An In-Depth Interview with Jonas Kaerlev, Creator and Lead Developer of ‘A Hat in Time’

Disclaimer: This interview has been edited for clarity and brevity.

A Hat in Time is the first game released by indie studio, Gears for Breakfast, but the collective background of the development team is impressive. Comprised of talented artists, programmers, ex-modders, and the like, the team represents a globe-spanning endeavor more than five years in the making.

After a wildly successful Kickstarter campaign in the summer of 2013, Gears for Breakfast set out to create a game inspired by classic 3D platforming collect-a-thons. The game finally released on October 5th, 2017 to resounding acclaim and applause.

In this interview, Jonas Kaerlev, the founder of GfB and lead developer on A Hat in Time, shares his thoughts on game design, indie development, and creating “fun.”

I was looking at some previous interviews you had done. There was one dating back to 2012 where you showed a screenshot of the game with Hat Kid, the visuals looking closer to Wind Waker than it does now. You’ve clearly had the idea for quite some time; so when did Hat Kid become ‘Hat Kid’? 

Jonas Kaerlev: Well, with the Wind Waker cel-shaded style, people tend to just call cel-shaded style Wind Waker which is fair. The reason we chose that style early on is because we still didn’t have a full grasp of modeling and normals, so it was a lot easier to hide up those mistakes that would otherwise be visible in normal shading; we pretty much just used it that way to hide our mistakes. Once we got a lot more comfortable with our skills and we felt like we were able to do more, then that’s when we started looking into alternative rendering methods, including shading, and that’s how Hat Kid has her current shading.

The design for Hat Kid is super old. She pretty much just started out as a block of characters: she had a sphere for a head, a block for a body, and a top hat. When the game exited prototype, we took our blocky model and said, “Okay, well how can we make this an interesting character?” Then one of our artists, Luigi Lucarelli, he just changed her up completely and made her such a lovable character. And that’s the Hat Kid we’ve got today!

It’s interesting to see the workarounds you used to get the same kind of quality and effects of bigger, older, AAA game studios. There are very clear lines of inspiration between A Hat in Time and older games like Super Mario Sunshine and Banjo-Kazooie. What are some inspirations that may not be as obvious?

JK: One of the more obvious not-so-obvious ones is Psychonauts, which released in 2004, I think, something along those lines. It’s a really good game, it’s one of Double Fine’s best games and it’s super funny. They’re coming out with a sequel too, which is cool, AND we got an official mod for Hat in Time! Because A Hat in Time has modding support and we wanted to kickstart it, get some really cool mods going, we released a few official mods and one of them is the mod for Psychonauts (with Double Fine’s blessing of course).

But I think some more less obvious games are probably the Paper Mario series. I think, once you get to play “Murder on the Owl Express,” which is one of our levels, you’ll definitely start to realize “Oh wait a minute, this is Paper Mario,” especially in writing and comedy style.

Another one is Left 4 Dead. Left 4 Dead is a first-person shooter where you mow down zombies, and you might think “Wait, it doesn’t have a lot to do with A Hat in Time.” But actually, if you look at the mission “The Big Parade”, it’s very much structured like a mission in Left 4 Dead, in that there are these moments where you have to survive, and then there comes the “Tank moment” and you have to do something.

So, you know, just because you’re making a cute platformer doesn’t mean you can’t look at games where you’re slaughtering zombies. You can take inspiration from anywhere, because a lot of mechanics are able to be reworked in an interesting way to fit your mechanics and what you’re trying to do.

Where did you get your start making games? When did the transition happen from hobbyist to indie dev? 

JK: I started out way back when I was just a wee lad. I used to make fan-games, specifically Nintendo fan-games, like crappy Mario platformers just for fun. I loved Nintendo games back then, and it was a neat way for me to do something, to get to play a game I could never play before and have my creative input be directly inside of a Nintendo-ish game.

It’s why I think fan-gaming is really cool and I want to support it as much as possible. Because you never know, you might look at a crappy fan-game and go “Aw, that’s just a crappy fan-game that guy’s never gonna grow up to be anything,” but a lot of people that work in the industry today have that background.

After I did fan-games I went over to do modding, and I used to mod Counter-Strike, Team Fortress, and Half-Life. They had a low barrier of access because there was an entire modding community around it, and it all started with me just wanting to make something fun for my friends. We used to love Left 4 Dead and I’d heard something about there being hidden modifications in the game. You could get a spotlight or something for your weapon, right, and that was never released in the game but apparently, it was still in the code. I thought, “Oh man, that’s super neat. What if I made it so that it would only spawn those power-ups as you went through the level?” Just so I could enjoy a game more with my friends. I thought it was so cool that I could make a game I enjoyed so much better.

Modding is such a cool way of getting started with making games, because in order to release a mod you need to have a pretty good understanding of how a modern AAA game engine works. If you know how a modern AAA game engine works, you’re probably gonna be a pretty good developer. A lot of people on A Hat in Time are ex-modders for that reason.

So yeah, if anyone out there is interested in making games you don’t have to start with your first game being an indie game. You can start just making shitty fan-games and eventually you’ll work up your skills to the point where you’re able to make a really good indie game.

Although you do have experience in three-dimensional space, working with TF2 and Counter-Strike doesn’t exactly strike me as “3D platforming.” What challenges came from moving 2D platforming mechanics to 3D?

JK: Way back when Unreal Engine 3 had just released, I was still doing mods in Source at this point (Source being the development engine for TF2, Counter-Strike, and such). I remember Source being a pain in the ass. You couldn’t just jump into a game and test your stuff; you had to boot up everything, it took forever, and it was so annoying. Super un-fun environment to work in.

But I had been eyeing Unreal Engine 3. They’d released a public version of it called Unreal Development Kit (UDK). I tried it out for the first time and I was like “Wow, this is amazing!” Graphically, it was leaps ahead of anything I’d seen in Source before. It just blew my mind. So, that just made me really want to make something, anything in Unreal Engine 3. One summer I sat down and said “Alright, I’ll just make, I don’t care, I’ll make whatever in Unreal Engine 3.” The only requirement I set for myself was that I had to finish it, it had to be done.

So I just started messing around. I put a small box guy into the game and he could jump around. He had a weapon as well, he had a sword and he could swing that sword to kill enemies. Because I wasn’t too good at programming back then, the sword-fighting was really boring compared to the platforming. It was a lot more fun just jumping around, and that’s when I made the connection “Well my combat sucks; let me just focus on the platforming part.” Slowly from there, it went on and on and became better and better until it eventually became A Hat in Time.

As I played through the game, one of the things I noticed was that a large part of the combat mechanics or the way you interact with the enemies is tied to your movement.  

JK: Yeah, we have a Sonic-ish homing attack. That came after our Alpha release where we were like, “Okay, nobody can hit any of these enemies on the head on their own, so let us just make it super easy for them.” It’s a bit more difficult in a platformer where you have full 360 camera control. It’s way harder to jump on an enemy than it is in like, Mario where it’s a fixed camera angle, which is partly why we decided to go with the homing attack thing.

Very early on we found out that people, when they’re playing A Hat in Time and they see an enemy, they just, they charge to the enemy. They don’t care. If they take damage, they don’t get it, right? We had a lot of enemies where they were like, “Alright, now they’re invincible and now you can’t attack them,” and people just didn’t understand it at all. I think, if you dig through the files, there’s like four fully programmed cut enemies in the game just because those did not work with what players wanted.

So instead, we made the enemies around the fact that the player literally just cannot hit them until we allow them to. Stuff like the Ninja Cats in Alpine Skyline are annoying, but they’re a pretty good illustration of what we wanted in enemies. Basically, they’re entirely invincible and even invisible: you can’t hit them at all until they allow you to. If we made the cats just be invincible, but not invisible, then players would just charge to them and not really understand why they couldn’t hit them. Even the Sleepy Raccoons, they drop from the sky when they wake up. Again, they’re in the air for a reason: it’s because otherwise, players will just try to hit them in various ways. That’s how our enemy design ended up.

I think the Ninja Cats are my favorite enemy in the game, just because we start out very easy. The Cat can steal one of your pons, which is your currency, and so players get pissed off and are like, “Alright, gimmie back my currency!” Then, later on, we turn it to MAX and we say “Okay, the Cat now steals your hat.” Players feel so helpless and so insulted that this Cat took their hat. And that’s super funny to me.

On this subject of cut enemies, what were other things (mechanics, features, etc.) that were cool in concept but bad in execution? What about the other way around? What didn’t initially work out that found its way into the final build?

JK: Early on in the game, I think it was around the Alpha build, if you attacked in the air Hat Kid would do a spin-attack. That sounds cool in theory because it means you can hit enemies in the air, but in reality, people just spam it and don’t understand anything. Not only that, but it made enemies in the air vulnerable again, so we dropped that pretty quickly. The community had gotten pretty attached to it, so some people, especially on our development team, were like “You can’t take that feature away; it’s too good!” But once they got accommodated to the fact that it was gone, it was a much better experience.

Another thing: in the Alpha build and the Beta build, hats were pretty much just cosmetic, right? But in the final game, hats are actually the main, active ability, while you’ve got badges that are passive abilities. We were kinda worried that wasn’t gonna turn out great, especially because we’d swapped around the entire mechanic, but in general, it turned out amazing! We were really happy how that ended up. We did that, I think, less than five months ago. It was very last-minute.

Really! The hats were that last-minute?

JK: Before, the hats were still there but [mechanically] they were badges instead. So the badges that are passive in the game now, they used to be active. They were swapped around, basically. When we made that change last-minute, we were like “Ah, we really hope this is gonna turn out great.” And it did! It turned out really fun.

Also, the Monday before we released we added a button that allows you to blow kisses and a button that allows you to tease enemies. People have been loving the hell out of that feature! Like, every screenshot we get on Twitter has that feature and it’s been so beloved, and the only reason we added that feature was because we played Yo! Noid 2, and Yo! Noid 2 has a “dab” button. That button made us realize it’s super fun to make your character do wacky stuff. We sort of took that idea and made Hat Kid blow kisses and tease enemies instead. We’re really happy with how the reception of that feature turned out to be.

There are a lot of easter eggs and little things to play around with in Hat in Time. “Play” and “Fun” are vague concepts, but the game manages to capture them wonderfully. What was your general philosophy in developing “fun”?

JK: Even just a year ago, we had a lot of missions that weren’t really fun. I think, specifically, Dead Bird Studio comes to mind as not being fun (Dead Bird Studio now is super fun), but that was before we added the “Billing System.” When you enter Dead Bird Studio you get these different bills for being mischievous, right? You get a bill for “Trespassing,” it says like you need to pay 7,000 currency, and if you push over a character it says “Assault on Characters,” you need to pay like 500. That feature alone completely changes how the player perceives the Chapter. It makes it less sterile because if that feature wasn’t there it would be more like plain platforming. But A Hat in Time isn’t just about hardcore platforming it’s about delivering an experience. Seeing people react to getting a bill of 100,000 currency, and just being like “Oh shit, I hope I can pay this off” and making jokes about “Oh, maybe I have to take out a loan or something,” it adds a lot to the game.

We called these features gimmicks, and when you think of gimmicks you might think “Oh, a gimmick is something that’s to the side and is optional”, and some people might like it and some people might hate it. But for A Hat in Time, gimmicks are like, the core thing. And it doesn’t make it a gimmicky game, it just means that that gimmick often changes how you perceive a level entirely. Do you remember the, what’s it called, the “Picture Perfect” mission?

Is that the one on the Parade level?

JK: Yeah, it’s the last one, it’s where you need to get a fanclub going. So early on, that level was just about going around and getting your picture taken and the gimmick that was added later on, that was the fanclub meter. And you know, it’s just a stupid UI element. It doesn’t really affect the level in any gameplay way, but it completely changes how you perceive the level because it gives you that feeling of oh, you’re in this scenario and you need to do this “something,” and this scenario is interesting because there’s a dynamic now: you need fans.

I think the level that has the best gimmick has to be “Murder on the Owl Express.” Because not only does “Murder on the Owl Express” have the suspects, it also has the timer, that guy that calls you up all the time and says “Hey, what are you doing, get off my train.” And those gimimcks don’t really actually do anything. But the player doesn’t know that.

And then the suspects; if you collect a certain number of suspects you can point your finger at a specific person. No matter who you pick it’s always the correct answer. The result isn’t too important, the journey is important. The player should have fun along the way. But just adding those gimmicks on top, like, it makes a huge difference because now the player feels like a detective. They’re going, “Oh, now I need to find suspects and I need to make sure that I get into every nook and cranny just to find those suspects.” So those gimmicks play a huge part in why I think A Hat in Time is so fun.

Another big reason is also we just cut a lot of stuff that wasn’t fun. For instance, I think modders will definitely find this out, but there are true, fully working missions in A Hat in Time, in the game files, that are just not in the game. They’re not playable, and we just cut those because we looked at them and were like “They’re not bad, they’re mediocre or a bit over mediocre.” It was a dialogue between did we want the game to feel long and kinda okay-ish, or did we want the player to have a blast and wanting more, right? And we just decided, well, short and sweet is better because the player is left wanting more, whereas long and mediocre is just gonna leave the player feeling “Meh.”

What was your reaction to seeing “Overwhelmingly Positive” on Steam?

JK: Ahhh, it was amazing! And not just on Steam, seeing “Overwhelmingly Positive,” but also seeing the reviews. We’re getting 8 and 9s out of 10, which is incredible because we were expecting like, 7 out of 10, right, because we’re a new studio, we’re not AAA, our game is super niche. We didn’t expect a lot of people would play it, which is wrong by the way because the sales have been incredible. Just seeing all the reception and seeing people rating it so high has been incredible.

It’s been really cool to see all the fan content. I think you guys did a good job of putting stuff in the game that promotes content creation. What are you hoping will come out of the modding scene? 

JK: We give people access to all of our tools. They can make just as good content as we can because they have our tools, and I really hope there is someone out there that just makes some really good levels, like full, big levels. That would be really cool. Like, there’s already a lot of small Time Rifts that are made by modders, and those are really cool as well, but it’d be cool to see a big scale level.

I’d love to play that just because when I play A Hat in Time, I can’t detach myself from the fact that I made this game so I’ll always see the small details and remember how something was made. But being able to play a custom level from the start? That’d be super cool. I’d love that.

I did a bit looking up on the internet and I saw that you had started Gears for Breakfast in the middle of school. How did you manage to balance your time? Between working on the game, being in school, and giving yourself time to breathe, that’s a lot to manage.

JK: Yeah, I graduated with a Master’s Degree in Computer Science a little over a year ago. So I was pretty much in university for the majority of development. University sucks up a lot of your time and you gotta prepare for that. You gotta make some time sacrifices in order to make that happen if you want to balance a game and your education. Obviously, don’t waste your education, especially not in the U.S., just because you want to make some crappy game. In Denmark, it’s a bit easier because education is free, but in the U.S. don’t waste that money. But yeah, it was tough at times.

Through all of that, did you find any time to play new or recent video games?

JK: No, dude, that’s the thing. I haven’t played a game for like a few years now. Well, I have played some games, like small games that have been able to eat up my time. But I actually made a rule for myself: I wouldn’t play Breath of the Wild until A Hat in Time was out. And I still haven’t played Breath of the Wild! I’m planning on playing it next week for the first time.

Being an indie developer with a significantly backed Kickstarter campaign, how important was transparency for you? Did you draw any lines when you established your PR approach?

JK: Well, on Kickstarter, transparency is difficult. You want the people who backed your project to know as much about your progress as possible. But not everyone is happy about transparency. When you look at the Double Fine Kickstarter campaign, “Double Fine Adventure,” people were dissatisfied by some of the things that were shown in the documentary. But that’s just because a lot of people don’t want to see how the sausage is made. So you kind of have to balance it on Kickstarter: it’s a balance between what you want to show and what you should show. And some people are upset. Some people want to see everything. But, you know, you’ve got to keep some things secret for their own sake.

Like, for instance, for A Hat in Time specifically, someone asked us, “Are you guys planning on making Hat in Time for the Nintendo Switch?” And that is not in our plan. Back when this question was asked, we had approached Nintendo about development kits, but because the Switch was all new they didn’t have any more in stock. They couldn’t deliver anything for us, because I imagine a lot of AAAs were asking them for stuff. So they just told us there were no dev kits available. And we’re like, alright, that’s fair. Fair enough. So we just let our fans know, we told the guy who asked us, “Yeah, we tried contacting Nintendo and they said there weren’t any Switch kits available.”

And people freaked, like, freaked out because we said that. They were like, “What do you mean Nintendo doesn’t have kits for you?! Are you trying to blackmail Nintendo?! Are you trying to smear Nintendo’s good name?” And we were like, “Wow, chill out.” So in terms of PR, some people are interested in knowing how progress is on a console and stuff, but you don’t want to tell them that because some people get super riled up about transparent stuff like that. Ever since, we’ve sort of just been quiet on the subject.

I think you have to be careful about being truly transparent because some people will take it to heart. I mean, it’s 2017, a lot of people get easily upset about a lot of stuff, even if you’re trying to be as transparent as you can. So you just have to be careful about that kind of stuff. Some people appreciate it, but some people get upset. Especially if you’ve watched the Double Fine documentary, there’s a part where they talk about splitting the game up in two because they’d spent all the money on part one or whatever. And people are still talking about that to this day.

But that’s the reality of game development. There are a lot of unexpected costs and projects often become much bigger than anticipated initially, and that’s just how the sausage is made. But some people, especially people who aren’t too familiar with game development, they just don’t want to know that. They just want to see the game is announced now and then one year from then they want to see that game is released, even though that game might have been in development for like, five years.

Especially for A Hat in Time, a lot of people are saying, “Oh, well, A Hat in Time took like four years, that’s a long time. Ugh, four years. That’s an eternity!” But, in reality, most projects take at least four years. So that’s just people being oblivious to how game development actually works, and a lot of people aren’t willing to be fully transparent about this fact for that reason. You have to strike a good balance when you handle Kickstarter and transparency.

With some of the more unsavory parts of community engagement, there were comparisons made to Yooka-Laylee and fights over which one was better. Did it ever give you any trouble? 

JK: Sure, I mean, that’s always going to happen. Even if Yooka-Laylee didn’t pop around they would’ve compared A Hat in Time to Wind Waker or Mario Sunshine or whatever game, right? There’s always going to be comparisons. Like, when a new game is introduced, the first thing people always say is, “Oh, that looks like another game.” And that’s just how people talk about games. That’s natural.

When we saw Yooka-Laylee arrive on Kickstarter we were like, “Aw, that’s so cool!” We can’t play A Hat in Time with a fresh mind because we’re so close to the project we’ll see everything, so Yooka-Laylee looked like something we would be able to play ourselves. But when it released to lukewarm reviews we were kinda worried that A Hat in Time was also going to be received with lukewarm reviews. But luckily that wasn’t the case! People don’t seem to care. People seem to love A Hat in Time, so our concerns were unfounded.

From your perspective, what do you think you guys brought to the table that was unique to your backgrounds?

JK: Well, I think for a lot of these revival games, where the creator goes in and tries to recreate an old game, they don’t fully understand what made their game good. And so they try to recreate it to the point, right? They try to encapsulate both the good and bad because they don’t fully understand the source material.

There is definitely a market for that; some people want that experience. They want the old games almost exactly, right? But for A Hat in Time we didn’t really want that kind of thing. Like, we acknowledged the games we love, like Mario Sunshine and Psychonauts and such, and… those are mediocre games. [laughs]

I love Mario Sunshine and I love Psychonauts, but they also have big parts of them that are problematic. Like in Mario Sunshine, it’s how the pacing completely stops halfway through, half the missions are boring, and Red Coin missions are copypasted. You have to recognize that the games have flaws, and once you recognize that you’re able to accommodate for it. So, for A Hat in Time we have no Red Coin missions for instance because we recognized that those aren’t too fun. Or at least, we didn’t feel that they were too fun.

And in stuff like Psychonauts, like, Psychonauts is such a funny game! It’s such a good game! But the pacing is also weird and it’s a bit too story heavy at times, it’s kinda difficult to get into it. And so we recognized that as well and tried to bring in the Psychonauts humor and scenarios but without being too story heavy, because we also recognized that players don’t want to read too much when they’re playing a 3D platformer.

What I really appreciated with what Hat in Time does is how it gives the player a lot of different choices in terms of how they set the pacing for themselves. How did you decide on what content would be in the game? When did you say, “Alright, this is the final product that we want to package and that we want the players to experience.”

JK: Choice is definitely a thing, and freedom of choice is something we were really big on for A Hat in Time. You can choose to go straight to the objectives or you can choose to just spend an hour wandering around. The final mission in the game is unlocked at 25 Time Pieces out of 40, right? That’s just a little over half. And that’s deliberate. A lot of people might say, “Oh, well you should put it higher. You should put it at 30 or 35.”

But we also recognized that some people, when they get to a certain point, they want to end it. We noticed especially one of the complaints for Yooka-Laylee was that it required an insane amount of Pagies (currency) to get to the final boss, and people were upset about that. You don’t want to overstay your welcome, so giving players the choice to exit out early makes it so that the game never turns sour. You never become upset with the game because you have the choice to just stop, which is really important.

And that kind of freedom of choice is also in our collectibles. Stuff like “Yarn”: you can decide whether to stitch an “Ice Hat” or a “Brewing Hat” or a “Fox Mask” or whatever, right? It’s not just a matter of the game saying, “No, you need this item to get to this thing” or “You need this item to get to this world.” You get to decide what you need. It’s not always useful all the time, but a lot of it will assist you in your journey. And so the player’s always motivated to collect those things because they are always a net gain to their adventure. Choice is mentally important to us for that reason.

But yeah, it was probably clear to us, I think, last year or so what kind of missions we wanted in the game. We knew that some people would say that our game was short. It’s not large for a AAA, but it’s large for an indie game, so we felt pretty comfortable with the number of missions we had and shipping it the way it was.

As you’ve tried to figure out the more formal side of game development, what challenges made you realize, “Oh, well I guess this is something we have to deal with now that we want to release our game to the public”?

JK: I think performance definitely came back to bite us in the ass. So, a lot of people assume that big AAA engines like the one we’re using, Unreal Engine 3, are optimized to hell and back, right? But that’s actually not the case. Even modern game engines are so primitive and so basic. There’s such a lot of room for improvement, especially in terms of performance. We applied so many performance improvements for Hat in Time, it’s insane. It’s almost its own new engine.

So that definitely up as a surprise to us, that we had to fix a lot of stuff we thought was obvious that should be fixed. I’d have loved to spend more time doing performance tweaks, but we had to release at some point. I think performance now is pretty good, but I’d love to do more of it to be honest.

One last question I had that I don’t think I’ve ever seen the answer to: Where did the name come from?

JK: Gears for Breakfast or A Hat in Time?

Well, both I guess!

JK: It’s kind of a long answer. A Hat in Time used to play in a way where you could diverge timelines. Imagine the Acts in A Hat in Time now, but where some Acts are only available in certain timelines. You could only play one timeline, and then when you got to the end of the timeline you could wind back and take all the timelines and whatever. We really liked that mechanic; it’s sort of related to what Wario Land II does. But, we also found that it reduced the playtime significantly, and a lot of people did not like the fact that felt short, even if they just hit one of many endings. So that’s why we changed the time mechanic.

Now “Time” just means the Time Pieces and “A Hat” just refers to Hat Kid. But way back it used to mean something completely different. If I had to name the project today, I’d probably name it something else. But, the name has stuck so it’s just what we ended up calling the game.

And Gears for Breakfast is just because it’s an unorthodox name. It makes you stop and think, and I like that. It’s cute. Because A Hat in Time is really quirky and, I feel, unorthodox in its own way, I kinda wanted the company name to reflect that as well.

‘A Hat in Time’ is now available on Steam, and will be coming to PS4/Xbox One this fall.

Links: Website | Steam | Facebook | Twitter