player owned games and assets.
Having these contrasts is part of what makes web3 gaming dynamic and exciting to participate in, so I’m not going to attempt to resolve them.
In fact, I think games with very divergent views can collectively succeed.
In that spirit, here are ten games I would love to see that are very different from trying to just replicate mainstream triple A games.
1. A Warhammer 40k style “Hobby“ Game
Warhammer 40k is one of the highest margin collectible hobbies that seems primed to make the transition into web3.
Players buy official models that implement seasonal rulesets any customize them in unique ways to give their armies a unique character and story.
NFTs are easier to modify in ways that doesn’t damage or deprecate the original asset and the idea of digital “painting” as a valuable past time has already been validated by projects BasePaint, which we covered in a prior issue.
The ecosystem is referred to as a “hobby” and supported by a monthly magazine called the White Dwarf (which I love and subscribe to), dedicated retail stores, events and even a library of books.
The game itself is a direct descendant of wargames (the original inspiration for games like Dungeons & Dragons which in turn inspired all modern RPGs). Arguably implementing the rules in a 3D game would make it a more seamless experience for both beginners and those short on time.
All in all, Warhammer 40k probably charges one of the highest premiums on plastic anywhere (even higher than Lego) and someone should be able to replicate their success in the land of value-added bits.
2. A hardcore simulation game a la Dwarf Fortress
Dwarf Fortress is known in the gaming circles as the “hardest game ever made”. It’s built in the style of “roguelikes”, a type of RPG rendered set in a two-dimensional space where each coordinate is rendered as an ASCII character.
I have a fondness for roguelikes, having played a few of them a long time ago and doing a fair amount of game programming in ASCII space (my first major program was a game of snake that I wrote entirely in Pascal on DOS).
However, the game is far deeper than a roguelike. Instead of controlling one character you control a simulated colony of dwarves. Everything in the game is procedurally generated but unlike Minecraft this even extends to items, NPC descriptions and personality and everything in between.
The game appeals to a certain kind of player and my hypothesis is that that kind of player would also be happy to tolerate the additional friction that is (currently) required to play web3 games.
The simplicity of a two-dimensional ASCII space and procedural generation also naturally lends itself to blockchains where storage comes at a premium.
Newer technologies like ZKML, trusted execution environments and coprocessors could be used here to favorably compress a large amount of off chain compute.
3. A geospatial game
I was at the Google campus when the internal demo of Ingress went live. I was running around with my Nexus tablet blown away by the immersive nature of it.
Years later, I did the same with Pokémon Go, walking around London in the hope of finding new Pokémon. Once we even went on a long walk in the middle of the night (1am) to hatch some eggs.
I do think web3 has the potential of adding something unique to geospatial games especially powered by DePIN.
For example, one of the issues facing Pokémon GO was players who faked their GPS location and built automated farms to acquire resources. With real hardware owned by the game or a partner, we could have verifiable location.
Another experiment could introduce composable real estate primitives and overlaying them onto real coordinates.
4. An autonomous farming game
I still find Stardew Valley one of the most inspiring stories in gaming, being built by a single developer over the span of 4.5 years.
It's a true Indy game with breakout levels of success.
And it should inspire web3 developers to experiment with the farming genre which seems highly suitable to crypto where it could be merged into a single autonomous world that others could build upon.
Imagine building a house where you can hang up other NFTs in your wallet as posters or leverage items and decorations from other games. Everyone’s Minecraft and Animal Crossing home base is special, it would be even more special if it truly felt like a digital home.
Account abstraction is already expanding the notion of a wallet, why not push it further and have a game become a wallet and aggregator too.
5. Mempool impact game
We are seeing a real evolution in the sophistication of in-game agents powered by adoption of LLMs in engines like UE5:
However, in crypto we still think of AIs as living “onchain” with their behavior described in smart contracts.
Allowing in-game agents and player owned bots to interact directly with the mempool would add interesting MEV-like dynamics to games and represents a completely new axis how to think about games.
Of course, this is only viable for games that exist on their own L3s but many games are going this route anyways so it’s not a significant restriction.
Another option is to introduce agent transactions programmatically in the sequencer which would permit the use of LLMs as a block-building hook.
This would be inspired by Worldchain’s idea of prioritized blockspace which we covered in a dedicated issue.
6. Time manipulation
Another idea that opens up when we consider how blockchain mechanics can be part of the “meta” is the idea of using block reorgs as time travel.
Especially in multiplayer games, the chain could put an explicit (and very high) cost of doing a block reorg which would revert prior player transactions and allow for novel mechanics to develop.
Similarly, slowing down time inspired by the Sandevistan in Cyberpunk would be possible through the manipulation of block times.
7. Custom execution environment
Another way that each game’s rollup could be modified is through a custom VM.
Notch’s 0x10c game prototype experimented with this concept of in-game programming but with blockchains programming doesn't need to happen in-game at all to be considered “part of it”.
Imagine a game which is built from the ground up using a custom VM that is also available to its players which fight over their ability to modify the world.
The result would be highly unpredictable and exciting, with players playing the role of citizen, developer and God all in one.
This was a really fun post to write and several of the ideas have been stuck in my head for a while.
I'd love to hear more from you if you've already seen these concepts in development or indeed want to turn one of them into a hackathon project or something even bigger.