Red Steel

2006.1016
Lead Game Designer Roman Campos-Oriola talks about Red Steel's Wii development cycle

Red Steel is looking like one of Wii's biggest launch games. Ubisoft hopes to push the boundaries of the genre with a new free-pointer control style and fancy sword-fighting mechanics, all operated using the motion sensing powers of the Wii Remote.

After a lengthy hands-on session with a near-final version, CVG grabbed Lead Game Designer, Roman Campos-Oriola, for a chat about the challenges of developing a totally new control system for a totally new piece of hardware. Enjoy.

Red Steel has come on a long way since E3. What have been the most significant improvements made to the game since then?

Campus-Oriola: E3 was the first time we showed the game to people outside of the development team - before E3 we couldn't even show the game to other Ubisoft employees. So that was the first time that we got any feedback at all, and after that the main thing we tweaked was the control. We aimed to make the controls tighter, really focusing on the detection of movement with the Wii Remote's pointer.

Since E3 all of the motion detection has changed, so now all of the sword actions are much easier to pull off. The hardware has also changed since E3. We were running it on Revolution dev kits back then. With the new hardware, we were able to tweak the graphics.

The bulk of the work was done on creating a better lighting system. We re-worked the lighting on the backgrounds and, more significantly, on character models. One thing that wasn't so good at E3 was the character models - when you spoke to them they looked almost like 2D characters and when you stood far away you couldn't see them at all, so we improved the lighting on them to fix this.

When Nintendo first approached you with the Wii controller, what were your initial thoughts, impressions and ideas?

Campus-Oriola: The first time we heard about the controller was from a Nintendo correspondent at E3 last year, who detailed the main motion-sensitive concept for the controller. He also told us that Nintendo wanted Ubisoft to make a first-person-shooter for the new console and its unique control system.

We spent the next three months thinking about what type of FPS control mechanics we could achieve with this controller - things like how you'd throw a grenade and adding swordplay. After our lengthy brainstorm, we decided to focus on four concepts and present those to Nintendo.

So we showed them how we wanted to use their new controller interface for an FPS. With direct movement interaction, this FPS would, for the first time in the genre, allow the player to aim anywhere on the screen, and we showed them the sword gameplay.

That demonstration took place in July last year in Kyoto, Japan. After showing Nintendo what we wanted to do, we were lead to another room in its Kyoto office where there were 13 different prototype mini-games to show off the controller. Then they showed us the Wii Remote, and the mini-games which each demonstrated different functions of the Remote, excluding the Remote speaker of course, because there was no speaker at that time.

We were really amazed by what we saw, and at that time we didn't even think about the technical power of the console. We we're too tied up in thoughts about what may or may not be possible with the controller. When we were eventually given the specs for the machine we realised that it wasn't that powerful, but we didn't care because of the controller. So our first ideas we're not at all focused on power, but on FPS play mechanics.

Red Steel demonstrates the fruits of your efforts, with loads of really cool and innovative play mechanics...

Campus-Oriola: Yeah, like the grenades for example. One thing that I find so frustrating with FPS games is that when you want to throw a grenade just two metres away, for example, just through a door, you always have to look down at your feet and throw the grenade downwards, often with excessive power. So in Red Steel, for the first time, you can open a door, make a rolling motion with your hand and just roll the grenade towards your nearby target.

Did you come across any problems with developing all these mechanics? Were there any ideas that you had to abandon?

Campus-Oriola: The main obstacle that we experienced with development of Red Steel was with the motion sensing functions. No-one moves in the same way. For example, if you ask someone to slash with a sword one guy will do it one way and other guy will do it differently.

In the beginning, we had the game only recognising a small range of movement - to perform a sword slash you had to move in a very particular way or the game wouldn't recognise it. That was the biggest mistake we made at first. In fact, operating the motion detection was the biggest challenge for the player, so we had to change that because the motion detection should not be a challenge to operate.

Talking of controls, Red Steel uses the analogue stick to control the player's movement, and the pointer to aim your gun and turn the view, as does Metroid Prime. That system, although strange at first, works well, but we've had huge debates on internet forums about what alternative FPS control systems might be possible on Wii. Do you test any alternative control systems?

Campus-Oriola: In the first three months of development we tried many different types of control mapping, pointer recognition and ways to move your character. We had discussions about this with the Metroid Prime 3 development team at Retro and came up with this system. I'm not saying that we have the best system possible, but I can say that, from what we tested, this system seems to work best.

So what other control styles did you test?

Campus-Oriola: In the beginning the idea was to keep the cursor locked to the middle of the screen. I don't think that is it better to have the pointer locked to the centre of the screen on Wii because, firstly, when you play FPS games with a mouse, when you reach the edge of the mouse mat, you lift the mouse and reposition it in the centre. With the Wii controller, you can't do that.

We tried to do something similar to a gyroscopic mouse (a mouse that you operate with motions in free space, similar to the Wii Remote). With this mouse, you simulate the 'lifting' of the mouse by releasing a trigger button on it, re-centre your position, then press the trigger again to continue operation. Before we even had a prototype of the Wii Remote, we constructed a prototype of Red Steel using a gyroscopic mouse.

But it doesn't work so well because when you're playing the game you have to concentrate on quickly releasing and re-pressing that trigger button, and it's not really intuitive. It's really frustrating when I read all the debates that take place on internet forums because we can't get on and tell them WE TRIED IT - IT'S NO GOOD!

Everything that's discussed - we've play-tested it. Nintendo has too; one prototype we played at Nintendo HQ in Kyoto was also like this and they didn't like it.

What about one-to-one direct motion detection on the sword-play; why did you choose not to use direct motion sensing like Wii Baseball?

Campus-Oriola: Perfect one-to-one motion recognition can't be done. What's interesting about Wii Baseball is that you assume the exact position of the character on the screen, so it feels like one-to-one motion detection. But if you hold the pad down like this *demonstrates holding the Wii Remote low and upside-down* and you flick it slightly you will see your character take a normal swing - not real one-to-one detection. But what's really important is not whether it's one-to-one or not, but that the player has the feeling of his motions on screen.

Another reason why we're not doing one-to-one motion with the swordplay is that holding a Wii Remote in your hand does not give the feedback of a normal samurai sword - the weight, the rebound of clashing blades.

We used animations to simulate the weight of the sword and the effects of your blade hitting something. Also, if you give someone the Remote and they will typically hold it at arm's length and flail around. That's not good, so that's why we used animation.