Baba is You preview: a puzzle game that pushing boundaries

Genre: Puzzler
Format: PC/Switch
Developer: Hempuli Ot
Publisher: Hempuli Oy
Release: 13 March 2019

Block-pushing games have been around since the days of Sokoban in the early 1980s, yet Finnish developer Arvi Teikari has managed to come up with an ingenious new take on an established format. Like Sokoban, Baba is You involves pushing blocks to solve puzzles; the twist here is that pushing blocks will fundamentally alter the game’s rules. Blocks can take the form of objects or individual words, so three blocks in a row might read, say, ‘Wall is stop’. Pushing one of the blocks, thus breaking up the sentence, will disable the collision detection on all the walls on the screen.

This is but one of numerous mind-boggling ways you can manipulate the game’s properties by pushing blocks and creating new sentences. The title itself is a reference to one of the most intriguing concepts in the game: replacing the word ‘Baba’ block in the sentence ‘Baba is you’ with a different block or object will switch the player’s control to that object. If the sentence is changed to read Key is You’, for example, the player will then be able to take control of the key, move it across the screen, and unlock a door.

With Baba is You filled with ingenious ideas like these, we just had to get in touch with Teikari to find out more. How did he came up with the concept? How did he playtest the game to ensure that, through all that rule-changing, players couldn’t break it entirely? And who is Baba, anyway? Just ahead of Baba is You’s release on the 13 March, here’s Arvi himself to give us the lowdown.

Block-pushing games have been around for decades, but the concept behind Baba is You seems completely new. I know you came up with the idea at a game jam, so which took longer – coming up with the concept, or implementing it?

Definitely the latter; Baba’s concept was very much one of those happy accidents where several ideas just happened to line up at a fortunate time. I think I got the idea during the night after the the game jam’s theme was announced, and spent the rest of the 48 hours we were given implementing it. Although of course I semi-passively tinkered with the concept in my head while coding etc.


Baba is You’s visual style’s simple yet appealing.

Was there a ‘eureka’ moment where you came up with the rule-pushing concept?

As far as I can recall the moment wasn’t very eureka-ish. The theme of the jam, ‘Not There’, made me think of logic operators and how in logic you can reverse a statement with a Not operator, and that eventually developed into a mental image of blocks of ice melting near lava unless the player declares that “Ice Is Not Melt”. I was fairly certain that the result would either not work in the timeframe I had or wouldn’t be fun to play, but decided to test it anyway, and luckily that turned out to be a good idea.

Can you describe some of the tools you’ve used to create the game – what have you found best for drawing and animating the sprites, for example?

I’ve used a game-creation tool called Multimedia Fusion 2, made by Clickteam, for nearly all my games so far. It offers a simple-but-fairly-powerful graphical interface and also a limited-but-usable sprite editor. The base program doesn’t allow scripting of any kind, but luckily for me a good friend, Lukas Meller, had shown me a plug-in a year or so prior that allowed integrating lua to the program. As a result most of Baba’s engine is scripted in lua.

When I’m not using the tools included in Multimedia Fusion 2 itself, I generally use a combination of Aseprite, Krita, occasionally GraphicsGale and a couple of older drawing programs for art, Notepad++ for scripting and Audacity & OpenMPT for audio. It feels like most non-subscription-based tools do specific things very well but lack some other specific features, so combining multiple tools has felt necessary.


Baba Is You contains over 200 puzzles, according to Teikari.

What was your pathway into making games? Do you consider yourself an artist first and a game maker second, or vice versa?

I don’t remember when I began to want to make games, but I must’ve been still in kindergarten. My cousin’s Super Nintendo and various PC games with level editors (Jetpack) and the The Incredible Machine series come to mind) probably contributed a lot to this fascination.

In primary school a schoolmate asked if I’d want to make games and showed Game Maker to me. GM’s scripting language was a bit much for my then-unable-to-understand- English-well self, so another schoolmate introducing me to the non-scripting-based The Games Factory, a predecessor of Multimedia Fusion, was a huge improvement. During that time there was a shortlived boom of young hobbyist game developers using tools like these, at least in Finland, and I guess the beginning of my gamedev hobby was somewhat a part or a result of that.

As for how I define myself, I guess I would call myself a game developer first, although I don’t really appreciate distinguishing that from being an artist. In case it helps, I paint a bunch and work as a pixel artist in another company so that accounts for some visual artistry.


“Some levels have gone through many many iterations of fixing and testing,”Teikari says.

I really like the game’s look. Did you try alternative styles before settling on this one?

The game started off with a very simplistic style due to the 48-hour time limit of the game jam it was made in; there are some stylistic shortcuts I tend to take when I need to hurry, and as a result the first version of Baba (still available via itch.io) uses a black background and default Windows palette. Once I started making the full version, I momentarily worked using the sprites of the game jam with updated colours, then made a couple mockups of what the game could look like and ultimately fairly quickly settled on the current wobbly style (looking at some pics, settling on the final style took about 1½ months). I’ve used a wobbly style like this in multiple previous games, so it was a relatively easy pick for me.

I’ve been consistently worried that the simplistic/pixelated look will drive people away, and as such it has been very encouraging to hear that people like the looks of the game. Thank you!

Aside from the music and visuals, have the mechanics and scope of the game changed much between the early version and the new one?

The scope of the game has inflated very heavily over time. At the beginning of development I added a lot of words that I felt could be interesting, and while I’ve retired a couple (Slip and Grab) for various reasons, most of the words have been left in. An older notebook scribble tells me that I initially estimated for there to be some 60 or so levels in total, if even that, which is amusing considering that I ended up with 200+ of them. I’ve been afraid that the game might become too exhausting to play, but at the same time the inflating scope has allowed me to really think through the systems of the game and figure out certain mechanics that really explore the concept deeply – the game would have been a much narrower experience had I gone with early concepts only. Of course other people have given lots of ideas and inspiration as well.

Is there a backstory to the game? Who and what is Baba?

There is a backstory of sorts and the game features a couple cutscenes (only in the form of intro- and ending-type events, though); however, the narrative of the game is very much more concerned with the rule system and the structure of the game world than with Baba or
the other beings seen in the game. As far as the game’s concerned, Baba is a cutesy sheep/cat/goat/dog who just happened to be settled with ‘Youness’ or a consciousness of sorts. In the jam version Baba was intended to be a robot with antennae, but a colleague, Jason Boyer, made some really sweet fanart of Baba, interpreting them as a humanoid goat creature (with the antennae becoming horns), and I decided to change Baba’s looks more to that direction as I updated the sprites.

Could Baba is You have been a horror game, do you think? It’s a game about warping the rules of reality, after all, which could be quite scary in a different context.

I don’t think I would’ve been able to turn Baba into a horror game; there would’ve been many pitfalls of accidental cheesiness and I think I would’ve fallen into them had I thought to try. I suppose even as the game is now there are some creepy elements in it, though, at least for some players, considering how meddling with the rules can alter and remove consciousnesses of beings.


Devising Baba is You’s puzzles has taken many hours of careful planning and testing.

Do you think Baba is You will help people better understand how games work? The way you push around rules in the game reminds me a little of Scratch, where you can plug scripts together in blocks.

It’s very hard to say at this point. The syntax Baba offers is quite unlike that of actual programming, so in that sense it might not be overly helpful, but on the other hand Baba could offer someone the spark to start studying actual programming languages in case they find the rule-shuffling similarly exciting.

What’s been the most challenging aspect of developing the game so far?

Depends a bit on how we define “developing”; marketing in today’s environment is exceedingly tough and scary, for example. But if we deal with purely game design -related issues, I’d say that two things come to mind first and foremost: composing of certain songs and writing code for the cutscenes. The former is purely an inspiration-related issue – some tunes in the game took me a long time to come up with, which made the process quite stressful. Somehow a musical theme for a forest area just didn’t show up until after waiting for it half a year or so. The latter is partially inspiration and partially coding-related issue; Baba’s basic engine is fairly rigid in order to accommodate all the rule-changing shenanigans, so implementing a more flexible system for cutscenes was very tough. Partially as a result of that figuring out the contents of the cutscenes was also difficult.

What’s your process for testing the game? Is it time-consuming to come up with the puzzles, and make sure they’re solvable?

Baba’s basic concept offers so much food for puzzle ideas that I mostly haven’t had to struggle to come up with them, with some notable exceptions. I’ve tried to follow a principle of having every level have as unique an interaction/gimmick as possible, both to keep the amount of levels down and to explore the concept more deeply. Looking at the level folder, it’d seem that there have been around 70-80 levels that have somehow not made it into the final game, so some ideas have definitely been stinkers. Generally the most common issue
hasn’t been that a puzzle wasn’t solvable, though, but rather that it had excessive unintended solutions (or that it was just frustrating/boring.)

As for testing, I’ve been lucky to have a bunch of friends and colleagues and helpful people who’ve kindly offered feedback throughout development. It’s easy to test that a level can be solved at all by myself, but said unintended solutions are way harder to spot and as a result some levels have gone through many many iterations of fixing and testing. Closer to release I’ve tried to approach testing in a bit more organized manner and that has been very helpful – watching a video of someone playing allows spotting oddities very easily.

Similarly, do you have to constantly check to make sure the player can’t break the game altogether?

As mentioned in the answer to the previous question, this has definitely been the more common issue with the game’s levels. In many cases I’ve gotten a nice alternative variant out of an unintended solution, though, in cases where a tester finds a way to solve a level that’s not what I had in mind but novel in a different way.


A tricky game to describe, Baba is You’s rules become obvious after a few minutes’ play.

The game’s concept might be difficult to describe verbally; do you think the advent of YouTube and streaming have made unusual game concepts easier to show to people?

Depends very much on the nature of the unusualness. Slow-paced, turn-based, difficult puzzle games are some of the hardest games to stream/make videos of in terms of garnering an audience, I feel, because the optimal way of playing them is usually very slow and non-flashy, and the satisfaction of solving a level might not be conveyed very well to an audience. I’d say that media formats like gifs (and gif-equivalents) have been very useful in sharing unusual concepts, though, since those allow cutting off the buildup and thinking and just displaying whatever the unusual element actually is. This isn’t to say that streaming/videos wasn’t helpful and useful, even to a game like Baba Is You.

I’m super looking forward to seeing someone stream Baba; I very much want to see their reaction to some of the twists of the game! Streams have been a very neat source of personal encouragement for me.

Have you thought about how the game could be localised into different languages, given that language is quite a big part of the game’s rules?

This has been given a lot of thought, and at this point it seems clear that the game is pretty much impossible to localise entirely to most languages without actually changing the puzzles. At this point we’re planning to eventually translate the UI, but unless we figure out something the actual game elements will have to remain as they are. To give an example, Russian doesn’t have a present tense “be” verb in the same way English does, so just the sentence “Baba Is You” isn’t translatable while keeping the word count and not resorting to more outlandish options.

What’s it been like when showing Baba is You to players at expos and festivals? Are you surprised at how quickly they pick the rules up? Or are there times when you’re surprised at how hard they find it?

I enjoy talking and presenting my stuff so my experiences at events have been mainly positive. I’ve definitely had both kinds of experiences – I think my general expectation has been that the game is hard to “get”, and in many cases I’ve been surprised at how quickly
people grasp the idea, while in others I’ve noted that people have intuitions about the game’s mechanics that differ enough from my own that I’ve wanted to clarify some of the implications of the game mechanics.

How much of a boost is it, as an indie developer, to have the support and interest of Nintendo?

In more business-y terms that is very hard to gauge, but there’s a lot of excitement about the Switch so I’d definitely assume that it has given a boost to the attention the game has gotten. On a personal level the impact has been very real, in any case – I’ve been a fan of Nintendo games my whole life so getting to release a game on a Nintendo console has been very cool. On top of that, I had an extremely sweet experience showcasing Baba at Nintendo’s Nindies booth during Gamescom last year, the crew at the booth was very kind and encouraging.

Baba is You releases on Switch and Windows on the 13 March.

Leave a Reply

Your email address will not be published. Required fields are marked *

More like this