Hanging on the wall is a television screen. It is massive. In your home, this screen would be luxurious; ludicrous, even. Here it is merely functional.
The television is connected to a computer; on the screen, a complicated spreadsheet outlining every facet of the creation of a video game, step by excruciating step. The game does not yet exist. It is being built at this very moment. It's little more than a few lines of code. Some art. A wealth of documents.
And this spreadsheet.
In another year (give or take) the game will be complete. You'll be able to download it on your PC, your Mac or maybe your mobile devices. You might enjoy it — the people making it certainly hope that you do — but if you saw it in the state in which it currently exists, you wouldn't even know what it was.
Month by month, milestone by milestone, that bit of code, those pieces of art and design will multiply and coalesce. Slowly, following the steps outlined on the spreadsheet on the massive screen on the conference room wall, the pieces of a video game will be assembled. And at any point some little thing may go slightly wrong, derailing the entire thing.
This is game development.
I'm sitting at the end of a long conference table witnessing a planning meeting for Hidden Path's upcoming tower defense game Defense Grid 2. Halfway down the table sits Dacey Willoughby, the project's associate producer. She carries herself like a nurse: kind, but tough. Empathetic, but determined. She's done this before, this game-making thing.
Willoughby worked at Warner Bros. (on Lord of the Rings: War in the North) and other places. She owned a business with her husband (a smoothie shop — it didn't work out). She worked at Hidden Path once before, but quit (to run the smoothie shop). Now she's part of one of the most experienced small development teams in the industry, working to corral the disparate pieces of Hidden Path's next game into some semblance of order.
Right now the conference room is Willoughby's workstation. She's setting milestones: scheduled deadlines the team must hit or risk losing money. Technically, these milestones have already been set, but Willoughby is adding detail; describing, in writing, exactly which parts of the game will be finished and on what dates and by whom.
Over the next 12 or so months, the team members behind DG2 will hit or miss these marks. Willoughby hopes they hit more than they miss, and she has some reason to be hopeful there. The team at Hidden Path is composed mostly of veterans. In the game industry as a whole, the average "time in" for game developers is approximately five years. At Hidden Path, it's 12.
This veteran crew is now making the official sequel to its first (and only) original title (as well as a couple of other games that team members can't talk about yet). But the process of game development is complex. Things go wrong. Creativity, always hard to bottle, reacts poorly to pressure. People fail.
This is game development.
Polygon has been promised exclusive access to the Defense Grid 2 development team over the next year: its meetings, its milestones, its design documents. Everything. We will bear witness firsthand to how a video game gets made, from the ground up. Everything will be on record and reported.
Hidden Path has, in short, given Polygon total access to the development of Defense Grid 2, and we will get to follow along as it struggles to make the game that could make or break the independent studio.
Our reports will be published in installments between now and whenever the game launches (projected: 2014).
This is part one.
A tough week
I'm sitting in a restaurant with the co-founders of Hidden Path. It's a little awkward. They're not used to press or visitors. And they're smart. Scary smart. The kind of smart where you understand all of the words they are speaking, but none of what they are saying makes any sense.
The kind of smart where, when tasked with making idle chitchat with a stranger over lunch, no one knows where to start.
Mike Austin jumps on the grenade, kicking things off by talking about wind and how it can knock over buildings, a phenomenon known as mechanical resonance. I've heard of this, but their discussion goes into atmospheres of intellect too thin for my wings to carry me. I can simply nod and smile and hope, at some point, I'll be able to interject and not sound like a moron.
Austin is the youngest among Hidden Path's founders by about 10 years, but he's had roughly the same number of years in the industry. He just graduated from school early.
Right now Austin is heading up development on a game that's not Defense Grid 2 and noodling with technology unrelated to any game in his spare time. One of his recent pet projects involved writing an algorithm that could recognize musical frequencies. Turns out it was harder than he expected — and for the same reason that wind can topple buildings.
As Austin made progress on his algorithm he discovered that low notes are harder to identify than high notes, and when he researched the problem he learned that it's a common one. Low notes essentially borrow each others' frequencies, overlapping and cascading. Which is why, if you have a building built to be susceptible to a low resonant frequency, it will topple easier than if it wasn't, because any one of several frequencies could trigger a collapse.
Mike Austin jumps on the grenade, kicking things off by talking about wind and how it can knock over buildings, a phenomenon known as mechanical resonance.
This revelation (expanded upon in great detail) marvels the assembled brains at the table, and the conversation then turns to a proposed NASA project someone heard about involving jet engines placed at the corners of buildings, to "steer" them from collapsing in high wind. And then we're back to music, talking about how graphing musical chords on paper produces visually appealing graphs for chords that also sound pleasant, and visually unappealing graphs for chords that don't sound as nice. And then we're on to piano tuning, and how tuners don't really understand any of that; they just measure the vibrations of the strings.
Then the food comes.
In the span of 15 minutes we've gone from staring awkwardly at our water glasses to full-on nerdery over seemingly mundane yet stunningly complicated mysteries of the universe. That's the kind of smart these guys are.
Hidden Path was founded in 2006, after the founders had "a tough week" at their old job. The job? They were the masters of the Xbox.
Austin, Mark Terrano, Jeff Pobst and recently departed Dave McCoy all worked together at Microsoft as part of the Xbox team's Advanced Technologies Group (ATG). Sort of a game development "A-Team," ATG would travel the world, meeting with third-party game developers, helping them design the best games they could for the original Xbox console. Solving problems.
Imagine the people who design games as experts in computer programming and mathematical computations. Now imagine the people those people call when they can't figure out how to make something work.
Need a higher frame rate? Call ATG. Can't get your game to work with Xbox Live? ATG. Want to squeeze more pixels out of that console's Nvidia graphics chip? ATG. They were the guys who essentially served at the pleasure of the game developers, doing whatever needed doing, with a smile, and expecting nothing in return. All with the blessing of Microsoft.
Then things changed.
Depending on who you ask, it was either politics or money or both, but at a certain point the mandate for ATG took a left turn and the members of ATG cashed in their time off and called it a day.
"We left in February of 2006," Austin says. "We were working out of my living room, because I had space. I made breakfast every morning in my kitchen."
"We were launching the 360 and [Pobst] said, 'Did you ever want to go back into games? What would you make?'" That's Mark Terrano speaking. He's the design director at Hidden Path. Also, he developed the classic strategy game Age of Empires 2: The Age of Kings. "I told him and he said, 'My God, we gotta go do that.'
"So we both said, 'OK, if we were gonna pick anybody we could work with in the whole industry, who would we pick?' That's how we started, with a list."
Then: Hidden Path.
The ATG boys looped in veteran money man James Garbarini after the fact, because none of them wanted to do that job. Garbarini was the CFO for FASA, creator of the MechWarrior universe, and ran part of that company for a time, after most of its assets were sold to Microsoft. When Hidden Path called, he was contemplating leaving the game industry entirely to teach middle school.
"The appeal was that there was not any single person who would be solely responsible or solely in charge," says Garbarini. "That and the experience and the intelligence these folks had who were going to be participants. They were extraordinary."
CEO Pobst explains the name:
"With the launch of the 360 and the launch of [Xbox Live], I would go out and visit a bunch of studios," he says, speaking of his time as the group lead for ATG. "We saw a tremendous number of studios. The things we saw were: one, no studio made games the same way as another. It doesn't happen. Two, we saw a lot of genre studios, or technology studios, or art studios who then kind of did OK on the other stuff.
Need a higher frame rate? Call ATG. Can't get your game to work with Xbox Live? ATG. Want to squeeze more pixels out of that console's Nvidia graphics chip? ATG.
"We felt like the most important thing for a game was fun. We didn't see many design-based studios that could do multiple genres, that could work and be successful in multiple ways.
"You needed technical expertise and you needed art. You needed all of that to do it. You couldn't just focus in one area.
"We saw a lot of developers that we felt really just focused on one area. And so we thought, 'What if we built a studio that really was about how you get to fun?' And so that was the conversation. Hidden Path was the way you get to fun."
For the first few years, the industry was kind to Hidden Path. The founders' tight relationships at Microsoft helped them land key deals and greased the wheels for the release of Hidden Path's first original game, the science fiction tower defense game Defense Grid: The Awakening, in 2008.
Then the world economy collapsed, and a year or so later, the industry that everyone believed was recession-proof revealed itself to be anything but.
By 2008, Hidden Path was working on games with each of the three major console makers: Nintendo, Sony and Microsoft. Then, one by one, those game deals collapsed. Hidden Path was left with nothing. Then: layoffs, scrambling and a lucky last-minute save in the form of investment from overseas and a request from Valve to begin working on several projects, one of which would become Counter-Strike: Global Offensive.
Then the world economy collapsed, and a year or so later, the industry that everyone believed was recession-proof revealed itself to be anything but.
It is now 2013. Hidden Path, the company forged from the ashes of Xbox's A-Team, has reinvented itself as a high-caliber indie, with the experience (and then some) of a major triple-A studio. If it were a car, it'd be a Tesla with a rocket engine.
"It's still hard for me to think of it as being a real company now, because it's more like we hired people to help us do our stuff," says Austin. "Then at some point we started doing real things and suddenly we'd been around for seven years, which is still crazy."
This seventh year could be the start of Hidden Path's rise to greatness. After seven years of ups and downs (a lot of downs), the company is now working on multiple titles again, but this time around the future looks considerably brighter.
In the summer of 2012, Hidden Path Kickstarted a long-overdue expansion to the original Defense Grid. The hope was that the campaign would raise enough money to fund a full sequel. It did not. So Pobst sent out a Hail Mary and the man who caught it was none other than the notorious angel investor Steven Dengler.
Says Pobst: "I wrote [Dengler] a letter and I said, 'Hey. There aren't many people who are willing to invest in games. It looks like this is something you're doing. We've got a good team here. We're making, I think, a good product. Is this something you would be interested in?'
"He wrote me back within a few hours and he said, 'Look, I get about 200 of these a week. I ignore most of them, except the games that I know. I'm one of your backers. I like your game. Let's talk.'"
By the end of 2012 Dengler had been to Hidden Path, seen what it could do and was sold. Pobst crafted a business plan, Dengler invested the cash and Defense Grid 2 ... is still not yet a game. But it will be. Soon. (They hope.)
Even though Hidden Path has made plenty of games before, no one knows how this one is going to turn out in the end. This is game development, and it's a long, long road.
Waiting is hard
The road starts with John Daud.
Daud is the lead designer of DG2. He'll be largely responsible for how much better or worse the game is compared to its predecessor. What he wants to be doing more than anything right now is digging in to the level design tools and making playable sections of the game he can try out and use to test his theories.
What he's doing instead is writing. Lots and lots of writing.
Daud is writing design documents, mainly, but also emails, scripts, reports. He's writing so much it's starting to make his wife nervous. But he's writing because he's nervous and, for the moment, there's nothing else for him to do.
What he's doing instead is writing. Lots and lots of writing.
The level design tool Daud will use later in the year is mostly not yet built, and what is built is broken; building roads causes it to crash. Just a few hours after I left the studio one Monday, Daud experienced one such crash and flew into a fit of rage, yelling and throwing small objects at his computer screen. When I arrived the next day, the other members of his team were still laughing about it.
Daud wasn't laughing.
Now he's sitting with me in a room Hidden Path uses for telephone meetings, wearing a Seattle Seahawks hoodie (un-ironically, he assures me). Daud is addicted to sports, and football is his favorite. He enjoys watching football because it requires a level of physical prowess he will never possess, and that entertains him. But, more than that, he enjoys studying how it works.
NFL football is, at its core, a game. And the various rule changes that hard-core fans bitch and moan about, season after season, reveal a dynamic between the game and its audience that is extraordinarily similar to the dynamics of long-running multiplayer video games.
"And it's a real-life problem," says Daud. "You know why they're doing it — the safety stuff. It's not just to be benevolent. It's self-preservation. Lawsuits, and keeping your players alive. I get it. It's just ... " He makes a pained, groaning noise. It's similar to the noise he makes when he talks about how much game development still needs to happen before he can really get his hands dirty designing DG2.
"It's not just to be benevolent. It's self-preservation. Lawsuits, and keeping your players alive. I get it. It's just ... "
Daud has been with Hidden Path since 2007. Before that, he was an artist at an outfit called 5000ft, where he worked on, among other things, some of the "many" sequels in the Army Men franchise. His team at 5000ft was underfunded and understaffed, forcing Daud to learn to do a lot, quickly. Up to and including assuming the role of level designer, which, for an artist, is highly unusual.
"All the artists were doing level design," Daud says. "I think we each got an equal portion of the level design stuff. Which meant basic scripting and laying the levels out and then playtest. Lots of playtest."
For another Army Men sequel, the license holders wanted to "make the Army Men real."
"They wanted them to be more human and they wanted [players] to be able to ... customize the characters. ... That direction changed after a while. I think they realized, 'Wait a minute. This is plastic men. We can't make them real.'
"So that changed."
Time and again, Daud got drafted into design duties because of his ability to write and communicate. Which led to more writing and communicating. Which, eventually — after "the unfortunate Daredevil game," a Disney Princess game, a poker game and the Aqua Teen Hunger Force game — led him to Hidden Path.
Six years later, he's leading the charge on what Hidden Path hopes will be its most successful original game to date. Although, as of this writing, "leading the charge" means writing. Refining his ideas in design documents and then breaking those ideas into discrete tasks — in writing — for the rest of the team to begin working on.
"I pride myself on the [design] documents, even knowing that they're going to evolve," he tells me. "They always evolve. At some point during this process, I will just consider those documents the starting point. I'll update them as we go along, probably for the next few months, but then it's like, 'OK, we'll do this by email or me shouting across the room or putting it on the whiteboard.'
"At some point it switches over from becoming a design spec to a historical documentation of what the design is. That seems to be the natural flow of design."
I've seen the design documents. To say they are "comprehensive" would be an understatement. Whereas some video game design documents read like bullet-pointed wish lists, or half-baked theories cribbed from cocktail napkins, Daud's DG2 design docs read like a review of a game that already exists, and is playable.
For Defense Grid 2, Hidden Path will [utilize] a more advanced internal technology platform developed in-house over the last two years. ... This [Hidden Worlds] engine is based upon the original Defense Grid work and now incorporates deferred rendering, a procedural terrain engine. It runs on the PC at a higher performance on the same minimum requirements as the original Defense Grid.
This new tech base includes visual features such as ambient occlusion, hemispherical lighting, supports an almost infinite number of lights in the scene with deferred lighting ... allows for more striking and exotic locations as we take the player to new planets. ...
A new story with multiple characters will be the glue connecting the game levels as the player progresses through the game. ... There will be improvement in alien animation vocabulary as well as methods of alien death and destruction. While no actual physics will be used, visual reactions to the world will appear to take place by the aliens through a limited set of animations.
This is taken from the 2,000-word main design document, one of more than 12 documents that Daud has written outlining every aspect of the game's design, from story to monetization. It is an exhaustive treatise not just on the design of DG2, but on the current state of the game industry itself and how DG2 will fit within it.
Something only a person with a great deal of experience — and a great deal of time — could have created.
"The ideas have just been coming up for years," Daud tells me. "Either by Michael [Austin] and Jeff [Pobst] and Mark [Terrano]. ... Because [Defense Grid has] just sort of been there on the back burner for years, there have been a lot of ideas and a lot of documents. There are a bunch of pitch documents.
"Basically [my part] was taking all that stuff and making sense of it."
Speaking to Daud, he gives the impression of a race car spinning its wheels, waiting for the green light. Defense Grid 2 is all but designed in his head, and he's ready to start tinkering with his theories and tweaking them. Seeing how the new towers (not yet built) will work with the new aliens (still being conceptualized). Adding in multiplayer (which he admits is a radically new concept for Defense Grid) and tweaking the UI (user interface — which he admits was not great in the first game).
Daud talks about these things as if he is already working on them, which, in his head, he probably is, and has been since around the middle of 2012. But in reality these things are weeks away from him being able to touch.
In the meantime — in addition to writing — Daud is trying to spend as much time as he can with his family. When the game development gets into full swing for him, his time may be short. Although at Hidden Path, the team very rarely uses the word "crunch" and not because they're hiding from it; it's because they don't do it. And that's part of why the studio is able to attract such a talented team in the shadow of their much more heavily funded colleagues in the Seattle area.
"Those are my levels. The buck stopped here. And so ... Yeah. Yeah, I don't know what would have happened."
"That's one of the reasons we're here," Daud says.
Daud tells the story of how his second child was born almost immediately after Defense Grid: Containment, the Defense Grid expansion Kickstarted as part of the same campaign that launched DG2, shipped. Luckily, Daud's bosses knew his wife was expecting and designed their schedule assuming he'd be gone towards the end, but childbirth is like game development in that anything can happen. If the baby had come early, it would have been a tense moment.
"That would have put it into perspective, I think," Daud says. "Those are my levels. The buck stopped here. And so ... Yeah. Yeah, I don't know what would have happened."
But he's convinced it would never have come to someone asking him to miss the birth of his child. Not at Hidden Path.
"I don't think they would ever require us to do that," he says. "That would kill me. I can't even imagine that.
"I've missed weddings and funerals [at other studios]. I'd like to think that ... It may not have saved the game, but it probably saved my career, I think. If that's any consolation. That's usually how I think about it, but it still grates on me. I know a lot of people probably feel the same way. But birth? Oh. No. I think I'd have to draw the line. That's not cool." He laughs. "That is not cool."
Dacey Willoughby is sitting in the Hidden Path conference room, typing and mousing and staring at the massive screen on the wall. Sitting beside her is DG2's lead artist, Lex Story. They're haggling over an animator.
"The biggest thing is, they want a sample mid-game level," says Willoughby. By "they," she means the game's investor, Dengler. "We're going to need to have a level."
Story: "The level will comprise of ... ? Towers that animate, correct?"
Willoughby: "Yes. They're going to put the alien stuff here, in sprint seven, at the end of next month. I think it would make sense to just do as many mobs as we can." (Mobs means "mobiles," which is video game speak for things that move. Enemies. In this case, aliens.)
Story: "If they're not animating? Where's my animator?"
Willoughby: "I need to talk to Jeff [Pobst] about the animator."
Story: "That would be Ron, correct?"
Willoughby: "I think after we have this plan and I present it to him and say we need an animator, they might be able to switch his project. At least for a little bit. I don't know if you think this is enough time to get all of the mobs done. ... It's three weeks."
Story: "All of the mobs done? Where did you get that number?"
He's starting to get antsy. Willoughby takes a mental step back. She's watching Story carefully, waiting to see if he's going to flip out, or if he's slowly coming to the same realization she's already reached — that there's a lot to do and not quite enough time.
Willoughby: "What do you mean?"
Story: "That's like May. That's three weeks."
Willoughby: "Yeah, exactly. Is that too little time?"
And so it goes. For nearly 20 minutes, Willoughby and Story hammer out exactly what pieces of art will be required to create a level for Dengler, and exactly how Story and his small team will be able to make that art in the time they have.
It's clear from the volume in the room that Story hates meetings. He likes working, he tells me later, and meetings don't feel like working. Especially when they come in the middle of the day, or without warning. They break his flow, and it takes him a long time to get the gears moving again. Willoughby knows this. She prepped him in advance, telling him this morning that he had a meeting this afternoon.
"She's awesome for the job she does," Story says later. "Because a producer is somebody, when you see them and they're walking down the hallway and they greet you, you say, 'Oh, shit, did I fuck up something?' or 'Shit, is something due?' That's the kind of producer I want. She provides all of that. She's like, 'Look, I'm gonna make you very responsible for the crap you said you were gonna do.'"
Story has been through all of this before. His resume reads like a history of notable games: BattleTech 4.0. Return to Zork. MechWarrior 2, 3 and 4. MechCommander. Crimson Skies. Supreme Commander. Space Siege. Then, at Hidden Path: Defense Grid, Left 4 Dead 2 and Counter-Strike: GO. DG2, in other words, is not his first rodeo.
"I was one of the only guys who said, 'I want to go into games and show them 3D. Everybody was like ... 'There's no such thing.'"
Nor is game development his first career. Story came to the industry by way of an aborted career in the culinary arts. Prior to that, he was in the U.S. Marine Corps.
It was his military experience that defined him. Sitting anxiously across from me in a small meeting room, elbows on the table, camouflage ball cap perched on his head, knee bouncing, sandaled feet poised to propel him (if need be) bodily out of the chair, he unwinds a tale of someone with an almost Forrest Gump-like tendency to appear in some of the most transformative events since the 1980s, in and out of the game industry. Starting with the civil war in El Salvador.
"It was a sketchy, stupid part of the whole ... I mean, my commander in chief was 'Ronnie Raygun,'" Story says, as if this explains everything. And in a way, it probably does.
Story's unit was involved in a conflict most Americans don't even think about and that the U.S. Congress refused to acknowledge at the time. It was tasked with recovering weapons sold to Central American warlords by the U.S. government. By "Ronnie Raygun." "We were basically making deals with bad people because they were our bad people. It was very ... ambiguous."
Story was what's called an "NBC" instructor, standing for Nuclear Biological Chemical. Basically the worst possible stuff to have thrown at you in war, and he was responsible for training fellow Marines how to survive it. After a series of "full engagements," he began to wonder: If he died in El Salvador, would anyone ever know why?
Soon after, he left the Corps, but was recalled for the first Gulf War — back when everyone assumed Saddam would gas attack invading U.S. forces, forcing Marines like Story to run around in 120 degree heat wearing rubber suits and suffocating protective gear.
"We're brutally honest. We don't put numbers out there because that's what we think the publisher wants to hear."
"We had to deal with a lot of ... but anyways," he says. The conversation trails off. Story's leg bounces. There's more to this part of his life, but he's not interested in sharing it. Or maybe he's not allowed.
On to the next chapter.
After leaving the Marine Corps, Story studied chemical engineering, got bored, went to art school, applied for an internship at PBS and ended up working as a post-production computer graphics animator for the McNeil-Lehrer News Hour. He also helped with the famous Civil War miniseries by Ken Burns. The one with voice actors reading soldiers' letters over images of old photographs and wood carvings. The documentary that, in short, changed the way documentaries are made.
The more Story dug in to creating computer graphics for television shows, the more convinced he became that he could make 3D graphics for video games.
"I was one of the only guys who said, 'I want to go into games and show them 3D,'" Story says. "Everybody was like ... 'There's no such thing.'" Story decided to prove them wrong.
After leaving McNeil-Lehrer, he cut his game development teeth working at Location Based Entertainment, then FASA, the MechWarrior creator that was eventually sold to Microsoft and which formed the core of Microsoft Game Studios during the early days of the original Xbox. Then (after a brief detour to go to culinary school) he went to Chris Taylor's Gas Powered Games, where he helped with Supreme Commander and other titles. When his longtime friend Dave McCoy called up (they worked together at PBS and again at FASA) and said he was co-founding Hidden Path, Story didn't hesitate. He knew there was an opportunity to do things the "right" way, and not necessarily the cheap way or the popular way.
"We're brutally honest," Story says. "We don't put numbers out there because that's what we think the publisher wants to hear. We put those numbers out there because we know what we're capable of. We 'fess up to it. We're like, 'Hey, look, it's gonna take this much time. It might cost you a little more, but at least you're going to get exactly what you wanted in the time that we know we can deliver it in.'
"I would work at Activision and it wasn't always '[hire] the best person.' It was '[hire] the cheapest person.' 'Really? You brought this guy in because you saved a nickel? That nickel in the end is going to cost you more and more down the road.'"
Back in the meeting room with Willoughby, Story is concerned about what exactly each milestone requirement entails. Citing a bullet point that reads "core housing," he growls that that doesn't even need to be on the schedule. It's a simple piece of geometry, he says. And besides, someone is already working on it. Willoughby wants to be specific. She believes that other members of the team understand what a core housing is, and that they need one. She wants it on the schedule so people know when it's coming, whether it's simple or not.
The mood is tense, but focused. Willoughby takes it all in stride.
"I think that the production role here is about ... kind of catering to what the partners are looking for, information-wise, and [then] catering to the team," she tells me after the meeting. "Different team members look for information in different ways. Like, engineers usually just like lists of things, whereas other people like to look at spreadsheets with colorful information on them. Tailoring to everybody's needs."
In the conference room, Story pushes back: Levels are not made in pieces, in advance. He doesn't know what a level will look like until he starts working on it, then adding things. For Defense Grid, he didn't know a level would need alien tentacles until he looked at it and decided that it was missing something ... like alien tentacles. So putting "alien tentacles" on the schedule would have been impossible. Which is, he says, what Willoughby is asking him to do now.
Willoughby relents, agreeing to change the wording, but the bullet point remains on the schedule. Story agrees to hit the mark.
"With our schedule and the way this game has come together for its development, we don't know, really, how long it's going to take," he tells me after the meeting, in the small room, leg still bouncing. "We know it's going to be a shitload of stuff, but we don't know how long it's going to take.
"Until I actually make it, I'm really just blowing smoke up your ass, because I don't know. Just give me a chance to do it, one piece, and then come back and ask me how long it's going to take."
For Story, the goal with DG2 is to make the game feel like a comprehensive experience, like the action is really happening, or could really happen. He complains about the original Defense Grid's gun towers looking like "towers with guns on them." For DG2 he wants them to be war machines.
"Being shot at changes a lot."
"They're not these decorative pieces and then you happen to have a nasty bit on the beautiful tower," he says. "It's not that. When you deploy this thing ... shit's breaking loose. You're like, 'Oh, fuck, we haven't turned that thing on in thousands of years and now we're turning it on?' That's the feeling I want.
"The first game, it was very ... everything was a relic. It was this old weapon system that was reactivated. In this one ... no, these people forgot. They're very happy. They live in this beautiful place. They all sit around and watch the sunset every day. They don't worry about horrible things like traffic. Then all of a sudden their whole world is disrupted because this alien force is coming to attack them. It's turning the world upside down. It's turning the world ugly. So we have to employ these things that ... the general populace doesn't even know they exist. It's a completely different feeling."
It's a feeling that Story, having been that thing no one knew existed, understands perfectly well.
"Being shot at changes a lot," he tells me. "How you look at what normal life is or what you can cope with. ... Games are emotional for me. They should be. When you look at them casually as, 'Oh, yeah, I just wiped out a couple hundred thousand people,' it dehumanizes you a bit.
"It's still entertainment, but there are certain things that are told in stories ... like when you read Homer, The Iliad. You're like, 'Uh, that's not a casual thing.' It's not as beautiful as Brad Pitt running up and down the beach in his sandals. It's pretty fuckin' horrible. I'm trying to get there. I'm trying to be a good storyteller, and aren't all game developers trying to be good storytellers?"
It's the beginning of the day, on a Monday. A few hours before Willoughby and Story will haggle over architecture in front of the giant television screen in the Hidden Path conference room. I'm sitting in a chair stolen from beside Willoughby's desk in the bullpen area where the DG2 team works.
The room is a mishmash of odd desks and random chairs. You get the sense the furniture has been accrued over time, or based on a person's individual preference. It's dark in here, but not dreary. Japanese screens wall off certain desks from certain others. Bookshelves form walls. Lamps shed a little light — but not too much. This is a place where people work, but there is energy and life. It's "lived-in."
"Things can go bad" is programmer speak for "disaster."
Around me, in their own chairs, the members of the DG2 team are arrayed in a circle. They're running through a scrum meeting.
Willoughby starts: "Today we're going to meet and go over the plan for this coming milestone that we're starting. Then we'll talk to everybody and see if everybody agrees and we'll know what people want to do. Then Lex and I are going to meet and go over the articles, so we can nail down some more specific tasks for you guys."
Going clockwise around the room, members of the DG2 team chime in to state what they are working on, what their goals are for the week and what they expect to, or would like to, be working on next. This is how the project stays on track. When so much of each person's work involves sitting alone, working at a computer, it's easy for each to burrow so deep into their own rabbit hole that they could derail another person's part of the project. Or miss the mark for not knowing what the mark actually was.
"Once you get them all talking in front of each other, then they discover things that they would not have otherwise known," Willoughby tells me later. "Or something that they assumed that they knew is actually not 100 percent what they thought they knew. It's revealing, even though they don't think it will be."
After Willoughby kicks it off, the meeting turns to Story, who briefly updates the team on the art he's working on with his team. Then the meeting turns to Matt Johnson, also known as "Twig." He's the lead programmer and right now the production of the entire game is resting on his shoulders. He's working on the level editor that Lead Designer John Daud needs to be finished so that he can start building levels. Johnson, as of this moment, is the linchpin.
"John's having issues with the editor," Johnson says. "I want to spend some time making sure it's better for him."
"Is that the flashing thing still?" asks Story, referring to an earlier bug.
Johnson: "It's a new thing. And I want to fix our road-generation stuff. It's causing crashes."
He's the lead programmer and right now the production of the entire game is resting on his shoulders.
Johnson describes the problem to me later. It's something to do with "threaded processing," "race conditions" and "critical sections." I'm following, but just barely.
"Two things running concurrently, both trying to fiddle with the same part in memory," Johnson explains. "One is using it and the other one changes it out from underneath the thing that's using it, and it's no longer getting what it expects. Things can go bad."
"Things can go bad" is programmer speak for "disaster."
Here's how I understand it: Let's say you're using the game's level editor to draw a road. You start drawing a road, putting a start point in one spot and an end point in another, and the program dutifully draws your roads. Then, for some reason, it stops. The program crashes.
This, in general terms, is what's called a bug. Something in the program code causes the program to stop working. Part of Johnson's job is finding those bugs and eliminating them. That's what he's working on now.
The particular bug involving roads has to do with two separate program "threads," basically operations that are running simultaneously. One of the threads is what's called the "main game loop," which keeps track of all of the objects in the game, what they are doing and what they should look like. The other thread is the rendering thread, which creates the graphics. What's happening is the two threads are trying to use the same section of code at the same time, which they're not supposed to do. One is supposed to be able to lock down the section of code it wants to use, creating a "critical section," and the other is supposed to wait for the first to finish. But that's not happening.
For some reason the main game loop isn't giving up its critical sections. It's like a traffic signal that's always green on one street, and always red on the other. Eventually cars will pile up on the red light and something will break. The rendering loop, unable to do its job, is piled up at the red light. Which is causing the program to crash. A bug.
Johnson's been working on the problem for about a day. He's not concerned about fixing it (to him it's just code) but he's not sure exactly what the problem is yet, and he can't move on to anything else until he figures it out.
This is game development.
"You write some stuff and then it starts behaving ... not like you expected it to," Johnson says, summing up the complexities of code creation in typically terse and entirely factually correct programmer language.
Johnson is called "The Keeper of Defense Grid" by other members of the team. He's the only employee at Hidden Path who's worked on every iteration of the game, focusing on little else but Defense Grid for the seven or eight years it's been in existence. Updates, expansions, ports. Then Containment, the Kickstarted DLC, and now DG2.
Before becoming The Keeper of Defense Grid, Johnson worked at Microsoft, on Office. Specifically on what's called the "UEX," or user interface team. That was much different from working on games. On a program like Office, there may be hundreds of programmers, but each works on a specific section of the program and all of the decisions are made from above. With games, particularly at Hidden Path, someone like Johnson has the freedom to take ownership of a project and make decisions. To be creative.
"The coolest part, in my opinion, is just figuring out how all the pieces fit together."
Like most programmers, Johnson got a computer when he was young. He taught himself to code and made a text adventure game at around age eight. As he grew older, he'd spend time in arcades, falling in love with the classic cabinet games of the '80s: Galaga, Dig Dug, Pac-Man. He always knew he'd be a programmer when he grew up. He participated in a special program for computer sciences at his Minnesota high school. He got his first programming job while he was still in high school. Getting into games, though — that was a bonus.
"The coolest part, in my opinion, is just figuring out how all the pieces fit together," he says. "I like thinking about ... 'We need to create a system to do X.' OK. What does that system look like? How does that interface with all of the other stuff in the game? I like that part of thinking about how something works and designing it in your head. Just the code architecture. Writing it down or thinking about it and then going ahead and implementing it.
"When you're done, you have this little thing, and it's like, 'Hey, look, I created this! It's cool! It works!' I don't know if everybody sees it that way, but I see coding as kind of a creative outlet, like an artist paints a picture. They have something they can show off to people. Creating software feels the same way to me. You have this little bit of something that you came up with and created. It just works."
Back in the scrum, Willoughby wraps up the meeting and asks if anyone has anything impeding their individual progress. Anything "blocking" them.
"I'm blocking," says Johnson.
Willoughby turns serious.
"Who are you blocking?" she asks.
"No, I'm building blocks," he says, deadpan.
There is laughter. This is a common joke. The meeting breaks and everyone returns to their desks to plug away at making a game that will (maybe) come to life in about a year.
This is game development. (The story continues with part two.)
This story is part of a series covering the development of Defense Grid 2. To read additional installments, please visit the Making of Defense Grid 2 page. This series will continue into mid-2014, the projected launch date for Defense Grid 2.
Editing: Matt Leone, Charlie Hall
Design / Layout: Warren Schultheis, Matthew Sullivan
Illustrations: Warren Schultheis
Photograph: Hidden Path Entertainment