‘We fill a gap for PlayStation Studios’ – Behind the scenes at Astro Bot’s Team Asobi
The PlayStation studio opens its doors to VGC and reveals the secrets behind PS5’s profound platformer
On the 15th floor of Sony Interactive Entertainment’s Tokyo building, we follow a trail of gold PlayStation coins to the entrance of Team Asobi, the unconventional studio behind PS5’s most brilliant platformer.
Stepping inside the open plan development floor, which is strewn with colourful mementos of PlayStation’s past and present, we’re initially struck by how spacious the room is for a first-party studio – especially in Tokyo, where floor space comes at a premium.
There’s room for a ping pong table, multiple sofas and break-out areas, a small reception decorated with awards, memorabilia, and photos of staff members, and even a traditional Japanese tatami room filled with retro consoles.
Then we realise it’s not the room that’s spacious, but the team that occupies it that is – for a first-party studio making AAA games, at least – surprisingly small.
EXCLUSIVE VIDEO: The Making of Astro Bot
At around 65 people, Team Asobi is PlayStation’s smallest first-party studio (at least of its established teams). Naughty Dog, Insomniac and Guerrilla all have over 400 employees – and even Housemarque and Media Molecule have virtually double the number of developers of the Astro Bot studio.
Yet, it’s often Asobi who passes knowledge to PlayStation’s blockbuster creators, particularly around the use of PS5’s most unique hardware features, such as its DualSense controller. Now with Astro Bot, its modest team has managed to create one of the most genuinely delightful video games of the modern console era in just three years. So what’s the secret?
Just like its imaginative adventures, Asobi is imbued with an inventive spirit. During our visit, nothing about the studio felt off-the-shelf: its structure, culture, development process and technology appear to be carefully considered for the kinds of playful toys it makes, utilising the bleeding edge of PlayStation hardware knowledge.
Its setup and team size would almost certainly strain if it were making the kinds of gargantuan blockbusters other PlayStation studios are known for, but we don’t believe Guerilla or Naughty Dog would be truly capable of making Astro Bot either.
A TEAM WITH PERSONALITY
“We think that bigger is not better always,” says gameplay programming lead Masayuki Yamada. “For Astro Bot, we think this team size is the best, because the interaction and discussion is very important with team members. If the team is bigger, the interaction is not that fluent.”
According to studio head Nicolas Doucet, part of the studio’s secret is being meticulous with who it hires – not just in their ability as game developers, but also their empathy as people.
“We’ve always kept a long-term approach of not just growing the team for one project, but growing the team for the future and to keep people,” he explains when asked why he’s decided to keep Asobi small, even after it became a standalone developer following the closure of Sony Japan Studio.
“For that reason we’ve been really, really cautious about growing the team slowly…. It’s important for us to ensure that every member of our team is bringing something special and unique and has a real ‘game’ mind,” he adds.
“When we do interviews, we check first for whether they know our games and are genuinely interested not just in the product itself, but also the soul behind it. Because if you’re able to relate to the game and to the consumer, and if you care about how the user will feel when they’re playing, then it’s likely that you will care about how the people around you every day feel, right?
“That’s an important conversation we have when we hire: besides the skills, it’s also about the attitude and the fit within the team. Sometimes, that means it takes us a long time, and our interview process can be tough because it’s not like you can absolutely tick every single box, and you never truly know until you can spend time together. But yeah, we are very cautious.”
Jamie Smith, Asobi’s British animation director best known for bringing to life the protagonist in The Last Guardian, explains that one of the things Asobi looks for from potential hires is being able to take on board criticism, which is essential within Asobi’s fast-paced, iterative production style.
“Personality is massive,” he says. “We need people to be open, we need people to be positive. To make a game like this with this kind of creativity and positivity, there’s only certain types of people who can do that and we need people who are super positive, open.”
He adds: “One of the most crucial things is the willingness and openness to take feedback, be able to take feedback in the right way, in a positive way, because we’re constantly giving each other feedback. There’s never any dishonest, like, ‘oh, that’s fantastic’, or ‘actually it’s not, but I don’t want to hurt that person’s feelings’… We’re honest and we have a team of people that respect that honesty and they don’t have any ego.”
WESTERN SPEED, EASTERN PERFECTION
Astro Bot is Asobi’s first game developed as a standalone studio, and although the developers we spoke to weren’t keen to discuss the closure and transition from Sony Japan Studio (though it’s clearly honoured in Astro Bot itself), Asobi has clearly implemented a unique working culture and identity all of its own.
Team Asobi – ‘asobi’ is Japanese for ‘play’ – was first formed in 2012 as an internal team within Sony Japan Studio. Utilising Doucet’s previous experience working on Eye Toy at PlayStation’s London Studio, the new team was tasked with creating technical demos for DualShock 4 and the PlayStation Camera.
Doucet brought to Japan not just expertise for augmented reality games, but also learnings from working within the culture of London Studio, as well as his previous position creating video games at Lego.
HOW MEDIA MOLECULE SAVED ASTRO
Astro’s origins trace back to an early PS4 demo designed to showcase its controller and camera. ‘Little AR Men’ would eventually evolve into The Playroom, and its blocky characters would become Astro – but if it wasn’t for an intervention from another PlayStation Studio, the character might not exist at all.
“Fun fact: [The Playroom Astro demo] was the only PS4 game that you play with the camera tilted down to the floor, and that was something that we had concerns about, because it meant it was a bit annoying having to move it,” explains Doucet.
“I was at a trip at Media Molecule in the UK and I was telling them about this demo, but we had to move the camera so maybe we should remove it. And one of them was like, ‘are you crazy?! This is really nice!’ And I remember going back and thinking, well if Media Molecule think it’s a good idea, maybe we should leave it in. We discussed it and ended up leaving it in. It’s a good thing we did!
From the latter, Asobi adopted five values that it decided should inform all of its work going forward; innovation, playfulness, universal appeal, polish, and a magic. From London, it adopted a fast, experimental design culture, with developers creating regular prototypes to test ideas and quickly gather feedback through playtesting.
“In London Studio I was lucky to work with people that were extremely smart about prototyping,” the studio head explains. “There was this one designer who was really good at making a methodology for fast prototyping and that was really for me a transformative moment when I thought, okay, if you’re going to be touching technologies and you know how things can fail easily, you’ve got to be able to get on your feet as fast as you can and in order to do that, rapid prototyping is super important.”
The walls of the Team Asobi studio are plastered with colourful post-it notes, each containing a scribbled gameplay idea. Doucet shows us how many of these drawings were turned into playable prototypes by a single developer in just two weeks, after which – if approved – they made it into the final game with few significant changes.
One video we’re shown demonstrates how a basic concept for a sumo enemy was turned into a playable demo by a single programmer. The demo is surprisingly content complete, with sounds, animations (albeit reused from existing assets) and even haptics. Once approved, this sumo enemy went through an art cycle, with some physics and tweaks added, but Doucet estimates that 80% of the final enemy is the same as that first test.
Another prototype we’re shown is “big hands” – an evolution of the monkey suit from Playroom, which allowed players to climb walls using the triggers to ‘grab’ rocks. This test made the suit part of the core platforming gameplay, allowing players to grab rocks and lift heavy objects while simultaneously running and jumping.
Again, like the sumo enemy, other than looking prettier, the final ‘big hands’ mechanic looks virtually unchanged in the final game compared to this two-week prototype. Even a boss fight we’re shown next is virtually feature complete in the test compared to the final game.
Nico explains that elements such as difficulty and art are tweaked right up to release, but otherwise it’s this quickfire testing philosophy which directly fuels the Astro Bot games’ inventive design, with their rapid conveyor belt of physics toys, gyro gadgets and unique gameplay mechanics.
“We often hear it, but in fact, efficient rapid prototyping is actually quite a difficult thing to put in place… there’s a massive learning curve,” he says. “When I came over to Japan, this two-week approach was something we had started cultivating in London Studio and then in terms of taking all of that knowledge into Asobi, it was all about trying to get the best of both worlds.
“Japan is really well known for being very meticulous and perfectionist, but it can also be sometimes slow as a result. So it’s about keeping this kind of speedy approach while still never taking away from that perfectionism that has made this country so amazing, and trying to combine the two.
“The moment we start progressing fast and our games come out fast, and people feel like, ‘okay, every two weeks I get to work on something fresh’, everybody buys into that process: you don’t have to convince people, it’s just that once you’re in it, it works, so we can be perfectionist and still be nimble. It works.”
This fast prototyping almost demands multidiscipline within Asobi’s team members, and again, highlights why it’s so picky with who it brings into the company. With developers able to touch different areas of the game, according to animation head Smith, it also creates an environment where everyone feels responsible for all areas of the final product – something that can sometimes be lost in giant teams.
“It’s not just the multidiscipline, it’s also that everybody has a say in everything,” he says. “The way we review work is completely open. It’s not a top-down process where one person makes a decision. What we do every two weeks is we show that two weeks’ progress and the whole team has a voice and gives feedback on what they like, and what they think needs improvement.
“For example, I’m not a game designer, I’m an animator, but if I play a level and I feel like there’s issues with the level or it can be improved, I’ve got a voice to say that and everybody on the team has a voice to say that.
“There’s a real openness about it, which is a real strength to the team because there’s an honesty about the quality. There’s no like, ‘oh, you can’t talk about that because you’re not in this department’. We can all comment on anything. The quality of the game is more important than anything.”
Art director Sebastian Brueckner, a German who previously worked on the Wipeout series, agrees: “There’s no need for having a big ego. It’s just all about having your mind and your head in the game.
“Understanding gameplay is one of the core things in this team. Really understanding what is the product that we’re making, how can I make it better, how can I improve it. And if you have a great idea that elevates the gameplay, elevates your world experience, then no one technically could stop you bringing that into the game.”
IN SONY’S SHADOW
It may not feel like it for those visiting Sony Interactive Entertainment’s Tokyo headquarters, which features a lobby decorated with Astro Bot as if it were the founding mascot of the PlayStation brand, but Astro Bot has only seen one full-sized release before this week – 2018’s VR game Rescue Mission – in addition to the free PS5 pack-in game, Astro’s Playroom.
Q&A: Tech of Astro Bot
Gameplay programming lead Masayuki Yamada
Astro Bot uses a custom engine. Why?
One of the benefits is we can do what we want to do. So if we use a generic game engine, if we want to do something strange, something different from the others, we need to investigate the game engine’s internal things. But if we use a custom engine, we can create what we want to do directly, so the speed is much faster. For us, the iteration is very important.
As a programmer, was it more fun working on Playroom or Astro Bot?
In the last game, we were working on the controller at the same time. In the current game, the controller was fixed. Which one was more fun? Both have good things, I think. So at first, we created a lot of prototypes, based on unfinished hardware. So it was very fun. But it means that all ideas couldn’t be included in the game. This time, we could select from a lot of those ideas and polish them up.
Did you make any considerations for controller battery life, considering this is a longer game?
We haven’t changed anything around based on battery life or anything like that.
The latter, which was designed as a tech demo for the PS5’s Dualsense controller, was hugely well received by players, and in the years since – as SIE’s lobby attests – the character organically became a mascot for the PlayStation brand, appearing at events such as the Champions League final. Doucet says this gave them the clout they needed to go big with the next Astro game, especially since PlayStation – which the last Astro game so wonderfully celebrated – marks its 30th anniversary this year.
Astro Bot is Asobi’s biggest game to date – twice the size of Rescue Mission, and ten times bigger than Playroom. The game revisits a mechanic from the VR title, where players save stranded robots. It also ramps up the nostalgia hook introduced in Playroom, with hundreds of cameos from PlayStation’s history, from the obvious to the highly obscure. Some of the most popular characters, such as Kratos from God of War, even have their own bespoke levels.
For the new game, Asobi wanted to double down the use of haptic feedback – a feature that Playroom arguably still uses better than any other PlayStation 5 title. Doucet says he initially showed PlayStation bosses a quote from a Washington Post feature to illustrate “the impossible standard” Asobi had set for its use of controller vibration, and a continued goal for the sequel.
Asobi quickly established a brainstorming team, dubbed ‘DualSense 2.0’, to create prototypes for new mechanics, but not necessarily containing Astro, such as horse racing, darts, balloon inflation, or tennis. Some tests, such as one featuring a sponge absorbing water, inspired new powerups for Astro. Others, such as a demo which had players feeling a texture on a wall, made it into the new game as a way of uncovering secrets.
The team also dug out ideas left over from when they were testing the DualSense while it was still being made, such as a chainsaw which players can use to cut through ice in one particularly satisfying game scene.
Doucet estimates that this kind of R&D work takes up about 10% of its typical game production time, but the inventive mindset is clearly something that runs deep through the culture of the team. And that’s not especially surprising when you realise that Sony’s main corporate headquarters is within stumbling distance.
“Team Asobi is the first game studio to get controller prototypes, because we’re literally across the street,” Toshimasa Aoki, a director on the PS5 product team, tells VGC. “Asobi’s mission to create magical experiences for everyone was also what we wanted to target with the PlayStation 5 console as a whole, so that was the number one reason we had a lot of work with them.
“The physical closeness was the cherry on the cake, the fact that we were able to just hand carry stuff over here, like literally coming out of the factory or the engineering desk that afternoon, bringing new controller prototypes to Yamada-san’s desk… If we only had one prototype, we would literally hand carry it here in a bag so that nobody could see it!”
He adds: “They might say, ‘I want to try this’, and we can go and get it made, or add more firmware, and try it… if they were in the US, we would have to ship it and wait. That turnaround is big, especially in the early prototyping times.”
Asobi’s gameplay tests using early DualSense prototypes would demonstrate if the hardware team’s ideas would translate into actual gameplay, and Aoki says this was crucial for his team to understand which features it should pursue further for the controller. In the process, this also helped out other PlayStation studios by providing them with demos for DualSense.
“This allowed us to understand both the player and creator experiences,” says Aoki. “With that, I think we were really successful with DualSense and having Astro’s Playroom pre-installed. Now they’re taking that to the next level with Astro Bot, and I’m really excited to see how players react.”
PLAYSTATION FAMILY
Team Asobi feels genuinely unique within the blockbuster machine that is PlayStation Studios – not just in terms of its size and working practices, but also the types of games that it’s making – a positioning that’s very deliberately off center from its company colleagues, Doucet says.
“If you consider the history of PlayStation and how every studio has positioned themselves, it’s interesting that – and this is also what I show management – as the audience of PlayStation grew, a lot of these [studios] went from making cartoon games to a little bit more mature, hyper-realistic games.
“That’s when we at Team Asobi as a studio decided that we really wanted to fill that gap, that used to be the gap filled by other studios 20-25 years ago. That’s really where we want to be: all ages, colourful etc. Even if we made a horror game, it would be a funny horror game.”
So far, Doucet has demonstrated four of Asobi’s five key values – the magic and innovation of DualSense, the playfulness of its design and animation, and its commitment to polish – but one equally important pillar remains the international makeup of its team.
He says that 75% of Asobi is Japanese, but it’s notable that many of its leads named in this article – including Doucet – are foreigners in Japan. The remaining 25% of the team covers at least 16 countries, and on top of its imported work practices, Asobi argues this puts it on a good footing for creating games with true international appeal.
“We really love and embrace the Japanese culture… but also we need to make games that resonate with the whole world,” says Doucet. “It’s very important when you work with family games because humour, for example, can be very different according to each culture. Having a way to discuss and check that is really important.”
What that means practically is that there’s a distinctly international menu in the team cafeteria, and the team communicates in a mix of Japanese and English, depending on the group. PlayStation also provides free lessons for both languages to Asobi’s employees – on work hours.
“The PlayStation Studio family today is largely abroad, so we need to be able to communicate. And besides work, it’s also like adding another string to people’s lives that is actually interesting because learning a language is something cool anyway, right?”
What lies next for Asobi is unknown (publicly, at least), but if Astro Bot is the enormous success it looks set to be, it’s inevitable that the studio will cement itself as one of Sony’s most important assets. One imagines that the team will have carte blanche on its next project to push its creativity even further.
One also imagines, given its key principles, that even if the game’s sales are stratospheric, Asobi won’t use that as leverage to grow its team too much, for fear of disrupting its creative process. It would probably be a good thing in terms of space, because we have the feeling the office is going to get a lot more awards come the end of the year.