On Game Design with Rich Hilleman (Part 1)

Rich HillemanRich Hilleman is the Chief Creative Director of EA. He is one of EA’s earliest employees and is best known for helping to build the juggernaut EA Sports business as the original producer of games including John Madden Football, NHL Hockey and Tiger Woods PGA Tour. This interview took place in April, 2012. For more from Rich, check out part 2 and part 3 of this interview.

EL: What are some of the games you’ve worked on in your 29-year career at EA?

RH: The very first game I worked on was a game called Chuck Yeager’s Advanced Flight Simulator, which then became Chuck Yeager’s Advanced Flight Trainer. We worked on a number of other simulations from that era with Lucasfilm and with others. We built driving games in that era which included Ferrari Formula One, an Indy 500 game. We also built Road Rash. I built the original Genesis version of Populous, of all crazy things. We built the first version of John Madden Football for the Genesis. We built the first version of NHL Hockey for the Genesis. Built the first Tiger Woods PGA Tour. Built American McGee’s Alice. I’m sure I’m forgetting other things I shouldn’t be forgetting, but I’m sure I’ve insulted somebody.

EL: [laughs] It’s okay. It’s good to have so many incredible hit classic games under your belt that that’s actually an issue.

So the question I start everybody off with is, what is game design?

RH: I think game design is the process of assembling the components that can make up a game to produce a desired experience in the player. There are a lot of different flavors of that I think. There are folks who build very prescriptive experiences. I worked on the Winged Commander series. We gave the user choices but trust me we didn’t give them that many choices. Apparently we don’t give them enough choices in Mass Effect anymore.

Those are games that the designer has a point of view about what they want you to experience. They want you to make some choices, but they want you to operate within a range so they can really produce a rich experience for you.

The other end of the spectrum is sports games which are really about creating the tools for somebody to be able to fulfill the fantasy they probably already have in their head. And sometimes that’s a very specific thing: they want to be a particular player in a particular place. Other times, they want to use it as a tool along with their imagination to realize something that you couldn’t even describe in advance.

And so for me game design is the process of either assembling that point of view in one case, or assembling the tools that allow your user to have that point of view in the other.

EL: I think sports games are a really interesting area because it’s such a specific area of simulation.

RH: Painfully specific.

EL: Having done so many sports games, how is the job of being a designer on FIFA or John Madden Football different from Seth Marinello’s job making levels on Dead Space?

RH: It has the illusion of being easier, but I’d make the case it’s harder. In Seth’s case, there is no right answer. The user doesn’t know what it’s supposed to be, he just knows whether he likes it or not. And so in that particular case, Seth’s job is to create an experience that has the right frequency, that has the right impact on the player to create an emotional narrative within the player that deepens their care for the outcome of the story over time.

Most of the time in a sports game, the player thinks they already know what the game is. They think they already know what the story is. One of your risks is that you either somehow negatively impact that, that you somehow don’t allow them to realize the story that they’re after, or that you intrude your own on them.

I think the reason that it’s harder is because what people think they know about sports is two characteristics that make it difficult. One is that it’s incomplete and the other one is that it’s often wrong. In modern American football, play calling and the execution of plays in a modern context is a responsibility of eleven players on one side to read the situation and make exactly the same decision at exactly the same time together.

Almost nobody understands that. It means that, if I give you control of a player, you need to understand the play that’s going on and need to understand the multiple approaches for the position that you’re playing. If you were playing defense in Madden, switching from player to player, that means you have to know eleven of those, not one of those. And you need to know eleven times three or four, probably.

So that is a complicated, realistic problem. If I give you that to solve, you will do nothing but fail. So our job is to give you what you think is the truth but really isn’t. That creates for you the sensation of authenticity. That’s usually equal measures of what I call “dirt,” which is the minutiae that makes up the specific and distinct characteristic of a sport combined with something you didn’t know before you showed up: something that we taught you about the sport that you never knew before.

That seems to be enough. The problem is, it’s a moving target and every year we have to improve it.

EL: Was there a time where you first started encountering this actual cognitive friction between building a feature in a sports game that was true to real life and then watching it fail to meet people’s expectations or fantasies about what the sport actually was and how they reacted?

RH: I didn’t learn it from sports, we learned it from flight simulators. What’s funny is that I came to sports products from doing flight simulators and driving simulators.

What that meant was that my perspective over here is very much shaped by the experiences we had over there. When we built Chuck Yeager’s Advanced Flight Trainer, it was a less pedantic and less articulate flight simulator than Microsoft’s flight simulator at the time. We also ran at four times the frame rate and had airplanes people cared about.

What did I learn out of that? Apparently not very much because I instantly went off and made another mistake. We tried to build an F-16 simulator to compete with Falcon and when Falcon shipped, it shipped with, I don’t know, a 160-page manual. Well, I don’t know if you remember, but in order to fire a missile on Falcon, you had to do like seven things. You had to identify the target, you had to range the radar to the right, you had to restrict the seeker head on the missile, you had to engage that seeker head, you had to receive a tone that it had been locked on, you had to lock the radar image to the tone, and then you had to arm it and fire the missile.

It was like eight things to fire the missile.

EL: That sounds like a very authentic simulation.

RH: It was a painfully authentic situation. Well the truth is, on F-16 Combat Pilot, we spent like a million dollars training those guys. And so if I give you a game that makes you do all the things that an F-16 makes you do, guess what? You never do anything, number one. Number two, the experience isn’t all that cool. To shoot down another airplane in an F-16 in a modern air combat: it’s a radar game. There’s a little blip on the screen and then I fire a missile at that blip and then the blip goes away.

And what people really want is Tom Cruise in Top Gun. They want to pull the trigger and shoot the thing down, and the whole thing happens in visual range, and the whole thing feels like it’s a mano-a-mano contest.

Modern jet combat has nothing to do with any of that. But that doesn’t mean that that’s not what people want. So I think it’s the classic example of when “the truth” and “the legend” are in conflict, print the legend, because that’s what people want.

So what we discovered was the right thing to do was to give them Tom Cruise with just a little bit more authenticity than they wanted. Call the missiles the real missiles. Have the right airplanes be the right airplanes. Maybe have them go equally fast: something that the user could track the difference and actually perceive that difference within the context of the game.

But we didn’t make them fly the different tactics. We didn’t make them fly. We didn’t make them use their weapon systems in a highly authentic way. We didn’t make them use radar systems in the coordinated fashion that the Soviet Union did. Most importantly, it turned out that for most of the 1980s and 90s, if you were a guy flying a jet fighter, you actually couldn’t fire the missile. The missiles were fired by the ground. Your job was to fly the airplane and then they fired the missiles. So that’s a distinctly unsatisfying expression of that.

What we learned by the time we got to sports was that we had been down that road already. We had already made that mistake of trying to present something that was so authentic it was painful.

And we’ve continued to have to solve that problem though. I think Madden to this day continues to be a problem where Madden is hard and football is hard. Together they’re nearly impossible. And so the new player problem for Madden is just a problem that we work on almost every year. We’re not solving it particularly well, but we’re working on it.

EL: It sounds like at its heart the key to doing a fun simulation game is delivering almost the Hollywood-level legend and not the actual simulation.

RH: The key thing is to recognize the reality you’re trying to create is the one in their head, not yours. And that if it doesn’t react favorably to that existing context in their head, it doesn’t matter if it’s true, it’s inauthentic.

Authenticity is based on the user’s experience and not reality. And sports are no different. It seems like all simulations, all things that are related to the real world, that’s how people think about them. It doesn’t matter if it’s Tony Hawk, for that matter.

EL: Yeah, Michael John and I talked about when he’s training designers sometimes he’ll teach them the “player thinking” which is, he’ll tell them, “I’m not listening to you until every sentence starts with, ‘The Player…’” And that sounds like it’s almost exactly what you just said with simulations. Frankly, in all videogames, it’s not about figuring out what’s true objectively. It’s about figuring what’s true in the player’s mind, and giving that to them.

RH: I think what’s ironic is in spite of the fact we don’t seem like one, actually we, at our best in particular simulations, are performance artists. And so, when you’re a DJ, it doesn’t matter if you’re right if the audience doesn’t dance. It doesn’t matter.

And I think in our case, that’s very much how it works. We’re looking for that response out of the user that says that we’ve engaged with their authenticity and their sense of anticipation with what’s going to happen in the game. And they’re drawing pictures and filling in spaces that I can never fill in their head. They’re having experiences that I couldn’t afford to give them. The power of simulations is what already exists in people’s heads. You fight that at your peril.

EL: As chief creative director here, what’s funny is that I worked for you for seven months and I’m not—

RH: [Laughs] You still don’t know what I do.

EL: I still don’t know exactly what your job as chief creative director means.

RH: There’s a dissonance between them and so the part of that that I think is the most actionable for me is really around three things. One of them is the quality of the design talent and production talent that we have as a company.

I invest in making sure that we are spending the time and space necessary within the university programs to foster the kinds of people that we want out of those programs, and then to identify the ones who are really great. And then to do secondary investments in those people, like you, to make sure that they’re ready for their futures. That’s the first thing that we do.

The other thing that we are responsible for is the state of the art of game design. For instance, Sandy, who’s in our group now, is with us because I believe that the free-to-play model will advance more rapidly in China than any other market. And that our understanding and exposure to that market and how it works will directly influence how successful it can be in the U.S., emulating that model when it happens.

For us, that’s an odd kind of sideways thing, but it’s actually really about game design at the bottom of all of that. And so I think the part that makes sense for that title is our advocacy for the role of designer and our advocacy for the discipline of design and the new things that will emerge in that space.

EL: Being a game designer means a lot of different things to a lot of different people when you’re trying to build: working with universities and the young game design talent here at EA and in the industry. What is the role of the modern game designer that you are helping to craft?

RH: A guy who builds a shooter like Seth, or a guy who builds a simulation like Sim City, or a guy who builds a social game, or a young lady who builds a social game, or a person who builds a mobile game, or a person who builds a triple-A console game, the problems you wrestle with become different because your audience is different, because your monetization systems are different, because your distribution is different, because the frequency and duration of the periods of time that people get to play it are different.

Increasingly what they share in common is a highly metrics-oriented relationship with their customer in the long term. If I try and get one thing across with the university programs of today, it is how to be in command of the information that your product expresses about how the player is playing it. To be in the business of changing those numbers, anticipating those changes, and explaining to the rest of your team what those things are and what they mean.

For a long time in this company and really early on, I think when you worked with me, designers were the lowest form of life in the company short of audio designers because on the average team they were outnumbered by artists 30-to-1, producers 10-to-1, and engineers 10-to-1. The only thing or person that they might outnumber is there might be three designers and one audio guy. “We’re going to go kick the dog now. We’ll beat up the audio guy.”

EL: You guys even told me to switch to being a producer because to do what I wanted to do at EA, I had to have that in my job title.

RH: You wanted to be in control in a way that I thought you needed to be a producer to do.

I think that’s changed. That doesn’t mean necessarily the designers are as much in charge, but I think that the increasing interest in telemetry and metrics have made the designer a job that we now understand how to evaluate. And I think the key issue before was the way that the company and most of the business evaluated a designer was about every 18 months when they shipped something. And the number of other factors that go into that equation dramatically swamp the designer’s real influence on that. Only if you can really take apart a product can you understand what the designer did versus the mistakes that somebody else did to them.

But, I think what’s interesting about this is, if you have metrics, if you have telemetry, and you have an ongoing live relationship with a customer, suddenly you can tell a good designer in about three weeks. And I think that’s really what’s changed is designers have a way to describe to their customers why they’re great and why you can depend on them in a way that very few members of the other teams actually can.

It’s gone from maybe the least understood and least measured component of the product to arguably the most in a very short period of time.

EL: It’s definitely the most measured. I feel the understanding just from our own experience with metrics on Dragon Age Legends. There’s a lot of room to grow there.

RH: The analysis portion of it—once you’ve acquired the numbers—doesn’t mean you know what they mean. And I think we’re still going through a lot of that.

EL: I think one of the most insightful things I’ve learned from talking to various metrics people in the past couple of weeks is actually that the people who do A/B test great and it really pays off. When you ask them how many of their tests have no effect, they’ll say most of them. Sixty percent or seventy percent of things you test have literally no effect, no significant change.

I wish I had known that twelve months ago. I could’ve made a hundred better decisions on my product had I just been able to say, “Hey, you know what? Seventy percent of the time, we’re going to see nothing. And when we see one percent change, that’s a huge win.”

RH: Yeah, knowing how to celebrate. I think a big chunk of that is why I think designers are starting to gain some headway is (A) they’re explaining those things, and (B) they understand when they can change them. When you tell somebody, “I’m going to make this number change by three percent,” and then you do three percent, and you do that like three times in a row, it’s fucking magic. To everybody else in the room, what you have done is magic.

Now the truth is you could probably explain to them why, in most cases, because for you to predict that, you’ve got some reason why. But for most of the people they’d just never bothered to think through the details enough to understand that that’s an anticipatable thing.

So simply the fact that you can anticipate it, you can forecast it in advance, and that you were right, there’s a point—as you’ve heard me describe before—you do that like three times in a row and the producer says, “Just leave him the fuck alone.” [laughs] “I don’t know what he does or how he does it, but he does shit that none of the rest of you know how to do. Leave him alone.”

If you were going to describe the end-state that designers want most, that might be it: “Leave me alone.” [laughs]

On Game Design with Rich Hilleman (Part 2)
On Game Design with Rich Hilleman (Part 3)

Michael John, General Manager of GlassLab
Paul Barnett, General Manager of EA Mythic
Jenova Chen, Chief Creative Officer at thatgamecompany