Interview with Eric Matthews, Creative Director of Sony Computer Entertainment Europe

"Kung Foo", a fighting game in the EyeToy collection.  Image from Gamespot.

"Kung Foo" allows players to punch and kick at ninjas overlaid onscreen. (Image from Gamespot)

In 2002, Sony Computer Entertainment’s (SCE) London Studio took on the unique challenge of developing video games for a very different type of controller.  The EyeToy USB camera for the PlayStation 2 projects a player’s image onscreen, where he or she can interact with overlaid objects by moving around and physically touching them. 

The result was a compilation of novel games called “EyeToy: Play”, the first release for the controller.  In “Wishy Washy” the players wipe soap suds off of dirty windows.  “Kung Foo” has players chop and kick at hordes of miniature attacking ninjas.  “Mirror Time” confounds players’ visual-motor control by having them grab at objects while their image is repeatedly flipped backward, upside down, and back again.

In a market where such models for player experience had no precedent, the EyeToy became an international hit selling millions of units.  Last month I spoke with Eric Matthews, the director of creative development for Sony Computer Entertainment Europe.  He works with the team at SCE London Studio, which has led innovation in the video game industry with titles including “SingStar”, “The Getaway”, and the “EyeToy: Play” series, now in its fifth iteration.  We discussed user interface design, the development of the EyeToy, and the role of usability in game design.

How did the idea for the EyeToy get its start?

"Mirror Time" inverts player's image repeatedly to confuse their sense of left, right, up, and down. (Image from Gamespot)

"Mirror Time" inverts the player's image repeatedly to confuse their sense of left, right, up, and down. (Image from Gamespot)

The idea of using a webcam as a video game interface originated from SCE’s research & development facilities in the US.  Dr. Richard Marks had been experimenting with things like motion control, color tracking, and augmented reality.  He had developed some prototypes using cameras, but there wasn’t yet any game experience associated with it.  Phil Harrison, then the head of Sony Computer Entertainment Europe development, showed Rick’s work to the various game development studios to see if anyone could think of directions it could be taken.

Ron Festejo, who’s one of the people on our team, really wanted to look into it.  From there, Rick assisted on the technical side, but it was London that made it into a game.

How did designing for the EyeToy differ from design for more conventional games?

We indentified four things to focus on.

First, you are the star of the game.  You’re seeing yourself up onscreen, so there’s a bit of vanity there.

Second, it had to be as much fun to watch as it was to play.  We wanted it to be a great social experience, and something where parents and kids could both engage in play together.

Third, the games couldn’t be better played on a controller.  It wouldn’t be worth doing if you could only do the very same things as you could in any other game, so it needed to be something unique to the controller.
Fourth, it had to be easy for people to intuit, so you don’t need to have the rules explained to you.   You couldn’t have a control scheme as complicated as you do for other games, where you’ve got sixteen buttons on a controller.  And of course that has the advantage of making it more accessible to people who normally aren’t as comfortable with games.

How did you design gameplay for the EyeToy?

We started by doing a bit of competitive research, looking at the landscape and what things may have been done along the same lines. 

Then we asked what is it good at, and what are its weaknesses?  For one, it had a magical “wow” factor of seeing yourself onscreen interacting with the game world, which got people particularly excited.  But while it could pick up big broad movements, but it was not very good at fine granular motion.  Another disadvantage was that it was sensitive to noise in the background.  We felt it worked well as a short-form medium, with fast play and a lot of variation in the games. 

Then we ran some brainstorming sessions to see where we could take the ideas.   We wanted to see how many different forms of interaction we could create, so we went really broad.  From there we would say okay, we’ve brainstormed an idea for a boxing game, why is that a valuable thing?  Why would you want to play that? 

After that we went into a very experimental prototyping phase so we could get a playable experience onscreen quickly to see if the ideas were fun and engaging.  We developed around thirty prototypes, including some wild stuff that never worked.  We ended up with twelve in the final game.

Can you give an example of how one of the games was designed?

Well “Wishy Washy” evolved from an idea where you had to climb to the top of a building.  I sat there with Pete Marshall, the lead programmer, and we started thinking “Okay, how would you see that?”  And we said well maybe the camera’s on the inside of the building looking out through the windows at you.  Then that turned into, well perhaps you’re washing the windows as you go up.  Then it was, you need to clean the window on each floor before you could go to the next one.  Through that process of just talking it through over a half hour late at night, an evolution of thinking turned it into a window washing game.

How did you interact with marketing?

We got marketing involved to collaborate on what games had the greatest potential and harvesting the best ideas, and getting out to look for an audience.  They found that it was something parents could play with their kids, that could bring down barriers and get people moving.  After the game was finished they put demonstrations of it in clubs and shopping malls for the launch, because it’s one of those things you don’t really get until you’ve played.

One of the defining moments came at The PlayStation Experience, a London event where we first showed the prototypes to the public.  We showed three games and the press went nuts.  Everyone from small children to hardcore Tony Hawk skaters with multiple piercings were lining up to play it.  The gaming press wrote that it was something very unlike what they’d seen before, and there was a lot of excitement.  At that point we started to think okay, we may have something here.

Up to then, no one had thought it was going to sell millions.  Well, let me amend that – we thought it would either sell less than 100,000 or more than a million but not in between.  It turns out it sold somewhere between 3 and 4 million units.

Did you do any testing of the game with users?

This is going to sound terrible, but EyeToy: Play was the first game where we did formal user testing, and that was only once the game was finished.  We had done ad hoc testing using the people in the QA department, children of coworkers, things like that.  But this was the first time we recruited real users and set up at a facility behind the one-way glass and so forth. 

And it was an absolute nightmare.  Once people were into the game they had a great time playing it, but they couldn’t get there quickly enough because the flow of the menus was too long winded.  People also couldn’t figure out the right distance to stand, and then someone would walk across the room in front of the camera and inadvertently trigger something onscreen.  It came to a head when the players kept accidentally cancelling the setup process for their profiles.  Around then Ron Festejo got up and said “Stop the blasted thing, I can’t bare it anymore!”

So we all found the first test very tough, but it was a very valuable learning experience.  From that point on, we’ve tested all of the games we develop.  I’m a big proponent of the value of testing with users, and we’re building our own facility in London.

How did you develop the process for testing games?

There was no company in Europe that focused on testing video games.  They were all working in consumer software or Web-based usability.  And some of that’s relevant, but some of it isn’t.  They had a heavy task orientation, where you’re given situation A, now show me how you’d do B, and that’s not as relevant to games.  They just hadn’t had the experience of applying their methodology to a game.  How do you test for fun?

So we worked with an agency that had a lot of experience in Web called AmberLight, and collaborated to create a process that we use to test all of our games.

So what is that process?

First, you decide what you want to test.  That includes obvious things – the fundamental control system, how it feels and how it compares to competitors.  Then there are things like signposting, navigation, how long it takes to complete objectives – primary, secondary, tertiary.  We do a peer group review, and get the experts to look at it before going into testing for what we think is going to work well, and what we think might not work so well.

Then we test what’s called the first publishable, which is typically one complete level of near-publishable quality with all of the controls and visuals in place – a vertical slice of the game.  This comes about halfway through the development lifecycle.  We test it with 10 people playing simultaneously, but separately from one another.

Following the test we have them complete a questionnaire.  We ask primarily about usability concerns, but also look at appeal to some extent.  We finish the testing by holding a roundtable with those people, though we’ll sometimes do individual interviews as well.

In the end, we document all of the issues we’ve found, make a top ten list from it and say okay, these are the things we want to fix for the next time round.  We’ll do testing 2-3 times in the lifecycle with 10 people each time, bringing in different demographics.

Alongside that, Mark Cerny had laid the foundations for testing at Sony in the US over the last 10 years.  They’re doing a lot of data capture as well, recording how long it takes to complete each level and looking for things like death clustering, which is when everyone’s dying at the same places in the game.  There’s now a strong culture of user testing at studios including Foster City and Santa Monica, and that’s been a big part of enabling them to increase the quality and relevance of the titles they produce.

And what’s the future of the EyeToy?

The PS3 now has its own USB camera — the PlayStation Eye, which was released with a game called Eye of Judgment that recognizes playing cards placed on a surface. 

We’ve also continued releasing EyeToy games for the PS2.   Two of them make use of an evolution of the EyeToy color tracking technology that identifies both motion and color, which gives a much deeper level of control.  One is EyeToy: Play PomPom Party, which is a cheerleading game that comes with a set of pom poms, pink and green.  Along with that is EyeToy:Play Hero, which comes with a green sword that you use to fight your way through the game.  On screen the sword can light up, take on flames, and so forth. 

And we’ve just announced EyePet, which is generating some excitement.  You might want to take a look at the video.


No comments yet

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: