2026-01-30
A short essay about the curse of abstracted thinking in game design and in product design.
The familiar space for the game design sits up top. Our mind is a great reasoner. We can put ourselves into believable situations, or rather, we can believe we have put ourselves in situations.
Many of my game design discussions start as a conversation. Conversation is the heart of a design process. This is what a studio or office space or garage provides: a landscape for the pulses to work their way. Extending the maximum radius of this system has been the work of the last three decades – the internet still falls short.
A meeting contains conversations. A video call sublimates conversation. Therefore a meeting can be done on a video call. Why does it feel wrong. This is where the reigns of reasoning rise up and take charge and we could spend a few paragraphs conceiving of the limitations of video call, contrasting video call to real conversations, and concluding with a way forward.
The problem is our abstraction.
The problem with game design through conversation (and any design for that matter), is the same issue of conversation by way of wires, is the same issue as solving problems through thought.
An example from the early days of our development of GUG. One year into development, the game felt incomplete. We needed to explore new game designs. We established a one-month 'exploration' period and agreed on a three axes we would implement. These would be features that we expanded the game with. At the end of the month we would test a prototype of these and evaluate them according to some metrics we decided in advance.
After one month (and a little bit of overtime into the next month), none of the explorations we made were integrated into the game. At the time, our retrospective revolved mostly around a time budget problem and a production problem. Given more time, some of these ideas could have matured.
But let’s consider abstraction: what were we actually talking about when we discussed these new game ideas. One of the core ideas was a world map. The game, for context, was a turn-based battler. The player moved between battles via a simplified map or a forced choice (think Inscryption or any classic JRPG or Slay The Spire). Our vision of the map was this further space where players could choose their direction, perhaps in a tile-based manner, and discover new battles in this space in a non-linear way. We settled on a restricted implementation of the map, a single path with one or two paths that allowed the player some choice. The experience felt flat, stilted, and entirely unnecessary.
I reflect on what our conversation about this map was, and it is clear that our collective mind model was stuck in a space that was not real. Real in the sense that the time, budget, technical advancements, and experience made it impossible. This is a curse of the abstracting mind. The craft of game design is to create this minimal illusion that relieves the larger fantasy built inside our brains. But what is that fantasy?
That fantasy is much more fluid. It flows between states, it expands outward, we are rushed towards new worlds, expansive lands. Where there should be holes are instead diligently filled spaces. Say to yourself now "expansive mountain planes" and watch behind your eyes as you pretend where you are. Without careful scrutiny you enter this space where anything is possible without testing it. In other words, you cannot perform science inside your mind. The ability to be objective requires careful energy. Walk your mind towards a flower in the field, bend down towards it, pluck it, inspect the stem, the texture – and the moment we stop this thought, the mind fills in the future. It makes up the rest of the interaction with a large and satisfying ellipses.
Ellipses are a curse of our limited mind...
We have to compact and compress in order to function. We could not deal with all that is up above if it was as precise as the way it worked in the real world. They help make communication efficient and allow us to stretch information between domains. When it comes to answering our questions honestly, our mind can only answer abstract questions with abstract answers.
What do we do? Anjan, the CEO of Daylight Computers has a quote that I like. Daylight Computers is working on a new hardware platform for a positive-tech future and they have released monochrome tablets for adults and kids to engage in technology in healthy ways. Anjan is quoted as telling his R&D team to follow the mantra: "See it. Buy it. Try it." To expand on Anjan's aphorism, I would rephrase it as: test your mind’s abstract design with the real world evidence. Anjan is referring to designs of old devices in this context. When a designers learns about an old 1970s computer hardware, it is tempting to look up images and learn about it online, build a model in their brain, and then consider the object understood. It is concerning how simple and satisfying it is to do this. The brevity of Anjan's phrasing is key too, because it is about using reality to challenge your design as soon as you can. The longer a design sits unchallenged in the fluid of thought, the longer it is left as an unchallenged reality, the further the reach of those nasty ellipses make it into its actualization.
Some call this a vision. The dream that requires steps to attain. But the point of a dream is never to be attained, so they say. But perhaps it is that attainment of dreams is never possible because they are not real.
What of the map feature we failed to implement into GUG? My critique turns towards the vision and not the execution. What were we dreaming? It is hard to remember all the details now, but I believe the collective vision of what that map provided was similar to what crosses my mind now: an expansive, open world with choices, freedom, discovery, exploration, unbounded play, creatures, design – and so on. As designers we have processes to fight these abstractions. Targeted questions that slice the bubble, cut it down until we find its smallest seed. Questions such as: what choices? How do they explore? What controls? What happens when button X is pressed in scenario Y? What creatures? How many? Which set of mechanics in which situation? How big is the world in miles? In meters? How frequent are encounters in miles? In meters? What are the maximum number of possible encounters given the number of enemies selected? Are enemies selected uniformly randomly? Cyclically? How do we measure replayability?
Answering these questions is a great start. Getting them out and on the board for the team to scrutinize helps subdivide the space, shave off the edges, and find the core of your minimal prototype. But there is still a missing step required after this. After having answered all these questions and figured out the technical, design, and conceptual limitations, one should return to the vision at hand. What was the point of the vision initially? Do the answers to the questions now satisfy the vision? This is the hardest part because the sum of the model will always be a subset of the vision. How can the vision even survive?
Reality is here to save us. We can use another person’s mind as an evaluation device. The process is simple: given a system that satisfies the constraints we have defined, what resulting mind model does another player experience? In other words, does your design produce the required vision in the mind of someone else. In that sense, your game has successfully communicated the abstraction to the player. This applies beyond games.
One similar, but wrong way to do this is through marketing. It is possible to market your vision quite easily. The tools of the marketer allows direct abstract-to-abstract delivery. This is great for customer acquisition. This solves nothing for customer satisfaction. This is why there are so many products that seem to exist that solve abstract problems. All those ellipses in our minds being tickled by all the products of other ellipses in the other minds. This is why scrolling through Kickstarter products is an ecstasy. All those dreams sold to our own dreams. Its expansive thinking feeding expansive thinking. In abstract land it is frighteningly easy to do everything with nothing. One image: we are traveling space. One phrase: we are rethinking our life choices. One statistic: we are considering a healthy lifestyle.
Yet we suffer from all the problems that these futures claim to solve. Take modularity as an example. I believe the abstract model we have in our head, this self-marketed vision of composable, indefinite parts, is so enticing in the group firmament that it is continuously reinvented every few years. There is something in our model we are believing is true of this system which the reality misses. It might be economic, scale-related, or technological – or, our visions override the complexity and confluence this system: what software do we build for this? Why do I want to take the effort to move these pieces around? Does the repositioning of pieces make me Batman or Mr. Fallout, or does it just make me have a heavy backpack?
If we want to feel like Batman, we need to test for Batman. We will discover that modularity does not get us there.
To close this off, we can define a science of design practice:
Abstract models inspire us. These abstractions are not real. Real models test our abstractions. A successful real model is one that realizes the abstraction in another mind. Therefore, the best real model is the one that realizes the closest target abstraction in another mind.
To do this, you have to make physically. Conversation is good, but you must build, test, and measure. What you measure, how you build, and what is good, is part of the design of your product. If you aren't answering these questions you are confabulating or having fun. It's fine to have fun, but you have to be a principled scientist. At best, this approach to building stops you from falling victim to other’s dreams.
Knowing that you are being sold abstractions is a great way to scrutinize a product. See it. Buy it. Use it. Take note of your abstractions. How far are they from what you were sold?