Laura Michet's Blog

Avoiding text entirely

I played a very small indie idle game recently, Settlemoon, which has very little text in the UI.

The lack of text honestly made it extremely difficult for me to understand and enjoy parts of this game. I don't want to spend this post slamming it, though, because I gather that it is a work of great enthusiasm and love from a very small team (found reddit posts suggesting it is a solo dev) and because I did, in the end, spend the time to finish it! So it must have been doing something right. I enjoyed myself even though I do not often play this kind of game.

I do encourage you to check out this trailer:

The trailer suggests a plot, ominous happenings, a potential main character, etc. It also shows how the story is communicated, which is through 2d pixel art letters and books superimposed over the game field, sometimes with cipher text on them.

I must admit: I did not solve any ciphers, because cipher-solving was not required to progress. I ended the game with pretty much zero understanding of who the major NPCs were, what their conflict was, or what they were in the act of doing while they were onscreen. There were also some mechanics I did not discover until hours and hours after they would have been useful for me to know about. The game did use precisely enough text (in the form of numbers and short labels) for me to understand the most critical mechanics, like posting quests. But everything else was up to me to discover by interpreting diagrams, or through experimentation.

A lot of game players and even game devs seem to think that forcing players to experiment is, like, often a valorous strategy, or somehow uniquely aligned with the interactivity of games. I do not agree - I think it's a lot more contextual than that.

Sometimes, players may feel that experimentation resonates with the rest of the game experience. Sometimes they may feel that it doesn't! I think this was a game where I could have dug much deeper into the experience I was having if more things had simply been explained to me. I think I would have had a better, more intellectually engaged time if more icons had been replaced by text, if more buttons had been labeled, and if the tutorials had come with text accompanying them.


I do want to point out, however, that avoiding a lot of text in your game is actually a huge advantage depending on your release plans and the number of people on your team. I really do understand why a lot of people do this.

One issue: players are bad at reading. Often, text on its own isn't enough. If you get burned enough times by players not reading, I can see why someone might want to replace that text with a kind of icon pictogram. But communicating with icons is really hard, too, and players are not always any better at interpreting icons. It may also be slower and more expensive to test and revise a communication system based on 2D art assets rather than text. So... it's kind of a wash. I think that the solution is repetition and redundancy - having both, not abandoning text.

Another issue with text: if you want to release in many regions, more text means more localization. Releasing games without proper localization can make some audiences angry, and using machine translation to translate text can produce a shitty result that has the same effect. If you want to release a game in as many places as cheaply as possible with perceived high-quality localization, then you probably want as little text as possible.

Localization also strictly limits how often you can edit the text of your game. Working with real translators means locking your content, sending them work, creating deadlines for them, agreeing on a schedule, fitting into their schedule somehow, and using LQA - localization QA - to make sure that every line was translated correctly. This usually means that you will not be able to change text whenever you want.

I've seen that a lot of developers find this extremely galling - it feels so simple and natural to adjust a line that it can be frustrating to lock your strings for loc and then realize you have to ship an imperfect line which would take you 3 seconds to correct in your own language. It sucks! But it's a requirement, and you have to be at peace with it, somehow.

However......... 2D image assets without text on them do not require this level of discipline and commitment. So if you've got a feature in a lot of flux very late in dev/overlapping with loc, then it can be wise, actually, to limit the amount of text you use to explain it.

Finally - and I have no idea if this is what was affecting Settlemoon or not, but I have strong suspicions - the pixel fidelity of your game might make using a lot of text feel difficult or inconvenient. A deliberately low pixel resolution might make text look ugly to you, might cause text to cover huge regions of the screen, or might require you to use multiple pixel densities on different regions of the screen (and then get your ass handed to you by the people who dislike that look). There are tons of incredible pixel fonts out there, but if you trap yourself with an overall aesthetic look which doesn't leave any room for text to exist gracefully alongside art... then I can see why you'd try to cut it out entirely.


I often have to be a text evangelist. I have to convince people that I will write exactly the correct amount of text to communicate something, and that it won't be overwhelming, and that it won't be boring. But one thing to keep in mind also is that a very large percentage of humans on planet earth find reading text to be extremely unpleasant on, like, a cognitive level.

You will certainly encounter game devs who say things like "everyone hates reading text, so I don't want any of it," or "text always communicates things way worse than art does," or "any time I have to read any text at all it ruins the entire vibe," or something else like this. I try to understand these arguments as coming from a perspective that probably cannot be changed. If someone is telling you that reading text makes their entire brain deeply miserable, then you should probably believe them.

It may be your responsibility, however, to convince them that other people enjoy, need, and want text, particularly explanatory text... and that bringing both art and text to the player at once, redundantly, perhaps also with audio if the budget is high enough... that's actually the safest thing to do. It can be very, very hard to convince creative leadership of this, but it's a fight worth fighting.

I am certain that there are some people who came through Settlemoon understanding who the characters were and what was happening to them, or accurately deducing the various mechanics and how to use them from UI alone, but I didn't. I'm definitely the kind of person who needs more than one method of communication to make things clear for me!

#game_development