Gamescom 2010: Kinect Interview - Kudo Tsunoda Talks Core Games, Response Times and the Future
Written Sunday, August 29, 2010 By Dan WebbView author's profile
There were two things on my mind when we walked in to interview Kudo Tsunoda, Kinect's Creative Director, at this year's Gamescom. 1. "Why does he always wear those glasses?" and 2. "Hmmm... Kinect, where do I start?"
I'm being serious here, there was so much confusion surrounding Kinect before we went in, that we knew that our 30 minutes with Kudo would throw up awkward question after awkward question. If we - the press - don't ask them, who will?
So sit back, grab a drink... and probably a pillow, and read the monstrous interview with Kinect's Creative Director and find out all the answers to questions like, just how many active players work with Kinect? Is a response time of 1:1 really possible with the device? What's with the dumbing down of the device? Where's the core gamer love? Find out all those answers and more below.
I don't think any of the games you just mentioned can be done on any other platform. Like Kinect Sports, they've got advanced football, so why are we not seeing football in any other kind of motion sports game? It's because they're not using full body motion, so you can't really make this kind of football game without being able to move your feet.
I think what the Kinect launch line-up does really well, is highlight the unique experiences that Kinect can do. I mean it's a lot different, being able to get into Kinectimals and sit there petting your pet tiger, being able to use, not only voice recognition, but having the animal respond to the tone of your voice and the way that you're communicating, then being able to do a lot of the full body tricks with your animal, like laying down on the ground and having your tiger play dead. That stuff can't be done at all in something like Nintendogs.
I think people who haven't played the experiences before would think that somehow they're similar, but it's really the unique stuff about Kinect that makes them very different.
I think it's interesting, because when you say core gamers, there isn't really an official definition of what 'core gamer' means, right? So, core gamers must be people who perhaps like to kill things – I'm gonna shoot something or saw it in half - and that's what makes a core gamer. I've been playing games since I was a little kid and so if I was to classify me, I'd be a core gamer. I think to me what a core gamers like is experiences that have very skill-based gameplay or games with a lot of depth, so the more they play the game, the better they become at it over time.
I think what core gamers hate is when they've been playing a game for two months and then someone comes in who's never played the game before and is able to beat them at the game – that's wrong, that's bad, because then obviously the game doesn't involve any skill. Even if you had a shooter where you played for three months and you and I went into an online match, and I could just beat you the first time I played; core gamers would hate that as well. And I don't think it's about the shooting or chopping people in half, it's about having the skill-based gameplay so you get better at the game over time. That's the kind of thing we've really focused on for the Kinect launch line-up – providing games where the longer you play, the better you get at it, there's always something new to learn and there are a variety of ways you can play.
I've been working with Kinect for a long time, so I've played the games for a while, and if you fired up any of the Kinect launch games and you just get in and beat me at the game, then okay, I'd agree with you and say that this isn't anything core gamers will like. If someone can come in and just start frantically flailing about and they start beating me even though I've been playing the game for a long time, well, that's just not how the experiences are. The longer you play, the better you get at them, and I think that's the kind of mechanic that gamers like.
Everyone talks about dividing players into core gamers and casual gamers, or whatever term you want to use, but it's funny that in my family, not a lot of them play video games, but they'll play cards or Boggle and that kind of thing. Now those kinds of games are very skill-based things too and the longer you play them, the better you get. That's a common thread that exists between all people, we all want to feel like the more you put in, the more you're progressing and the more you get out of it.
That's something we've really focused on in all the Kinect launch experiences and that's not just Joy Ride or the Sonic game. It's funny you mention Sonic, because would you say that previous versions of Sonic are a core game or a casual game?
But even the old-school Sonic games, you'd consider those to be core games, right? Even if we were delivering an experience with Kinect where you'd never heard of Sonic before and we told you about this game that's a little blue hedgehog running around collecting gold rings, then you'd say where are the core games? Well that totally was a core game, and I think that because the interface is so simple and you don't have a lot of complicated controls to learn, it does provide a deep gameplay experience that requires skill.
When we talk about Kinect being accessible, it's not about making it casual so that a six-year old kid can get into it and beat me at any of the Kinect games. It's about making it so that anyone can play without having to learn any controls and get that skill-based gameplay that I think everybody enjoys.
Kinect has a very carefully tailored launch line-up, but one thing we haven't seen yet, is third-parties or indeed Microsoft incorporating Kinect functionality into their existing franchises. Looking at PlayStation Move support: EA is adding it to Tiger Woods, Ubisoft is adding it to RUSE, it'll be in Killzone, SOCOM – all big franchises. Yet Halo: Reach and Fable III are shipping without Kinect support. Is that because it's difficult to implement Kinect into those games or is there a reluctance to add Kinect to those games?
I don't think it's any of those things, as the way you're going to make a super-awesome experience is not by taking something that exists for one type of controller and just porting it over to another controller. If you think about the way first-person shooters evolved, they started on PC and everyone played them on PC with the keyboard and mouse interface. When shooters started getting ported over to consoles, they would just take the existing build for a PC interface and swap it over to a console controller, they wouldn't play very well and so everyone would say, oh look, you can't make a first-person shooter on a console, it's not made for that type of controller, which is like what you're saying now.
The Halo team, instead of taking another controller for an existing device and pushing it over to a new interface, made a first-person shooter exclusively for that device, from the ground up. And the game was awesome, so now everybody plays them on controllers and hardly anybody plays first-person shooters on the PC any more. But everybody said the same thing, that first-person shooters will never work on consoles, it's only going to work on a PC, it's not responsive enough, there'll be all these problems and blah, blah, blah. If you build something specifically for the device and take advantage of what the device is, then you can make an awesome experience.
Now, Bungie does an awesome job of making Halo, and I'm not going to be one to tell them how to make a Halo game, because they do it a lot better than I would. But I would personally think that, Halo has been designed and built over a long period of time, specifically to be played on a controller, so to try and just cram Kinect into it and still call it Halo, you're going to have that same thing where shooters went from the PC to the consoles. That's not saying that it can't be done or that it's not going to be awesome, and like I said, hardly anybody plays shooters on the PC any more. It's all about the console now and that's the stuff we really focus on with Kinect. It's all very new and innovative technology and the best experiences that you're going to get will be ones made specifically for that, versus taking something and just pushing it over to Kinect.
Have you seen any prototypes for first-person shooters on Kinect? Is there anything you've seen that might take that genre, for instance, and advance it?
Obviously, we're very focused on talking about the Kinect launch line-up and we're not really talking about things that haven't been announced, here at Gamescom. Hypothetically, just as a design person, if I was going to try and think of a way of doing a first-person shooter, you could get up and hold a gun up with your fingers, aim by pointing around, move by actually moving in and out with the gun as you're aiming, getting strafing on your shoulders, getting head-tracking with your head, using your front hand to shoot the gun and throwing grenades by performing an overarm action, that would be one way you could do a first-person shooter, which I just made up on the spot.
So, have you seen anything like that?
It's not something that we're talking about right now. We're just very focused on the launch line-up. We showed some other new games at E3 this year, like the new Star Wars game, the Forza Kinect game, but the stuff we're talking about now is already announced titles.
Is Kinect accurate and sophisticated enough for that kind of play? We believe that it once had twice as many tracking points, and patents suggest that it was once able to read sign language and so on. What concessions have you had to push down the price point and turn several thousand pounds worth of motion-camera tech into a more affordable Microsoft product?
Let's just take the tracking points for example. When you start developing something – and obviously Kinect is a new technology – you don't necessarily understand what it takes to track a human body and it's a lot easier when you're building something to start with tracking all of these different points and as you start building the stuff, you're like well, to track everything we need for the human body, we don't really need all of these points.
It's not that the technology can't track as many points or that it has anything to do with cost, it's just that any of the experiences right now, you can just get in and anything you move with your body, it instantly does on-screen. So, there's nothing in any of the experiences that we've compromised because of the number of points and it certainly has nothing to do with cost. If you don't need to track so many points and you have the same fidelity of response in the human motion and translating that into the game, there's no reason to carry all the extra points and be processing them, so instead you can devote that to getting all of the other things that you want to do with the games.
I think that's just normal game development and with any game, you want to use the processing power as efficiently as possible, so you use the exact amount you need to get the kind of fidelity and experience that you want.
Kinect had its own processor at one time. Surely dropping that was a cost decision?
Well, the stuff you try and figure out is, how are you going to be able to deliver everything you want in the experiences? So at some point, you've got to be like, okay, we know how much processing Kinect is going to require when we start off development and obviously you don't want to lose any of the things that are important to Xbox customers. I mean graphical fidelity is something that Xbox has always been known for, and you want to make sure that you still hit that level of graphical fidelity and the speed at which the game runs. You want to make sure that still works well and what we found is that the overhead and processing power that it takes to run Kinect is small enough that you're not losing any of those things.
And so we showed one of the best games from Microsoft that's really known for awesome graphics and running at 60 frames-per-second, and that's Forza, right? Forza is an amazing graphical show-piece that shows the importance of the responsiveness and running at a high frame rate. So we had Forza with Kinect running at E3, and the graphical fidelity was actually improved in some areas; it's still running at 60 frames-per-second and so there's just no need to have that extra stuff.
It's never a question of compromising and having a worse experience, because we're trying to remove processing from the sensor, it's just when you start out, you don't really understand how much processing Kinect is going to take or the optimisation you're going to get out. We're always erring on the side of providing the best experience for the user and as we get the stuff implemented, all of a sudden we have something like Forza running at the awesome level of graphical quality that you want, running at 60 frames-per-second using Kinect, you just don't need that stuff.
There's a lot of confusion about how many active players can Kinect track. How many players can it actually track?
Again, we just try and optimise everything around creating the best possible experience for everyone, like track and field for instance, you end up being in split-screen because as people get farther away in the race, you can't keep them all together in one area. So at some point, having more than two people doing that, cutting the screen up into really small slices makes it a very bad experience. Dance Central though, has lots of people dancing, involved in the game and they're able to keep everyone within one screen space.
So what you're saying is, that for the launch line-up, the focus is on two player games, but in the future, it's possible to track more than two active players?
The distance issue is still an area of confusion too. Some are saying it's about 2 metres, whereas the guidelines say 1.7-3 metres.
Since the sensor can see the whole room in 3D, something like Kinect Adventures will be able to customise the game experience based upon where you're standing in your living room. If you're standing at 1.5 metres or so, it'll customise the experience for being in a smaller room and if you're back 3-3.5 metres, then it'll change that gameplay experience as well, based on your living room and how much space you have. Obviously playing something like football might not be ideal at certain distances [where you can't see the feet], but you can play pretty much anywhere, whether it's side to side or whatever within that kind of space.
One of the biggest core games to be shown for Kinect at E3 was Ubisoft's Child of Eden, which seemed to be right for the audience, so it had a big response. Do you think it was a mistake for Microsoft not to focus on that kind of core game at E3 and what do you think it'll take to win those kind of gamers at E3 over to your side?
Say I was to make a shooter right now, and I'll call it, I don't know... 'Shooter Dooter'. Not the best name, I know. But what do you think it would take to get people at E3 excited about that?
Put a controller in their hands.
Right, assuming there's a controller in your hands, what is it going to take to make people want to go play the game and go out and buy it? What's going to make them think it's a good game?
A compelling story? The type of thing that made Halo successful. The things you were talking about earlier: great gameplay, increasing your skills over time.
That's right. Great gameplay and increasing skill over time. The only way you can really get any of that is not by seeing it in a video online, but by playing it. That's how you know a game has great gameplay, because you actually play it. And that's what I just say when I come to any of these shows, is those Kinect experiences do have that kind of great, skill-based gameplay, there is depth of gameplay in there and so I think all it takes for any of those things to be demonstrated is for people to actually play it.
It's harder when you just look at anything online and there's a reason this is an interactive industry. It's because you need to interact with it and so the thing I say to everybody is, I totally understand everybody's going to have their opinions and that's all very good, but once people get in and play it and try the games, that's how you can see whether they're fun and have great gameplay, skill-based gameplay and depth of gameplay.
Those are the things I feel very good about when it comes to Kinect, is getting people in and playing the games, and the reactions are always great. Whether it's at Gamescom and the trade shows, whether it's at E3 in the booths, people who play games or don't play games, once they get in and try the stuff, they have a really awesome experience. The same goes for any game, whether it's Kinect or controller-based, when people play, that's what gets them excited.
Some of the biggest reactions to Kinect have been from games that are immediately recognisable to gamers, like Forza for instance, which has since become a big talking point. Everyone expects Dance Central will be huge. When are we going to see more of these sort of recognisable games that will appeal more to existing Xbox gamers?
Again, that's what I'm trying to say. What I think core or Xbox gamers like is what you've already said: great gameplay, skill-based gameplay and those are already in the titles. If you're asking me when I think there's going to be shooters or games where you're chopping people in half or whatever with Kinect, I mean obviously third-party developers are working on their own stuff and so me as a first-party will be managing what third-party people are doing.
But again, I would just say to anybody, we announced a Star Wars game at E3, so certainly that one is coming and you saw Forza and Dance Central, like you said and these are all things that give us a good mix of content. There really are a number of enjoyable experiences for different kinds of people and with all the games that are in the Kinect line-up, there's a good mix. Some are from existing genres, some are new types of experiences and it's easy if you haven't played before to see something that you know already, and extrapolate in your mind what it'll be, as you've already had experience with that. But it all comes back to, if you can go in and play – and if you watch people playing at the pods, they're having a super-fun time – and that's what motivates and convinces people that it's a good experience.
The response time seems like it could still be an issue. Is true 1:1 tracking actually possible with Kinect?
Would it be something that could be fixed with updates anyway?
It's not something that we're fixing. We've done tests back at the office where we have a TV, flash a light on the screen and we'll have a controller, and have someone there hitting a button, and someone with Kinect clapping as soon as they see the light on the screen, and both things go just as fast as one another. Whether you're using a controller or making a gesture, it doesn't make a difference. It's all human response time and so it's not something we're trying to fix with software updates or anything.
Look at Xbox Live, it's a lot different when it first came out to what it is today. It's evolved over time with a lot of new features and a lot of new experiences, and so when you talk about software updates, that's the kind of stuff you'll see with Kinect. Lots of new features coming online, new experiences we can give people without them having to go and buy a whole new console and we already have the 1:1 stuff working really well and the response time is just as fast as pushing a button or making a gesture. Software updates will be used to implement new features and deliver new experiences.
So, 1:1 is already there with Kinect, it's just up to the developers to calibrate their games to get the most out of it?
Yeah, and you can already see in like Kinect Adventures and Kinect Sports that they're doing a lot of stuff with 1:1 tracking. Kinect Joy Ride obviously needs to be very responsive, because it's a high-speed car game and those things are working really well already.
Kinect is scheduled for a November 4th release in North America. Everywhere else will have to wait for a release date.