Go Bananas: ‘TOSS!🍌’ Now Available on Meta Quest 2 + Pro

When TOSS!🍌 was pitched to the Vertigo Games team shortly after they joined Embracer in 2020, they knew they had something special on their hands. A unique blend of a VR platformer and primate-fueled parkour action, its accessible yet challenging gameplay has players throwing themselves across dozens of increasingly difficult courses as they collect parts to repair their crash-landed spaceship. Thanks to a ghost leaderboard system, you can compete against other monkeys around the world—or challenge your friends to show them who’s boss. Agility and timing are the name of the game, which is available now on Meta Quest 2 and Pro for $19.99 USD.

While publisher Vertigo is a household name in the VR community, you may not have heard of the developer just yet. A small, three-person studio, Agera Games has been working on TOSS!🍌 for just over five years. We sat down with Founder & CTO Simon Bergqvist to learn more.

SB: I grew up on a farm in a rural part of Sweden, where I started programming at age 12. I’ve spent an unfathomable amount of hours on my computer gaming, browsing, programming, photoshopping, making music, and working.

I remember the very first time in life when I realized that games actually consist of many small pieces that real people have created. I used to play a lot of The Sims 2 with my sister, and I got into modding forums and started downloading mods. Somewhere around 2006, I made my very own texture pack, which I recently learned still exists out there. It was very bad - I had just taken a couple of pictures off the internet and swapped out the textures of a couple of paintings in the game. One of them was the The Sims 2 logo itself, but it was flipped, so you had to look at it in a mirror for it to look right. I still uploaded it.

A few years later, I got accepted into an upper-secondary school with gamedev on its curriculum, followed by the University of Skövde where I studied Game Development Programming, which has been the source of most of my gamedev network. I would say at least 90% of my knowledge comes from game jams and actually making games, rather than something I learned in school.

Since then, I’ve been the founder and co-founder of three different game companies, one of them being Agera Games. TOSS!🍌 is our first commercial title together as a studio.

SB: Albin Ahlstedt Grafman and I met in our local game development community called The Great Journey. We did some game jamming together, and then Albin won the seed money for a company in a pitch contest. We realized that we had similar ambitions and complementary skills, so when the idea for this game came up, we went ahead and did it.

Eddie Klint joined us a couple of years down the road. He was studying game development at a program where I was the programming teacher, and he needed a place to intern. After the internship, we took him out for ice cream and asked him to join the company.

Being such a small team lets us be creative, make quick decisions, and involve the perspective of our many roles in each of our meetings. There might be an opportunity for us to grow with the release of TOSS!🍌, but we’re not planning on rushing into things too quickly.

Simon Bergqvist: We were sitting around a table at our local game developer community one Thursday night, discussing games and virtual reality. I remember being very adamant that walking around with joystick controls in VR feels so backwards and outdated. And so, the million-banana question appeared: How can we make movement in VR fun?

By grabbing on to your surroundings and dragging yourself wherever you’d like to go! The bananas, however, didn’t come into the picture until about a year and a half later. We started out by making a quite simple climbing game, which hundreds of people came to play at the Sweden Game Conference where we exhibited it.

Getting Over It with Bennett Foddy, released the year before, was quite popular at the time. We drew inspiration for it, and the first level we made was this huge tower with an especially hard jump, which is probably why we implemented the “toss” mechanic in the first place. Also, you only had a limited amount of energy before you had to stop and recharge. It was a game that was high-risk, and we encouraged you to think before you moved. It wasn’t very fun.

With the success at the conference, we started doing some proper market research and found out we weren’t the first climbing game in VR out there—but one of our features stood out: tossing.

We tried to get rid of climbing altogether while also going a really weird route of wanting to be a rhythm game, where your actions controlled the music that was playing. You tossed yourself between these pads that came at you and tried to go as far as you could. We even threw some vaporwave in there. It... didn’t turn out great.

We pivoted once again, and in late 2019/early 2020, we can officially say that TOSS!🍌 was born. We were making a platformer in VR, along with level progression and obstacles and enemies and friendlies—and who can possibly be so strong as to toss themselves around like it was nothing? A banana-munching monkey in the sky, of course!

SB: There has been a bit of exploration into other non-traditional technologies. We once made a spin-off called TOSS!🍌: Banana Catch where you used the silhouette of your entire body to try to catch all the falling bananas using camera-based body tracking. It was due to be part of a showing at a shopping mall in our city, Karlstad, Sweden. The pandemic put a pin in that though.

SB: At least seven.

SB: Hah, there are so many to choose from! Like the one time we 3D-printed a 6′7″ banana and hung it on our office wall. Or the time we spent an entire meeting climbing trees outside of our office to understand momentum. Or the time we experimented with pulling your eyes apart from each other in VR. That last one isn’t really something we would recommend.

Or the time we got a retro banana phone as a gift from our friends over at Cuebricks. Fun fact about Cuebricks: Not only did they create the sound and music for TOSS!🍌, they also produced the music and soundboard for American Song Contest. One night they texted us being super excited and telling us that Snoop Dogg, who was hosting the show, was rapping over music they’d created. You can find it if you search for it. So, uh, in a way, we kind of share producers with Snoop Dogg. Which is cool, I guess.

SB: It’s in the details. I’ve spent so long working on it, and the most fun I have is always when we’re coming up with something new: when we see the potential a new feature can bring, how it fits in with the rest of the game and the world. Whatever feature or thing we did last is my favorite thing.

I’m currently prototyping something very secret that I’ve not told a living soul about yet, not even my team. If we can make it work, I think it will definitely become my favorite part of the game.

SB: HYPED. Some of the Reddit posts we made a few years ago when the game was early on in development blew up and got us many hundreds of people playing our demo overnight. We also exhibited the game at Steam Next Fest once and people loved it, praising its smoothness and asking us to release it right away.

I would like to thank the community for being so aggressively patient. Many thought the game was pretty much close to done during our public demos two to three years ago, but we wanted to add more stuff to it and make it prettier. We’ve added lots of obstacles and enemies, worked very hard at the level progression to keep the game fun, added a bunch of features, and made it work on many platforms. Looking back, it’s such a different game now, and I can tell you that it is all for the better.

SB: The day Quest was announced, we did a 180 and decided to make everything mobile-first from here on out, as it was obviously the future of VR. I had prior experience working with mobile games and knew that overdraw and draw calls would likely be our biggest concerns. We also looked at the recommended polycounts and, just to make life easier for everyone, I lied to my team and told them our polycount and draw call budget was 20% lower than it actually was. We, of course, went over my stated budget, which proved that this was a good choice and I would recommend you to do the same.

We then laid down some ground rules. First, we decided to design the entire game around having no-to-minimal transparency, as that’s the usual overdraw culprit. In the end, we’ve broken this rule a few times in the UI and with particles, but for the most part, it actually holds true, even in the UI. This became a pillar we came back to for every single graphical choice, and we shaped the entire aesthetic with this in mind.

We also looked at the various ways Unity would allow us to batch models. There’s GPU instancing, dynamic batching, static batching, and of course baking your models after they’ve been laid out. I recommend you play around with this a lot before you come to any conclusions regarding your own game, as there’s a lot of complexity that comes into play when you start factoring in how far away objects are, whether you’re inside using rooms or outside with everything in sight all the time (not too fond of us making this design choice out of an optimization perspective), and lots of other things.

Our final optimization pass was mostly cleaning up levels and batching some animations. CPU-wise, we just had to be mindful not to dump too much computing into a single frame and spread it out a bit—outside of that, our code ran pretty smoothly.

We also followed pretty much all the basic guidelines provided in the Quest documentation, which gave us a good start.

SB: Get comfortable with some flat game development first.

Do game jams. Seriously, do game jams. Just game jam after game jam. And lots of them.

If I were to create a new VR project completely from scratch right now, the most important thing would be to make literally as much of the game as possible runnable and testable without putting a VR headset on. There’s a special kind of headache involved with putting your headset on for the umpteenth time that day just to see if this one button in this one menu is correctly placed. Save the VR sessions for testing only VR-specific features and general game testing to see if it’s fun.

SB: We’re preparing the launch to be as epic as possible, and TOSS!🍌 still has more content to come. There’s a reason why we were pulling our eyes apart in VR the other week—and not just morbid curiosity.

As to what comes after TOSS!🍌, we have some seeds and saplings slowly growing. They’re going to need a bit of time and a lot of love before they’re ready to be seen by the world though—not just from us, but also from our community. What would you like to see?

SB: One member of our team has promised to get a tattoo of a banana if we get to 100,000 copies sold. We’re trying to convince Vertigo to let us put up a public counter so everyone can contribute. Let us know if you want to see it as much as we do! 🍌🐒