Kinect Is Discontinued: Why Did the Device Built to Free Players' Hands Fail?

For players, what matters most is the thrill a game delivers. If losing the controller doesn't add to that thrill, it is pointless — but why, exactly, didn't it?

Kinect was the motion-sensing device Microsoft released alongside the Xbox 360.

It was once thought capable of freeing players' hands so that everyone could enjoy games; Microsoft saw it as a breakthrough technology that would change how games were played, with huge potential in other fields of technology too. Its sales at the time were astonishing — Guinness World Records listed it as the "fastest-selling consumer electronics device." Yet seven years later, on October 26, when Microsoft announced that Kinect was being discontinued, no one was surprised and no one mourned the technology. Instead they went and left comments under ancient Kinect promo videos: "I can't believe people actually fell for this back then?"

The verdict sounds like hindsight, but Microsoft really did make a mistake: it assumed players were desperate to be rid of the controller, and that its own mission was to free their hands.

There was also an unspoken premise: throw away the controller, that symbol of the barrier to entry, and more people would become gamers.

Seven years on, the Switch did free players' hands — but it did so more cleverly, still letting them hold a familiar stick. Precise button input will always suit core gamers better than waving your bare hands around. Sony, meanwhile, has been testing another route into console gaming for the masses: turn the smartphone everyone already carries into a "controller" with an app, so a living-room gathering can play together.

Kinect, for its part, ended up as a piece of trial and error.

It Began With a Sensor That Turned Sound and Motion Into Magic... but There Was No Steering Wheel

Project Natal came out of nowhere.

2008 and 2009 were the Xbox 360's peak years. Three years after launch, its sales had just been nudged past by the later-arriving Nintendo Wii, but with solid hardware and excellent games the 360 performed steadily throughout. Looking back, many critics are convinced it was the best console of that generation.

Over a console's fairly long life cycle, though, there are always experimental upgrades and add-on peripherals meant to inject some novelty, hold the attention of new and old players alike, and lift later sales. At E3 2009, the "upgrade" Microsoft brought was a project code-named Natal: "A revolutionary new way to play: no controller required. If you know how to wave your arms, shake your hips or talk, you and your friends can jump right in — life experience is the only experience you need."

This became Kinect, a blend of "kinetics" and "connection." It was an Xbox 360 peripheral that used facial and motion recognition to let players drop the controller entirely, working the console's interface with voice and gesture. All you had to do was move.

No buttons, no sticks, so nothing complicated to operate. In Microsoft's vision, once those obstacles were gone, people with no controller experience and the all-thumbs crowd who were bad at games could play as freely as core gamers. That was the point Microsoft hammered on again and again at the E3 keynote.

"The controller is a barrier," said Don Mattrick, then president of Microsoft's Interactive Entertainment Business. "It draws a line between gamers and everyone else... Can we break through the controller? Can we make you, the person, the controller? We can!"

In the keynote and the demos that followed, players really did work as the "controller": they waved their arms through the air pretending to swing a lightsaber and summon the Force; they hopped back and forth in front of the TV pretending to skateboard, and ran in place pretending to race; in a demo code-named Project Milo, they pulled faces at a virtual character as if talking to a real person, and put out an arm as if catching an object thrown to them. Kinect did all of it by capturing motion, and it really did look like "magic."

Even Steven Spielberg came on stage to vouch for it: "What Microsoft has done is not about redesigning the steering wheel — there is no steering wheel at all."

Comparing games to driving was a precise metaphor, and behind it hid a second meaning: driving with no steering wheel is a dangerous thing to do. On the surface everything had been simplified, so the obstacles in a newcomer's way disappeared. But for games that need precise input to get anything done, with no real buttons or sticks and only a player's imprecise posture standing in for them, the experience was simply baffling.

In the end Microsoft's aim was to "simplify" input, and the point of simplifying was to widen the console's audience so players of any age and any skill could join in. At that year's E3, everything from Kinectimals to Kinect Disneyland was demoed on stage by a group of children who could hardly be called players at all. The moment they flailed and squealed while reaching out to stroke a virtual tiger, it was clear Kinect would never become a core gamer's favorite thing.

Put this back in its context. Nintendo and Sony had both launched or were developing motion-sensing devices of their own. That same year Sony began experimenting with pairing the PlayStation Eye with a motion wand, which later became PS Move; and Nintendo's motion-controlled Wii Sports had long since been a huge hit, with the Wii Remote especially good at pointing and swinging, letting sports games show what they could do. That playful, physical experience was also what made the Wii deliver on its "living-room entertainment" positioning. Living-room entertainment was what every console maker was straining for, because it meant the product had broken past its original audience. And in 2008, two years after Wii Sports launched, the Wii's US sales passed the Xbox 360's.

"The launch of Project Natal had nothing to do with Sony or Nintendo," said Shane Kim, then VP of business development at Microsoft's entertainment division, in an interview. "This really was just about breaking down barriers and reaching a much bigger market, because controllers are a barrier — they look intimidating."

Half Magic, Half Frustration

After two years of E3 hype, Kinect went on sale in the US on November 4, 2010, for $150. Times Square was packed; Microsoft employees danced together to motion games under a giant banner reading "You are the controller."

Kinect sold 8 million units in its first 60 days, entering the Guinness Book of World Records as the "fastest-selling consumer electronics device." Across its seven-year life it sold 35 million units, half of them in that first year.

The consumer's decision is easy to understand: a Kinect with no controller looked like fun, and Wii Sports before it had been such a good time. There was no reason to suspect from the start that this would turn out badly.

Microsoft had also spent $500 million on marketing, packaging Kinect as the most promising new technology going — not only for games but for classroom teaching, medicine, learning an instrument, space exploration. Even Oprah Winfrey gave away Xbox 360s and Kinects on her show. And among the first wave of outlets to review it, the raves came from mainstream media outside the seasoned gaming press (Engadget, PC World, Kotaku and Eurogamer all stayed cautious).

Then the problems started to show.

Small ones at first. You needed a lot of space for Kinect's motion capture to work properly; if the living room wasn't big enough you had to start shifting furniture, and anyone in a smaller place couldn't play at all, which ruled out student housing almost entirely. And of the 17 Kinect games at launch, only Dance Central struck people as fun. Most of the rest seemed to be there so you could try out the new system, stretching your arms and legs: "there just aren't any good games."

Worse came later. Even players who started out patient with Kinect were disappointed wave after wave: nearly every Kinect game fell into family entertainment or fitness, they were all much alike, and the quality was generally mediocre. Players had paid an extra $150 for Kinect, and they hadn't wanted repetitive, dull, inaccurate experiences — and these genres were not what the Xbox 360's core audience liked in the first place.

The big publishers had said they would support Kinect. At E3, EA and Ubisoft promised to bring motion control to their sports franchises and to the Tom Clancy series. It turned out those games did not suit Kinect. With its lag and its imprecise readings, Kinect actually made the games worse: responsiveness fell off badly and the whole thing became sluggish, out of control and unpleasant. (One player was even shown a red card in FIFA for swearing into the voice mic.)

And a player who wanted to properly play a Tom Clancy-style shooter would never drop the controller and let Kinect capture his movements instead. True core gamers treat precision as perfection. They don't mind complicated button combinations, or the "barrier" of a controller binding their hands, because what they want is a perfect combo or a headshot timed exactly right. Waving at a television doesn't deliver that. It feels too easy, even a bit stupid; it erases the gap between beginner and expert and makes the grind of getting good stop mattering. In a sense it strips video games of their meaning.

A more obvious reality was that publishers had no need to support Kinect. They had a settled process for making AAA games, and that process fit a world with controllers. No players were complaining that controllers felt clumsy, so there was no reason to "simplify" the way Microsoft wanted. An Xbox insider said: "A game like Halo doesn't need Kinect at all. It already has a very precise, very detailed control scheme, and for the sake of balance you can't hand Kinect players that big an advantage either."

For developers, in short, abandoning an existing process to make Kinect games was a thankless job.

No big titles meant Kinect was unpopular, which meant fewer developers wanted to put effort into it. Back then this formed a vicious circle much like today's "no blockbuster VR games" problem. Kinect had become a fairly useless device, and the motion games that felt novel at first quickly grew boring.

The shortage of good games killed Microsoft's living-room dream. Kinect plainly never won over core gamers, and it never brought in the broad audience it had imagined either — those people already owned a Wii, or found Kinect games less novel than promised, and the device's later reputation was more than enough to talk them out of buying one.

The Project Milo demo that once looked so magical — talking and gesturing at a virtual character on screen — answered no real need either. Players don't judge a game by casual surface interaction between the room and the screen; they judge it by gameplay and story, which are what create immersion. And this much-anticipated project was never actually built.

A Futile Rescue

Kinect kept growing by being bundled into discounted Xbox 360 deals. In February 2013, Microsoft's published figures showed 76 million 360 consoles and 24 million Kinects sold.

The bundle came back at the Xbox One's launch. When the next-generation Xbox One arrived in November 2013, Microsoft said the console would ship with an updated Kinect, which would be an indispensable part of it. That pushed the price to $499, a full $100 above its rival the PS4.

Core gamers, of course, hated the idea. Nobody wanted to pay an extra $100 for a Kinect that had nothing to do with how they played, and it still needed a lot of room to work. It amounted to Microsoft saying, "Please don't buy an Xbox One."

That drove players straight to Sony. The two consoles launched in the same month, and by the second month PS4 sales had easily passed the Xbox One's. Six months later the quarterly reports showed a clear gap: Sony sold 3.5 million consoles (PS4 and PS3) in the first quarter of 2014, Microsoft 1.1 million (Xbox One and 360). The estimated split between people favoring the PS4 and the Xbox One was roughly 7:3. Other things fed that leaning, exclusive games among them, but nobody doubts that "$499 with a Kinect attached" badly damaged the will to buy.

Meanwhile, though Microsoft kept insisting "Kinect is the future," little about the device had really changed. It may have improved in limited technical ways, but it still hadn't solved the central problem of poor precision and poor control, and no good studio was interested in making Kinect games, so nothing good came out. The only Kinect game to launch alongside the Xbox One was Fighter Within, which scored 23% on Metacritic.

Once it realized players genuinely did not want Kinect, Microsoft finally gave ground. The last straw came in April 2014, when Kinect Sports Rivals — built by a 150-person team at Microsoft's Rare studio to show off everything the new Kinect could do — failed miserably, which eventually cost 15 Rare developers their jobs. Seeing how it would end, Microsoft dropped the bundle in May, and players could at last buy an Xbox One without Kinect for $399. Xbox One sales revived and the gap with the PS4 began to close.

Nobody was making Kinect games any more, not even Microsoft. On the Xbox One S, released in 2016, it removed the built-in Kinect port and told players to use a separate USB adapter.

So Kinect had been dying slowly for a long time. Its last days were spent burning off inventory. If there is any consolation, it is that Microsoft picked up a lot of technical experience building Kinect, which indirectly produced the sensing and recognition technology behind HoloLens and Windows Hello.

Cover image from NBC and Kinect Disneyland.