Six weeks ago, Cevat Yerli sent his wife and kids on a vacation to France. He wasn't going to be any use to them for a month and a half, he figured, with crunch time that would keep him at the office until 11 or 12 p.m. — or, on some days, until 5 a.m. Growing a beard, stressing out and making a video game.
As CEO, president and founder of Crytek, Yerli currently oversees six games, a company and a technology engine from his office in Frankfurt. But for the past month and a half, he's been head down on what he says has been his studio's "most challenging," "most expensive" and "most demanding" title to make, Ryse: Son of Rome. Meeting with project directors, reviewing builds and giving feedback.
"It's a little bit of a dictatorship right now," he laughs, while sitting in a Crytek cafeteria in late May. "There's too much at stake."
This week, if all goes according to plan, Yerli will appear on stage to reveal the game publicly during Microsoft's Electronic Entertainment Expo press conference in Los Angeles. A big moment for any studio, magnified in this case by a development cycle that started seven years ago.
Since 2006, when Yerli thought up the idea that would become Ryse, the game has changed art styles, settings and development teams. It's gone over budget and shifted from first-person to third-person, from Xbox 360 to Xbox One and from gamepad controls to motion controls and back again.
"I would say we almost over-designed it, over-engineered it," Yerli says.
But, he says, without that work, Ryse wouldn't be the game it is today, one that he thinks has a shot at challenging established action game franchises with its first entry, as a launch title for Microsoft's new console.
The early days
In 2006, Crytek had one hit under its belt, the sunny island shooter Far Cry. Microsoft had just released Xbox 360, and its Kinect motion control device was still a prototype at Israeli startup PrimeSense.
Crytek was deep in development on a Far Cry spiritual successor, but Yerli wanted the studio to expand and tackle multiple projects.
Yerli had an idea for a pair of games that could work in tandem: a massively multiplayer online game called Kings, and a first-person action role-playing game called Kingdoms. Early concept art within Crytek showed hooded characters with swords, loosely resembling the protagonists from Ubisoft's Assassin's Creed franchise, which at the time had yet to be released.
Kings and Kingdoms started small, with staff dabbling in prototypes and fleshing out the concepts enough to pitch them to publishers. "It really was just putting together a little bit of a universe," says Nick Button-Brown, general manager of games at Crytek.
"I knew that the life of a close-up experience, a visceral experience, is something that [players] could relate to really well. There's no reason why not."
Crytek pitched both games to Microsoft in 2009, and heard that the publisher wasn't interested in Kings but wanted to continue talking about Kingdoms.
Button-Brown pulls up a video in a Crytek conference room, showing what the team had in mind. It starts with a medieval courtyard on fire, with a hero fighting off enemies one by one using first-person melee attacks; bloody and aggressive. At the time, a slide in the pitch presentation estimated a Q1 2011 release.
"This wasn't a working game; this was very much a proof of concept," Button-Brown says. "If this was a working game, we'd probably have shipped it. ... We were trying to prove that the first-person melee worked and could be balanced and could be fun for a whole game."
"I knew that the life of a close-up experience, a visceral experience, is something that [players] could relate to really well," says Yerli. "There's no reason why not. I couldn't find a counterargument, effectively."
Cevat Yerli (above) sits with a sculpture from a cancelled shooter called Redemption, which Crytek worked on for two years with Electronic Arts before throwing in the towel.
Yerli works with a member of the Crytek art team.
Kingdoms concept art from 2006
As Microsoft got wind of the game, the pitch made its way to current Microsoft Studios Corporate Vice President Phil Spencer, who liked that Crytek wanted to branch out. He thought Kingdoms could fill a hole in Microsoft's first-party lineup, which wasn't strong in the melee combat genre, and he liked that it would cement Crytek's CryEngine technology on Xbox hardware for other developers who wanted to license it.
"I kind of drew some analogies to the first time I met with Epic around [Gears of War]," says Spencer. "You know, a studio that had done some PC things and some cross-platform things and wanted to go big on something that was a little different for them."
Spencer notes that when Microsoft and Crytek began talking, Kinect still hadn't been announced publicly. As talks progressed, the two sides came to agree that Kingdoms would not only make sense for Microsoft, but would make sense as a Kinect game as well.
"It's really weird looking back at something from four years ago, because at the time it felt like this was a great pitch, and I'm looking back going, 'This was a terrible pitch,'" says Button-Brown.
"We've evolved a fair bit since then."
Budapest to Frankfurt
At the 2010 Electronic Entertainment Expo, Microsoft announced Codename: Kingdoms via a teaser trailer filled with live-action footage of warriors covered in dirt, dripping with blood.
Behind the scenes, the game had begun to evolve — its controls shifted to Kinect, with players holding their arms in the air to simulate holding a sword and shield, and the setting shifted from "medieval high fantasy" to a more realistic Rome.
"We were trying not to show that it was Roman at that stage," says Button-Brown. "There were kind of hints of Roman [in the trailer], but that was our secret we weren't sharing."
At the time, Crytek's Budapest studio was leading development on the title, and while team members at Crytek don't get specific in our interviews about what may have gone wrong, they say they decided to pull the game away from the Budapest team in early 2011.
By E3 in June 2011, Crytek would settle on the name Ryse, lay off staff in the Budapest office, transform the rest of the Budapest office into a mobile game team, move Ryse development to Crytek's home base in Frankfurt and question whether Ryse should be a Kinect game in the first place.
Amongst all this, Crytek showed a second trailer with the first public glimpse at what the game could look like — revealing Kinect sword and shield controls, though not using actual gameplay footage. Behind the scenes, the game's control scheme and approach was in limbo, and Yerli describes the trailer as a "market test" to see how the public would react.
It was one of many tests Yerli and team would do to determine the game's direction, now that it was back in Frankfurt — just down the hall from Yerli's office — where he'd be able to keep a closer eye on it.
Codename: Kingdoms in 2010
Ryse in 2011
Throw everything at the wall
While Ryse evolved through its teenage years — largely, 2011 — Yerli and team were willing to experiment to see what would work. They assigned groups to test concepts, even using external teams to prototype ideas on a contract basis.
One came from Crackdown developer Ruffian Games, as a report on game site Kotaku pointed to last year. Crytek hired Ruffian to try a player versus player multiplayer mode, which didn't end up making the cut for the final game. Yerli says that's no reflection on Ruffian's work, though; it was just one of many ideas tested and left out.
Another, led by Crytek's Frankfurt studio, was to make Ryse more a cinematic experience than a traditional game — "essentially an on-rails interactive movie," says Rasmus Hoejengaard, Crytek's director of creative development and one of the main forces behind the game's story.
Yerli liked the idea, as a sort of revival of '80s arcade games like playable cartoon Dragon's Lair. It would be on-rails with branches, heavily guided, four hours long and offer combat and dialog choices. Some team members thought it would lean into what Kinect could do well and would enable the game to showcase the CryEngine's visuals. But eventually the studio put it aside.
"It just felt difficult to sell that to a player that expects to play a game," says Button-Brown.
"The funny thing, though, is it was this pitch [that] actually what got us to where we are now," says Hoejengaard.
The moment of truth
As 2011 went on, Crytek still hadn't made a final decision on Ryse's interface.
Some on the team liked the Kinect controls, and were proud of the progress they had made with combat and navigation prototypes. They'd developed tricks to disguise input lag, and were working on getting around motion control design issues. "We cracked navigation, and we were 80 percent there in first-person combat," says Design Director Patrick Esteves.
Others at the company were worried about players getting tired with extended play, and Kinect not having the precision that players wanted.
So the team built three prototypes to help make the decision. One used Kinect controls. Another used Kinect along with a traditional controller, so the player could move their head to look around but the combat went through the gamepad. And a third was strictly on the controller, with a new third-person camera angle showing the character on screen during gameplay for the first time.
"[We were both worried] core gamers may not yet be convinced to use Kinect."
The team figured it could put these three side by side, survey staff at Crytek and make the call.
"We tried to make it as scientific as possible," says Esteves.
"Everyone played these three," he says. "And [we all answered], 'What resonates with you as a gamer, being a developer on this team?' And honestly, it was pretty much 30 percent like this, 30 percent like this and 30 percent like this."
Back to square one, then.
Crytek expanded its reach, testing the concepts with others staff knew outside the company, and in Microsoft's user research labs.
Eventually the numbers started to skew toward the controller-only prototype. Largely, the feedback came in, players preferred seeing their character. Ryse had become a heavily story-based game by that point, with a focus on a main character named Marius, and Crytek's research said players liked seeing him on the screen.
As time went on, the team also realized that going the controller-based, third-person route would lead to a higher quality game. "When you have to teach players how to just do basic actions, you're going to spend a lot of your time doing that rather than making them be immersed in the world," says Esteves. "That was kind of the tipping point."
Crytek came to the decision to focus on the controller, then sold Microsoft on the idea.
"Microsoft and Crytek sat together at a table and said, 'Are we still doing this game Kinect-based?' Then effectively we and Microsoft, though nobody dared to say it to each other [at first], found that this was not the right way to go forward," says Yerli. "[We were both worried] core gamers may not yet be convinced to use Kinect."
The move to Xbox One
After changing the setting, art style, control scheme, development team, perspective and control scheme again, Crytek had one last major shift waiting for it: the switch to Microsoft's next-generation Xbox, recently revealed as Xbox One.
As Yerli tells it, this officially came along in 2012, fairly late in the game's development, but the team was planning for a next-gen shift much earlier. When the interface changes pushed back the game's planned release date, Yerli suspected that Microsoft would want the game to move to its next hardware.
"We kind of knew it," he says, "because it wouldn't make sense to [ship on 360 then] ... But at that point [Xbox One] wasn't on the table yet."
"It didn't take rocket science to say, 'Boy, this would be a nice game to line up with the launch window of the new generation.'"
As a company selling an engine as well as making games, Crytek was eager to be a part of it. But since Microsoft couldn't commit to the new console as of 2011, Yerli and the team decided to develop the game for what they guessed would be a next-gen visual target, hoping it would work out, noting that the team was "running really high-end pipelines and research anyway," so the amount of work to jump between platforms wouldn't be massive.
Microsoft's Phil Spencer says it's not that Crytek was playing a game of chicken with Microsoft, but that they were working together and just didn't know the launch date for sure at that point.
"When we started looking at the date for Xbox One and this game in production," says Spencer, "it didn't take rocket science to say, 'Boy, this would be a nice game to line up with the launch window of the new generation. ... There's hardware schedules, there's platform schedules, there's game schedules, and all of these things intertwine. You imagine you've gotta land all three of those to make sure you understand what your launch window is for the platform and then what games you want to have there. It's the fun part of watching the sausage get made. ...
Why Ryse and not Rise? "We wanted to protect [the copyright for] Rise with an 'i' but we couldn't, so we made a 'y' out of it," says Yerli. ... "The 'y' suggestion actually came from Microsoft."
"We've done this in other generations as well, where you've got a lineup of games [that move between platforms]. The games, at some creative level, exist without the platform, if you just think about the creative of what the game is — if you think of Marius as a Roman general fighting, his story arc, the way you want to play it — some of those things are absent of what platform they're gonna ship on."
It also seemed like a natural fit, Spencer says, given Crytek's background of pushing game graphics forward — to see Ryse on Xbox One is to see a game that visually looks like a clear step beyond what's on Xbox 360. And one of the reasons that both Microsoft and Crytek say it made sense to put Ryse on Xbox One was to establish the CryEngine on the new hardware.
Although he makes the connection, Spencer cautions against drawing too many parallels between Epic Games with Gears of War on Xbox 360 and Crytek with Ryse on Xbox One, and says that part of the benefit of having Crytek develop a launch game was that Crytek's tech would go on to benefit other developers looking to develop for the system. Similar to how Epic built up its Unreal Engine on the back of Gears of War and then saw a lot of success in licensing that tech out.
"If you look at signing what was called Warfare at the time — Gears of War — on 360, it meant that we were going to get Epic focused on getting their engine onto the platform early," Spencer says. "Which obviously was good for their game, and our game that [we were] working on together, but also good for things like Mass Effect and other games that were coming out that wanted to use Unreal, because you knew that the support would be there for our platform. I think it helped. If you went and looked at the number of Unreal games, either from us or third parties, I think getting Epic focused on 360, or [when it was known by the codename] Xenon, early with Unreal was a good thing."
The Ryse DNA
Ryse in 2013 bears little resemblance to Codename: Kingdoms from 2006, but team members say it wouldn't be what it is today without having gone through what it did.
When Ryse debuts at E3 this week, the footage will highlight the spectacle, with fire-charged boulders flying at hero Marius and masses of warriors fighting alongside him. Even subtle background touches, like cloth flapping in the wind, are being designed to showcase what the new hardware can do in a more realistic way than Xbox 360. But those who look closely will also see how the past informed the game in its current state.
"I always say to the team, 'Hey, last gen is about body language. Next gen is about facial performance or full performance capture and adding filmic techniques.'"
Shifting development from first-person to third-person initially made Crytek team members feel like they missed fighting enemies up close, for instance, leading to a closer-than-normal third-person camera view and a heavy emphasis on enemy emotions and facial reactions in Ryse. Which fed into a combat system where many enemies only take a few hits to kill. Which fed into the idea of combat "flow" and moving gracefully through groups of enemies. Which fed into a rule that players can't get hit by off-screen enemies.
The chain reaction formed what Crytek refers to as Ryse's DNA — "efficient" short-range sword and shield attacks that give combat a "death on the neck" feel, according to Yerli, with score multipliers building up based on how well the player performs, a focus mode where the player stays in real-time while enemies slow down and an execution "meta game" for each enemy that determines what short- and long-term upgrades players earn.
If Ryse sells well, Yerli has plans in mind for a sequel. "The title's not supposed to be a one-off," he says.
Away from combat, the game also presents over-the-top action scenarios and involved in-game cinematics using extensive performance capture in conjunction with Gollum actor Andy Serkis' company The Imaginarium Studios, a story focusing on "the two sides" of hero Marius and his revenge story to find those who tortured his family, all wrapped in an art deco-influenced visual style that sets up a realistic-looking Rome.
Ask the team members at Crytek, and the word they will often use to describe the game is "immersive" — a big goal for the team is to use the next-gen visual performance upgrades, cinematic techniques and Kinect features to make players feel like part of the world.
"I always say to the team, 'Hey, last gen is about body language. Next gen is about facial performance or full performance capture and adding filmic techniques,'" says Yerli.
And the team's experiments with Kinect haven't gone to waste either. While Ryse is primarily a controller-based game at this point, Crytek is still experimenting to see what Kinect motion control and voice features make sense to layer on top of that before launch.
At the moment, that means players can use voice commands to control their troops — telling them to arm a catapult or charge toward enemies, giving them a morale boost. But depending on how Crytek's experiments pan out, players might be able to hold their controller up at certain moments to block incoming arrow attacks, or taunt enemies to make them angry — which would make them more challenging to kill, but would also drive the player's multiplier score count up as a risk/reward.
Yerli makes it clear that players who don't want to use these features won't have to — they'll all have button replacements — though he thinks players will want to speak to feel like a leader of their army. And he says that in some cases players will be able to perform a command faster by speaking than by using the controller, which could lead to better combat flow and therefore a better score.
"Ryse was a challenge, and is still a challenge. It's going to be a challenge until it's in the shops."
Some of the more direct pre-third-person concepts have carried over as well, from select story moments to a co-op multiplayer mode where the player fights as a gladiator in a colosseum, to Smart Glass content that Microsoft plans to detail down the line.
"You can't say we would have ever gotten to the point in the game where we are now had we not gone through all of those steps," says Button-Brown.
Never too late
While Ryse's path to completion looks relatively stable at this point, Yerli says Crytek team members have plenty of work left to finish, and will keep experimenting until someone pulls the game away from them.
"That's the way it works with Crytek," he says. "We are not giving up working on stuff until the last moment. We are not locking it four months before, six months before. ... If it's a new [intellectual property], we are going right to beta, which can be like six weeks [out]."
But for Yerli, there's an element of personal satisfaction in reaching this point — as the person who thought up the idea for Kingdoms in 2006 and the only one at Crytek who's been creatively involved in the game for its entire lifespan, he's both nervous and relieved to see the game in its current state.
"Ryse was a challenge, and is still a challenge," he says. "It's going to be a challenge until it's in the shops. ... When at times certain things didn't happen as I wanted them to happen, I never thought, 'OK, let's put it in stop.' But I had my doubts that we could pull it off in the kind of quality I'm looking at."
"Building games is a creative endeavor," says Spencer. "If you think about the number of games that start as a piece of paper, as a pitch, whether that's from an internal studio or an external studio, that actually turn into running bits on a machine, it's probably one in a hundred. And there are things that we start that don't finish, and that's not because of failure. In fact, I think it's a good thing that there's a certain number of things that we start that we don't finish, because it means we're pushing hard to try to break out of just doing what everybody does. ...
"We could always see that it was gonna make sense for us to get a game done [with Ryse]. It was something that we believed in all along. How it was gonna get done, when it was gonna get done — those are variables that change for a lot of games. ... I've definitely had games that have taken longer to really find their feet than this one."