In the spring of 2021 we started talking about how blockchains could be used to make a new kind of game. Between us we’d built more than 100 games in our careers already, but we were taken with the idea that this web3 thing offered a genuinely new design space.
Early on we took the view that we should focus on games that wouldn’t be possible without decentralised technologies. Designing new kinds of game experiences is a harder task than revising existing ones, but then as now, we feel it leads to a more interesting destination.
Shortly after signing an office lease and making some great hires, the Loot NFT was released. An onchain game asset generated from a smart contract, designed to be used in game projects. It challenged the idea of what a game asset should be, and felt like a good idea for us to cut our web3 teeth on.
We built The Crypt, a fully onchain game running directly on Ethereum, whose gameplay involved collaborating with other holders of Loot to defeat onchain dungeons. It was a co-op game that made use of a new feature for videogames: interoperability.
The tweet announcing the Loot project
A screenshot of our first game, The Crypt
We learned how to write a smart contract in solidity, how to optimise it, how to design a game with a two-minute transaction time and how to handle community-owned IP. A whole set of things that you don’t have to consider when making a PlayStation game.
The result proved really popular. All dungeons were defeated within 15 minutes, so we did what any self-respecting game developer should do after a successful game launch - we went to the pub.
A group of game developers in a pub in the middle of winter, feeling very happy with themselves!
We continued developing chapters of The Crypt, trying out new ideas like ‘gas-less’ transactions (using a Polygon relay) and dungeons that required a combination of Loot-derived NFTs to succeed.
A surprise in Chapter 3 of The Crypt was when an unknown developer extended our onchain codebase to find optimal raiding strategies. The idea of modding is not new to games, but seeing your core game experience evolve permissionlessly was. The composablilty of onchain games became a key area of interest to us in subsequent games.
After a seed investment round led by Bitkraft, and a mixture of game- and web3- focused funds, we started thinking about our next game. We wanted to extend the world of The Crypt, but needed to work with the constraints inherent in blockchain technology. So real-time combat was out, but a slower-paced game in a persistent shared world seemed like it could work. Let’s make an MMO! How hard can that be?!
An early screenshot of Dawnseekers.
We moved blockchain from Ethereum to Starknet and created some early prototypes, which focussed on exploring an onchain map. We also met a whole bunch of people working on similar projects, brought together by the idea of Autonomous Worlds. Along the way we built infrastructure to compensate for the shortcomings of using a blockchain as a game server, including client-side proofs and account abstraction.
By Christmas 2022 we recognised that the scope of Dawnseekers was getting too large, and that our key feature of composability was getting buried. We took a step backwards to go forwards, stripping back the game and bringing composability front and centre. Downstream shares DNA with Dawnseekers but is much clearer in its mission.
Composability abound in Downstream!
Games like Dark Forest and OPCraft included composability, where the community had expanded the functionality of the game. We wanted to take that idea further by building composability into the game design itself, so that the game’s features and strategies would evolve as the community built new parts of the game. Furthermore, we wanted all players to see these new parts with a minimum of headache, so we built technology that presented them to all players as they were deployed.