Design Philosophy (Part 3: Why Design Badly on Purpose?)

i’ve talked about good and bad design, and one of the things that I’ve wanted to point out is that there’s a separation between good design from a purely design-oriented perspective and something good that is designed badly, whether by accident or on purpose. I think it’s easy to stop thinking about good design at the precise moment something works– once it does what it’s intended to, there’s a temptation to say “okay, good design here” and move on. I think that’s an oversimplification, and a potentially harmful one; it leaves us out of looking to continually improve our designs.


That being said, ideal design is not always possible, or even desirable. There are a wide variety of reasons why we might want to intentionally design something less well than we could, or even design it badly on purpose. I want to talk today about a few reasons we might do this.

In Service To A Greater Design

I’ve talked before about friction, and how important it is to a variety of (notably interactive) experiences. Friction is attained by intentionally slowing things down, making them less efficient, less direct, and in general, more frustrating in small ways. Friction in games tends to take the form of one feature being designed inefficiently to make the overall game’s design better; we make decisions about one area to improve the whole.

Outside of games, there are things like speed limiters or manual transmissions on cars. A car with a speed limiter is not as well designed as one without that kind of hardcoded restraint, but those cars help make the roads safer, contributing to the overall design of the transportation network. Similarly, while cars with manual transmissions are more efficient and give the driver finer control, they’re also more difficult to use and require more training to operate– cars without them, while less efficient and offering less control, allow more drivers to use cars without needing to overcome the training hurdle, and allowing a driver to pay attention to a wider variety of things while driving without compromising safety quite as much.


Catering To A Specific Audience

Sometimes, the specific audience for something lends itself to otherwise bad design. Whether it’s a niche crowd that prefers a very specific sort of content, an enthusiast who is either intentionally pursuing older or less refined hobby materials, or a group of hardcore gamers that wants their games excessively punishing and challenging, catering to these groups will require sacrifices on the design side.

In a lot of these cases, people will just be “used to” something, and even if a superior design exists, the transition or new rules are so uncomfortable that the improvement in design gets lost. Inertia is a powerful force, and it’s often a lot easier to continue using what’s already known than attempt to improve a system, even if that system is poorly designed. The saying usually associated with this is “if it ain’t broke, don’t fix it”, but I tend to think that this misses the need to continually self-improve. That having been said, incremental improvement is a solid way to shift, and that often requires intentionally bad design just to ease the overall transition.


Creating Art

Put simply, there is a line where the science of design meets the art of design, and these two facets of the whole tend to clash. People’s reactions to design are often emotional, rather than practical, and an objectively good design may still inspire negative emotions, and vice-versa. Art is doing this intentionally, evoking emotional responses intentionally, and in the service of creating art, design can and must be sacrificed.

Everything science has to say about aerodynamics may make it clear that curved shapes are more capable of flight, but an artist creating a beautiful painting of a flying machine for a culture that despises curved shapes is going to give up the superior curved design in favor of the desired emotional response.

box kite

Not Enough Resources

Sometimes, something needs to just be “good enough”. It’s a reality of pretty much every creative industry, and realistically the vast majority of designs are going to need to settle, rather than get continually refined to near-perfection. It’s just too resource-intensive to pursue ideal design constantly, so corners often need to get cut (or rounded, depending on the tools you’re using).

While a necessity of staying sane in a creative field, I think it’s still important not to conflate “good enough design” with “good design”– settling for poor design is fine, but convincing oneself that the “good enough design” is actually intended and is therefore above criticism or improvement is something of a dangerous trap. It leads us to stop thinking about how to improve, and I’m firmly of the opinion that we should never stop pursing improvement.


To Teach

AI shouldn’t run into obvious traps. Systems shouldn’t obscure parts of themselves or mislead their users as far as their functionality. Games shouldn’t be wildly unbalanced or turn on cheat codes for the player. All of these are useful in the context of teaching people how things work. The AI that runs into a trap shows you that the trap is there, so it isn’t an unfair surprise. Systems may obscure parts of themselves or lie about their capabilities to break themselves down into more manageable chunks. Early levels in games will often make you invincible, or otherwise have an overwhelming advantage, just so that you can learn the controls and the basic functions.

Inflatable water wings make you a less efficient, less capable swimmer, but they’re GREAT for teaching children to swim and generally be comfortable in the water. They’re a temporary measure to be shed later on, like many teaching tools. Once we’re capable of understanding how to swim, we tend to automatically understand why floaties are unnecessary and really just hold us back. Similarly, once we’re capable of seeing the traps, understanding the systems, or competently playing a new game, we no longer need the teaching tools.

Again, it’s important to understand that the teaching tools aren’t indicative of the good design of whatever is being taught, in the same way that putting scaffolding around a building isn’t a good design for the long term, but is useful for short-term construction until the building no longer needs it. Much like things that are in service to a greater design, teaching tools aid the overall work, rather than necessarily being good design in and of themselves. A game that never sheds its tutorial cheats, or whose AI consistently runs into obvious traps gets savaged in reviews; the game needs to shed those constructs because while they’re useful for the short term, they’re bad for the long term.


Design is a really big concept that’s often treated like something much smaller. Part of my inspiration to write this series stemmed from a discussion I had about free-to-play games, that lean heavily on randomness as a skill equalizer / spoiler and are in some cases explicitly built to draw money out of players at the cost of the game’s overall design quality. Candy Crush is a well-loved game but it frequently supplies you with unwinnable boards and carefully constructed design failures to drive you to make purchases. These things aren’t good design for the game, but they serve the product as a whole.

I find the distinction between “good design” and “justifiably bad design” to be a really important, really interesting one, which is why I’ve spent so many words over the last few days talking about it, both here and elsewhere. At a philosophical level, I don’t think being able to justify something makes it right, and that philosophy extends to things other than moral imperatives for me.

Hopefully this was interesting for some folks. Thanks for reading!

Leave a Reply

Your email address will not be published.