There's just no way to judge a controller without physically holding it and walking an FPS mile with it as your shoes.

I spent the early 90's going from one giant, gaudy "flight stick" to the next, never knowing that what I really wanted was a pair of miniature thumb-sticks. My mind is wide open to stuff like this.

Given 15 different controller designs, use logic to figure out what order you test them/build them. But at the end, it comes down to the feel.

I'm down to my last one that works perfectly, and considering building a USB one since all of the joystick parts are still manufactured and available from CH.

http://ec3.images-amazon.com/images/P/B000IBRN7O.01-ABT3ZC4T...

When you read into the details of these two pads they're developing, it sounds extremely exciting.

Some of the highest resolution and performant sensors ever put into a controller.

When combined with the supposed "configuration utility" whereby we have hopefully powerful control to map and bind to will, I can see this being a big win.

I like my 360 controller for PC gaming for many reasons: pain from mouse/kbb overuse being one, comfort when sitting and playing for another, and the simple fact that many games are made for controller and ported to kbb/mouse as well.

So you talk about an ergonomic, high quality, extremely performant, Linux native, configurable, built-for-PC controller?

Color me very interested.

[0]: http://smoothmouse.com/

    defaults write .GlobalPreferences com.apple.mouse.scaling -1
    defaults write .GlobalPreferences com.apple.trackpad.scaling -1

Touchpads (and mice) don't work like thumbsticks.

Touchpads/mice measure a vector with the size equal to distance traveled. Thumbsticks measure a vector with a size equal to the deflection from center.

Thumbsticks are terrible for mousing tasks (e.g., moving a virtual cursor [rts games]), and mice/touchpads are terrible for thumbstick tasks (e.g., playing a racing/flying game).

http://mobile.logitech.com/en-us/mice-pointers/trackballs

the thumbs trackball is more precise. But still not good. A good and cheap microsoft mouse is always with me when i'm playing FPSs.

Also, even if the control wasn't an issue, no trackball has more than 300DPI. EVER. The minimum for mouses is 400. I don't know why that is. But even the overly expensive Kensington ones are crappy even to click on a 10x10 button on regular usage, let alone to headshot in a FPS.

I have trouble believing that any device that relies on thumb-only input is going to be anywhere near as good as a mouse, but I hope I'm proven wrong.

I used a Logitech Trackman for seven years, and the biggest issue was that the force required to move the ball around wasn't constant; it would vary depending on how gunked up the little rollers inside had gotten. I had to clean mine out every day to keep it smooth.

My 1600DPI L-Trac disagrees.

The positioning on the Steampad helps, too. Imagine a D-pad or thumbstick built into your laptop in the place where a trackpad normally goes. That's what laptop trackpads are like.

It's a controller, it will never be as precise as a mouse. Trust me when this thing comes out people will be complaining about how they're suddenly losing every game of TF2.

I don't understand the problem with traditional mouse+kb for couch gaming. I've been using it for years and haven't encountered any drawbacks (using wireless mouse+kb combo). Am able to play online FPS and RTS, both of which require extreme precision and extreme low latency, at very high levels.

You use a solid mousemat, place that on the couch beside you. The keyboard goes in your lap. Kick your feet up and game away. It works great.

This controller looks like trying to reinvent the wheel, and I just don't think that it'll work. The problem which all of these schemes have is that they make you uncomptetive versus mouse+kb foes in multiplayer games. No one has ever been able to match the precision and control of mouse+kb, and even if this solution is far better than previously attempts, it will still likely not be anywhere near the traditional control.

For 'casual' single player games, it'll work great. I'm sure that's a large part of the market for SteamBox. But for the hardcore, it probably won't work at all. I think they'd be better off finding a solution which incorporates traditional mouse+kb than trying to work around it, but that's just my opinion from the outside looking in.

I've played console games with three other people scrunched up on a love seat; lying on the couch; sitting crosslegged on the floor; and sprawled out in a beanbag chair. Sometimes I get tired of having my hands in my lap, so I stretch them out behind my head. If I want to relocate, I can keep playing as I walk across the room, or pause and carry the controller in one hand and my drink in the other.

That's the sort of flexibility I want in a "couch controller," not just sitting still on the couch by myself.

"I’m a happy Steam customer happily using my happy mouse and keyboard. I don’t want a controller? You can’t make a sentence into a question by just putting a question-mark at the end. But we’re happy you’re happy, and by all means keep using whatever input method makes sense for you. Rest assured, we won’t abandon you. We love mice and keyboards, too."

And the problem here is not negligible. CoD series games will be miserable to play with a gamepad versus mouse+kb'ers, and those games would likely be quite popular on SteamBox.

"why not find a solution which incorporates mouse+kb instead of avoiding it?"

It sounds like they feel the same way to do. Mouse + Keyboard is already viable across a very wide spectrum of games. They don't need "a solution which incorporates mouse+kb", because that already is a solution (as you note, there's no real problem there).

However, for those that want a controller experience, there's not a viable solution. Controllers work across a pretty narrow spectrum of games. That is the problem that they're trying to address.

If you want/need to sell something in the millions, it kind of needs to be least common denominator. Gamers understand controllers as the basic interface for tv gaming, so they will buy them.

Gamers don't buy keyboards to put on their TV's so Valve is being smart by not forcing a square peg in a round hole. They want the Steam Machines to sell. A mouse and keyboard won't sell.

1. cultural expecations. consoles have controller gizmos. people expect it. its a different experience than just running an HDMI cable from your PC to a TV and calling it a day. Its a console. It lives next to your TV all the time. Its supposed to have a controller, because Nintendo's NES had one back in 1985. Cultural expectations and traditions are a powerful force.

2. there are many games that are actually superior to play with a controller than with a mouse and keyboard. Platformer games come to mind. Thats a major genre of consoles and always has been, and alot of indie game companies are exploring the genre again on the steam platform. At the moment I have to use an XBox controller plugged in to my PC to play. If Valve offered their own controller I'd gladly use it instead (assuming its good).

If you haven't tried before, I highly recommend doing time trials with KB+M. It's a completely different game.

I suppose my "complaint" is that they could do a better job of bridging the gap, but perhaps I am wrong. You do make a good point that the mouse+kb may be too cumbersome for some (most?) consoler switchers.

I do wish that I could see what sort of solution Valve could come up with if forced to keep the primary mouse+kb, though. Some sort of streamlined solution which you partly imagine in your post.

I always wondered if something like that would be possible.. So you'd have all your precision mouse movement without the akward balancing of loose things while sitting on the couch, or getting up and sitting down.

http://www.wired.com/gamelife/2007/12/phantom-keyboar/

Aaaand looks like the keyboard actually made it to market, there's even one on ebay. I had no idea.

hoverfield: http://www.correlatedmagnetics.com/products/hoverfield-non-c...

Someone did some tests a while back with a gamepad-trackball (trackball replacing the right thumbstick). They found that pro PC FPS gamers did just as well with it as with mouse+keyboard, after a short adjustment period.

Here's some quotes from the most recent attempt (out of three that I know of) to crowdfund one of those (http://www.kickstarter.com/projects/485185897/revolve-goes-w...):

"...I attained a level of accuracy I had never experienced without a mouse and keyboard combination." - Gerry Block, IGN

"You will spend less time fighting your controls and more time fighting bad guys." - Grant Collier, Co-Founder of Infinity Ward and Co-Creator of Call of Duty

Of course it failed anyway, because gamers all went "LOOKS DUMB LOL" and refused to read further. I was really hoping that Valve's controller would be a trackballpad; if it actually made it to market, I'm confident that ongoing positive reviews would lead to popular adoption.

It is a reason that console games come with built in auto aim on all FPS shooters...

But come now, my original statement was hardly all that bold. Obviously the reason that gamepad aiming sucks is just that thumbstick aiming sucks. Replace the thumbstick, solve the problem. (And as a bonus, get analog movement as well as analog aiming--the traditional WASD doesn't allow for speed gradations.) I used to game exclusively with a trackball and never had any troubles. The only reason I stopped is that nobody makes trackballs for left-handers anymore that are any damn good. And if you don't believe me, I just gave you a couple of quotes from gaming professionals.

As for Starcraft, you're right, it wouldn't work; not because of the pointing device, but because a gamepad doesn't have enough buttons to manage all the hotkeys. In my original post, I called out FPSes specifically for that reason.

Edit: Rereading your comment, I guess you might mean PC gamers who are sitting on their couches. I don't think that's the market this is after; it's trying to get console gamers onto Steam by providing them a cool new product that will work well with the games they have. This makes that a much smaller shift for those gamers than the keyboard and mouse. Will they still get owned by more traditional PC gamers (since they're now competing directly), as you suggest? I guess that's possible, but that's just saying "console gaming needs to stay separate", not "they should use a keyboard and mouse for console games", which just isn't on the table any more than PC gamers are going to switch to a handheld controller.

See: CoD, Battlefield, Quake, etc. All prefer exact aiming with a mouse over fudging analog input sticks with aim assist.

By your wording, no, console gamers don't use keyboards / mice, because the games they play on consoles are not compatible with them, and are not designed to use them. Very few genres (fighting, brawlers) sometimes prefer controllers to keyboards, and when they do it is usually a personal choice, or the market has better alternatives to both (ie, fight sticks).

Mouse is a pointing input - for a movement of 1 unit, your view changes by 1 unit. If you stop moving the mouse, your view stops moving.

Controller stick is a velocity input - for a movement of 1 unit, your view starts moving at a speed of 1 unit. If you bring the stick back to center, your view stops moving.

It could be that the Steam controller trackpad is precise enough to be a pointing input sensor, which could enable precise aiming. But it's too early to tell before the first reviews are out.

Joystick calibration utilities, for instance, tend to take stick position as a position input, and ISTR a number of PC sims that preferred joystick input but could use a mouse like a non-auto-centering joystick.

From what I can find online, the PS3 controller is 10-bit per axis (1024 positions on either axis), the Xbox 360 is either 8-bit (256 positions) or 16-bit (64k positions.)

A 25-way joystick of the type you describe wouldn't be usable for most console games, whether as a position or velocity input.

Much more important is the platform market share for hardcore games, like CoD and Battlefield. It simply isn't true that people prefer the PC version [1], despite the inferior controls. CoD sells vastly more copies for console than for PC [2]. It's not even close.

If Valve wants to break into that market and compete with Xbox and PS, it needs a controller.

[1] Though I do, and it sounds like you do too.

[2] http://www.gamasutra.com/view/news/31995/Exclusive_Black_Ops... (I recognize this doesn't include Steam sales, but the point still stands)

Who buys retail games anymore?

Console will always be at disadvantage to the PC in this genre due to its low frame rate (usually 121+ fps is considered slandered for the genre in order to make use of a 120hz refresh rate). Consoles currently output about 20 - 30 fps, not really acceptable for competitive play in a FPS.

It'll be interesting to see how well they manage to deal with this.

Keeping thumbs oriented seems to be the biggest challenge of a controller like this, but I as hoping the haptic feedback would help keep your thumbs in place.

The analog trackpad does have an advantage over a keyboard but not over a mouse, i think.

With two analog sticks/trackpads/mice/etc. you can get the same range of control over both. That's what the two "planes" give you.

In most FPSes your move commands are binary, "W" or nothing. You can alter them a bit with combinations, Shift-W to sprint or perhaps Ctrl-W to crawl but you're still limited to a handful of preset speeds.

I could easily imagine being able to use a continuous axis to encapsulate walk/jog/sprint being quite a bit more intuitive and precise.

Maybe its just all of us aren't used to doing those tasks with a controller as we are used to it by a mouse (btw long time halo veteran so am more comfortable on controllers for FPS games) but i think a learning curve would prevent high level players from switching to controllers becuase it would (albeit temporarily) really shunt their skills