The Pitfalls of Motion Control *Spin-Off

You are correct, patsu. Perhaps I was too hasty.

I have personally seen people struggle with motion control. Is it inneptitude or preferance? Doesn't really matter. Motion control, especially when it's not an option, can seriously alienate potential players.

 

I think one of the best examples of how the sixaxis could make a game better would be a version of fight night where you control the upper body with the tilt axis. I've discussed this idea before. That you never keep the controller 100% still would only make it look more natural. And then you can bend forward, bend backward, lean to the left, lean to the right, turn your right shoulder forward, turn your left shoulder forward, etc. You don't need extreme motions, but they can make a big difference to the game, and you don't need to change any of the other controls.

That would be a good example of how the tilt can add something to the gameplay that you would otherwise never be able to do. And I think it could work for just about every fighting game.

Similarly, I think motion control could work great for looking around and even aiming in FPS games. I think what you need to do is have a button that you press to activate the looking around/aiming, and let it go to recenter automatically.

 

And the mouse comes with its own line of disadvantages. I don't want to use a kb and mouse in my console setting as it would create problems for where I would use the mouse on my recliner and I'd have to sit the kb on my lap. Some of the same issues created by motion control. If I want to lean back and relax, it creates issues where sixaxis and wiimote won't work because I may not have the freedom to move my arms as required. And I have to make these sacrifices for what? A gimmicky way of achieving the same thing they can do with dual analog? There may be some scenarios where motion control would be more intuitive or work better on the whole, but the lack of precision that seems to be offered at the moment makes those even more limited/questionable.

I don't think the devices are good enough to replace all aspects of the pads just yet for the vast majority of game types. At this point, to me, it seems like shipping a steering wheel with a console and trying to adapt every game to make it work, sure its great for racing games, but after that you're just wishing you had a gamepad.

 

Also, and sorry for this being in a new post - blame the PS3's browser - in my experience sixaxis is for many things much more precise than an analog stick. A good example that shows this off is probably a game like Mercury for Wii (See the Eurogamer preview). Also, think of games like GT, where many never adjusted to the analog stick and actually still use the (heavily assisted) gamepad. Tilt steering may just end up being better, especially for driving slower cars where precise control over your curve leads to additional tenths of seconds being gained.

 

I don't think lack of precision is it. It's a combination of game design, perception/acceptance, effectiveness (where precision is one of the key factors) and audience/market size. I also disagree that motion control is inferior to analog stick in general.

Mainstream console players feel right at home with analog sticks since the latter has more than 10 years to draw in these people (and well-tested content !).

As you mentioned, a mouse is more precise but less applicable in the living room ...unless you are a FPS player "trained" by over 10 years of mouse usage. As it stands, most people claim that the mouse plays better than analog sticks for FPS.

I think a mouse is also better for manipulation games like RTS, Sudoku, LittleBigPlanet Builder than Dualshock. A Sudoku fan refused to play the game on my PS3 because of the analog sticks. She'd rather play it on _paper_ or a PC. A coffee table mouse would work for her though.

Casuals find motion control more attractive as proven by Wii. Now it will need the same 10 year cycle to attract its own crowd and titles. If Wii dominates this generation for many years, it's a little hard to argue which is more mainstream to the consumers. Where SIXAXIS is concerned, I enjoyed flying using SIXAXIS, the feeling is a little different from analog stick control.

My view is similar to RancidLunchmeat in the sense that different controls are for different situations/segments. But I disagree that a motion control device is necessarily (always) less precise. Even if Lair is controlled by analog sticks, it will continue to have a large turning radius when you don't slow down. 180 degree turn may work better but only 1-2 times across 10 tries (not a major break). People simply want their familiar controls back.

This I agree.

 

For the most part I've really enjoyed motion controls on the Wii and feel it's brought a new,more intuitive and immersive feel to playing games.
I feel they should be the future of gaming control. I just hope they continue to refine them and hope developers become more consistent in their implementation.

 

I'm wondering how much of it is longtime gamers being used to being really good at game control and feeling like frustrated newbs. You should watch someone who doesn't play video games try to play Wind Waker or something sometime and truly wrestle with the analog stick.

Maybe once motion controls are normative and we've all been playing with them for 5 years, they'll be like second nature the way dual thumbsticks are now.

 

I'm on your team, patsu. You heard it here first.

go. team.



edit below:

I wrote about my experience with the new firmware, and it's improvements to motion control, in the 1.92 update thread.

Long story short. My experience with motion control in Formula One was that theimplemtation was very good, but that the actual motion controlling hardware just could not quite deliver the goods for a game as harsh* as F1.

After the update, F1 is amazing with it. I like it much better than the stick, which I've been using since Gran Turismo 1 came out.

So perhaps there is more than meets the eye with the motion control systems. Something in the PS3's new firmware makes a subtle, but significant difference in the way the system maps the signal from the controller. It's just a little more sharp and predictable. More like buttons. And it seems to lose nothing in the upgrade.

Lair has spawned alot of talk about motion control and I'm glad to be in the discussion here.

go. b3d.

*Seriously, the game is nerve-wracking, demanding, rewarding and visceral. It's harsh.

 

If I were Sony, I'd release a Linux framework with a renderer and motion detection, and let the homebrew community explore possibilities. They'd find what works and what doesn't for the fun of it, and do it for free!

 

I agree. Or just release a Windows sixaxis driver. After all, why not?

 

I agree. IMHO, a good SIXAXIS control scheme is as important as solutions like EDGE and PSSG.

Does the UT3 mods only allow visual changes ? Can UT3 take control mods ?

 

Hate to be the one to bring it up, but... could the problems with Sixaxis be related to how it seemed to be an afterthought, introduced so close to the release and thus the finishing of the PS3 hardware? IIRC the legal problems about the rumble feature seemed to go on forever and there was a time when they could no longer hope to resolve it and include the feature on the final PS3, so they had to come up with something new. Then they've seen the reception and interest in the Wiimote and decided to go for motion control too. But they haven't had the time to properly complete the hw design, and that's why it has problems... Could this be the case here?

 

Well, I think I agree with you except for one aspect of it. The hardware isn't the issue here. On the other hand, the motion control SDK was only released just over a year ago. But having played with motion controls extensively, I'm pretty confident that the hardware seems to be more than fine. However, with the FW 1.92 update increasing the quality of the response to the motion controls a bit further, that seems to indicate there were some improvements possible in software.

 

I think the old paradigm of inputs triggering animations doesn't work with gestures. You need something different. If you look closely at what games do, outside of using a button to fire, there actually is a delay between the button press and the execution. Mario doesn't do a 180 instantaneously, the Prince isn't in the air the moment you hit the button, and Samus doesn't instantly turn into a morph ball. But this is exacerbated with gestures, because it takes longer to input a gesture than a button press, so now you're not just delayed by the animation, but by the time it takes to input.

I think the connection between input and animation needs to be reworked with Wii games. I have some ideas about how things could be improved, although I don't have the means to implement them. For example, doing a throwing motion and then watching a throwing animation to throw a grenade just doesn't work. It feels awkward, and it's hard to learn. What you could do instead is tie the playback speed of the canned throwing animation to the velocity of the controller is moving (you could even make it really simple by tying the speed to the velocity component of only 1 axis), and have speed_grenade = speed_final. That would minimize delay a great deal and improve the recognition of the input, since now you're not trying to do any gesture recognition. Further, I think it would make the action more learnable, since the player will see and feel near-instantaneous feedback to his actions. And since the feedback is constrained to a predefined animation, you'd get rid of that feeling of frustration players have when their gestures aren't recognized. The animation would just be moving forward (possibly backward), either more quickly or more slowly.

Maybe this specific example would need a good bit of work, but I think this sort of thinking, which is basically what Wii Golf does, will work better than gesture recognition.

 

That is why I suggested that if you want any gesture based thing, maybe it is better to indicate first with a button that you are going to do a gesture, so that then any motion from there on can be listened to precisely and instantly. Just one button could do the trick.

Also, where possible, you should tie the motion to the animation. Say that you're going to do a throw using a gesture, then move the arm backwards and then forwards for instance. Just an example. The combination could do the trick.

 

The R&D has been going on for a while based on some early Sony patent. Other parts of PS3 like the Blu-ray drive was completed even later. So they should have ample time to deliver a good SIXAXIS hardware.

The law suit may have triggered the motion control plan, but I doubt they based it on Wii from the very beginning. They had to design it with power consumption and dimension in mind. They also added USB-pairing (instead of Bluetooh pairing) for security and simplicity/robustness. The controller engineers had to make sure it can be mass produced reliably and cheaply too.

However they may not have enough time, resources and corporate culture to design a comprehensive UI guideline and middleware for using SIXAXIS though. There were/are just too many things going in parallel. It seems that Sony only aimed for ease-of-integration based on the Incognito's "It only took us 1-2 weeks to put it in" comment.

Based on the game reviews, I'd say it has more to do with issues like user acceptance, game design and implementations, plus isolated manufacturing issues.

It would be familiar to mouse users but may make the motion sensing redundant though. Ideally, the vendors should follow fearsomepirate's method (i.e., The in-game events and interactions are performed as part of the motion/gesture .... rather than after the motion). That way, there is no need to coordinate button presses with large movements. Just swing away.

 

Folklore has yanks, balance, and shakes
there is far more implementation than a simple yank

Heavenly Sword also has an implimentation of yanking the controller to jump
I also dislike it because the success rate is 30% for me. (at least in the demo)
I would much rather revert it to using X as the control.

Well, so far the sixaxis implementations aren't out of the world but many are good, practical applications. At least in my opinion.

 

Yes. The idea of motion in PS's controller was an old one. Perhaps the technology wasn't as good as Sony hoped for but felt they needed to put it in now, to either compensate lack of rumble or compete with Nintendo, or both. But if they put it in, they must have felt it had sufficient accuracy to be usable. Blowing money on a useless gimmick seem implausible to me. Certainly Nintendo were relying on the same principles and they bet the entire farm on it.

Also other examples haven't had problems, so it seems a software issue rather than hardware issue. Have people seen the homebrew application of the guy controlling the robot with sixaxis? It shows good accuracy, with no obvious drift or the like.


It's a bit twitchy, but I put that down to no filtering on the input. Anyhow, the point is apparently not every motion detection application is dodgy (haven't used sixaxis myself so I'm relying on other feedback!). It could be some gestures the hardware can't pick up, but there's nothing in the hardware that's preventing accurate motion gameplay if you work to it's limits.

 

In comparing a motion control to an analog control, it would not be that hard to generate statistical data that demonstrated the inferiority of the motion control in accurately and consistently gathering user input.

A button press failing? Unlikely. A gesture not properly registering with a sensor bar, or a dozen other minor issues interfering with accurate readings? Much more likely. Throw in the simple physical problem of it being easier to repeatedly press a button than mimic a gesture, and the problem is obvious.

A big part of the problem is simply a lack of developer discrimination - motion control doesn't make sense in many circumstances - controlling a dragon isn't a natural motion, it's not something that will be intuitive for me, and as a hardcore gamer (the audience Lair seems to target) I'd feel far more comfortable using a joystick in a flight simulation than tilting my hands.

 

The point of motion controls to me seems to be, at least in part, that it isn't right 100% of the time right away. Instead of likening it to a game that presses button, it is more akin to a sport. You wouldn't expect to play well at tennis picking up the racquet the first time. You wouldn't expect wielding a lacrosse crosse to be as easy to master as pressing a button. Sports offer a challenge in learning to control your motion. Motion controls must inherently offer the same challenge. The problem is probably one of developers understanding this and players understanding this. Existing gamers expect it to be as quick and accurate as button-presses without regards for physical limits that they'd happily expect in other physical games.