Ion Storm Austin’s Deus Ex was a critical and commercial success when it was released in the summer of 2000, laying the groundwork for an enduring franchise.
More importantly, it inspired a generation of game designers to dig into the nooks, crannies, and conveniently-placed ducts of what we now shorthand as "immersive sims," those games that ask players to make interesting choices in how they want to confront or slip past obstacles in their path.
Shortly after the game shipped, game director Warren Spector wrote a broad postmortem of the project. Today at GDC, he revisited the subject after 17 years to offer some fresh insight into how the groundbreaking game came to be.
“People always ask me which of my games are my favorite; don’t ever ask a game designer that,” said Spector. “The closest I ever get to answering is saying that the game I’m most proud of is Deus Ex.”
Spector acknowledges that a lot of what he's talking about is standard practice in the game industry now, but he reminds fellow devs that it absolutely was not common practice 20 years ago when the game was first conceived.
“Conceptually, I thought of Deus Ex as a genre-busting game, which let me say, really enamored us to the marketing folks. They loved that,” said Sepctor. “If you ever want to make a marketing person unhappy, mash some genres together.”
The game was envisioned as a mix of first-person shooter, adventure game, and RPG. But Spector traces the history of its development back to playing tabletop games, long before he became a game designer.
"I would not be here, you would not be here, if not for that game of Dungeons and Dragons"
“Let me tell you where it began -- this is going to get a little embarrassing,” said Spector. “It all began with Dungeons and Dragons.”
In 1978, to be precise, when he started playing D&D with a new Dungeon Master.
“I would not be here, you would not be here, if not for that game of Dungeons and Dragons,” said Spector.
The Dungeon Master of that game was “cyberpunk guru” Bruce Sterling, and Spector says the experience was meaningful not because of the story being told, but because of how it was being told -- by Spector and his friends, under the guidance of Sterling.
“The story belonged to Bruce, but every detail belonged to us,” said Spector. “I was completely hooked; I played in that campaign for ten years.”
So how does that lead to Deus Ex? Spector says it inspired his entire career in game design, a long-running attempt on his part to try and recreate that experience he had playing D&D for the first time in ‘78.
“That’s been my life mission: to recreate that feeling,” he said. “Every game I’ve worked on, every single one, has been trying to engage players in the telling of the story. My only hope is to do it a little better every time.”
So Spector became a game designer, and joined Origin Systems in 1989 to work on games like Ultima Underworld. But at some point he got tired of making those games, and he decided to “try and swing for the fences.”
“I was sick to death of making games about guys in plate armor swinging swords,” said Spector. He says he wanted to make what he saw as a “real-world role-playing game” in which the player’s choices would say more about the player themselves than their in-game avatar
“I don’t care about your puppet; I care about you,” he said.
He also wanted every player to get to the end of the story; he wanted to tell a story with players the way Sterling told a story with him and his friends back in the ‘70s.
At Origin Systems he pitched a game concept that allowed him to do all this: something called (Trouble)shooter, which was a sort of hard-boiled noir game. Electronic Arts and Origin weren’t interested at all, and Spector shelved the idea. A few years past, he left Origin and joined Looking Glass, where a game called Thief was being developed.
Spector says he had a moment while playing Thief: he hit a part that was too tough to sneak through, and impossible to fight through. So he asked the team to make the player character tougher, so he could fight his way through; but the team balked, saying that wasn’t the point of Thief. The game was about sneaking, after all; it wouldn’t make sense to let players smash through a roadblock.
That, says Spector, was when he got the idea to go back to (Trouble)shooter. So he left Looking Glass shortly before Thief shipped, and wound up nearly signing a contract with Electronic Arts to make a totally different game.
“That’s when John Romero called me and said don’t sign the contract; join me at Ion Storm and make the game of your dreams,” said Spector. “Who can say no to that?”
And so Ion Storm Austin was born, with the game that would become Deus Ex as its debut project. But before work got underway, devs may appreciate Spector's recollection of asking himself a series of questions he calls the 6 + 2 + 1 -- a vetting process he says he applies to all games he works on.
“Why I don’t call them the nine questions, I don’t know,” said Spector. “But if I can’t answer any of these questions I don’t make the game.”
- What’s the core idea? Can you describe the core of the game in 2-3 sentences?
- Why do this game?
- What are the development challenges?
- How well-suited to games is the idea?
- What’s the player fantasy? (If the fantasy and goals aren’t there, it’s probably a bad idea)
- What does the player do? (What are the “verbs” of the game?)
- Has anyone done this before?
- What’s the one new thing? (“You can always find one thing that hasn’t been done before [in games], even if you’re making a My Little Pony game.”)
- Do you have something to say? (“In Deus Ex I wanted to explore all sorts of big issues,” said Spector. “And I wanted players to explore those things in ways that only games could do.”)
All of these questions were answerable for Deus Ex, so in 1997 work on the project got underway at Ion Storm Austin.
“If we were gonna plug into the real cultural zeitgeist, we had to play close attention to things in the real world,” said Spector. Things like mechanically-augmented soldiers, or the rising tide of terrorism.
“You didn’t have to be a genius, even in 1997, to see that the news was increasingly filled with reports of terrorism,” he added.
”And really interesting to us was the rise of nanotechnology; the dangers of that seemed fun to explore. But most importantly, everywhere you looked in 1997, it seemed like there was a conspiracy theory,” said Spector. “The world of Deus Ex was being created all around us; we didn’t have to make anything up. It was great!”
In short, Spector says the game was designed to answer five big questions:
- What would happen if you mashed up the adventure game, shooter, and RPG genres?
- What would happen if you dropped a secret agent into a world wih no black and white -- just shades of grey?
- What if all conspiracy theories were true?
- What does it mean to be human in world with augmentation?
- How should the world be? Would the world be better off ruled by a secret cabal, a sentient AI that connected all humankind, or plunged into a new dark age?
“The game was going to have no bosses to kill; it was about deciding how the world should be,” said Spector. “You could answer those questions through your play choices.”
To explore these questions in a game, Spector says he established some “commandments” for the team to follow:
Ion Storm Austin Commandments (or, the Deus Ex rules of roleplaying)
- Always show the goal. Players should always see what they’re trying to accomplish.
- Problems, not puzzles. “It’s an obstacle course, not a jigsaw puzzle.”
- No forced failures. “Failure’s not fun.”
- It’s “role” not “roll” -- ie.. “It’s about playing a role, not rolling dice,” said Spector. “Why do we still have character classes and skill levels and die rolls?”
- If there’s cool stuff happening, players should be doing it. Players do the cool stuff; NPCs watch the players do the cool stuff.
- Lay out constant rewards to drive players onward.
- As the player gets better, make the game harder.
- Think 3D. “I believe 3D maps can’t be laid out on graph paper; if players aren’t looking up and down constantly, you may as well make a 2D game.”
- Be connected. Tunnels from A to B aren’t interesting, says Spector; interconnected spaces are.
- Every problem should have multiple solutions.
By the time the team moved from the concept phase to pre-production, the Ion Storm Austin team had grown from six to about 30 people. It was a “totally dysfunctional team,” according to Spector, and the team had “about six-ish months” of pre-production. The design document was cut down to 270 pages (“which nobody read”) and the team entered full production.
“You don’t want to know how many times I was told to just make a shooter, and I said no,” said Spector.
“Then there was the day we hit pre-alpha and realized the game was not fun; that was a good one,” recalled Spector. By this time, September of 1999, Spector says the game had finally come together and the team found out it wasn’t fun.
“Luckily, Eidos saw the potential of the game and gave us more time,” said Spector. “By June of 2000 it was ready to ship. It was the game of my dreams.”
He seems to really mean this, too; Spector says at the start of every project he closes his eyes and envisions what a game will be. By the time Deus Ex had shipped, he looked at it and recalled that “every single detail had changed, but in spirit it was exactly the same. It was kind of amazing; it’s the only time that’s ever happened.”
So what positive lessons were learned from the experience that fellow devs might appreciate? As you might expect, Spector had a list:
Don’t give up
“The lesson here is that if there’s a game you have to make, never give up, because someone’s gonna be stupid enough to give you the money some day.”
Have clear goals
“We wanted players thinking about who they were in the world...we wanted players to think about how they wanted to behave in the world,” said Spector. “We wanted them to feel like they were actually in the world….everything in the game had to be based on something real.”
“If there’s anyone in the audience here who worked on Deus Ex, I’m sure you hate me,” added Spector.
Don’t skimp on pre-production
“You can always use more, but we did okay,” said Spector.
Keep your game design organic -- be open to change
“Our game systems didn’t work as well in reality as I thought they would on paper,” said Spector.
Always have a playable build
“We always knew where we were, even if that was painful,” said Spector, explaining that the team would build out envisioned missions early, in limited form, to test how they played. These “proto-missions” were small and ugly, but critical to the game’s production because it helped the team see things that didn’t work, early enough to do something meaningful about it.
Also, says Spector, make sure to playtest with real people
“Do not let your publisher know you’re doing this,” said Spector, recounting how Ion Storm Austin would bring folks in to do playtests without telling Eidos. “Listening to their responses, I called that particular milestone the ‘wow these missions suck’ milestone.”
License good tech
“That was something we did right; we licensed Unreal Tournament,” said Spector. “Adding what we needed was much easier than it could have been...three programmers made Deus Ex. They were very overworked.”
So what went wrong? What should you know to avoid in your own game projects?
The team structure didn’t work
“I had two people qualified to be lead designers so instead of picking one, I gave them both roles as lead designers with design teams,” said Spector. “Don’t do that. I thought I could manage the tension between the teams, and I couldn’t...I had to call one of the design teams 1 and the other one team A, because neither would be team B or 2.”
However, Spector took pains to say one thing he did right was in choosing who to hire. “Those guys do not get enough credit for the creative aspects of the game, the fact that we shipped at all, and managing the team.”
Goals were too big and unrealistic
“Harvey Smith and another designer, Steve Powers, came to me one day and said ‘you cannot make this game. We cannot make this story,'” recalled Spector. “So we cut a bunch, and the game was better for it.”
He recommends you limit yourself in a kind of creative box, and ensure your entire team is operating within it.
More risks should have been frontloaded
Put simply: Spector says devs should prototype more systems earlier so you can iterate and make them better -- or throw out what doesn’t work.
Be careful about licensing tech
“This is something that I said we did right, but it’s also something we did wrong,” said Spector. “Unreal Tournament was not designed to have a deep sim and allow fighting, sneaking and talking...we had to layer all that stuff in on top. It worked, but….we ended up faking a lot of stuff, to be frank.”
He also added that the engine presented the team from being able to create the large, open areas they envisioned, and he feels it significantly impacted the game.
“It was a mistake to try and recreate the real world...in 1997,” said Spector.
Any publicity is not good publicity
“Ion Storm was not a quiet company; I don’t know how many of you remember this,” said Spector, pulling up an image of an ad for Ion Storm Dallas’ Daikatana, also released in 2000. He went on to mention Playboy shoots in the office, articles about pay rates at Ion Storm, and more, noting that “it’s hard for everyone on your team to be productive when everyone’s screaming about your company.”
But when the game finally released in 2000, Spector says the results were gratifying: it sold well, reviewed well, and most importantly (according to him) players got excited about the notion of making meaningful choices in the game.
“Be ready to define success for any game you make; if you can’t define success, you are not ready to make that game,” said Spector. “Conversation was my definition of success.”
He’s talking about the sort of conversations that are now a common goal in game design: the “what did you do? How did you get through this part? What ending did you choose?” conversations between players who are playing through the same game in different ways.
“That was what Deus Ex was all about. It wasn’t about what the development team thought they world should be, but what the player thought the world should be,” said Sterling. “Like Bruce Sterling and every other Dungeon Master, we provided the skeleton of the experience and the players put meat on the bones.”
“Here’s my final thought for all of you: I sincerely hope that everyone of you gets the opportunity to work on the game of your dreams some day,” said Spector. “And I hope every one of you gets to work on something that’s still relevant 17 years after you make it, and has a life beyond you.”