John Romero isn’t done with the first-person shooter genre he helped create

John Romero made a name for himself in game development as one of the original creators of Wolfenstein, Doom, and Quake. He helped establish the first-person shooter genre in the 1990s, and those games are now a multibillion-dolllar market. And Romero is still involved.

He recently teamed up with fellow id Software cofounder Adrian Carmack to make a shooter game called Blackroom. Their new startup, Night Work Games, launched a Kickstarter crowdfunding campaign and then shelved it after four days to create a more elaborate demo.

His announcement coincided with the reboot of Doom by Bethesda Softworks, which now owns id Software. I talked with Romero in a fireside chat at the Gamelab 2016 event in Barcelona last week. We roamed all over the place, touching on topics such as making games in Ireland and his feelings about Doom.

Romero and Carmack are creating their new game in Galway, Ireland, where Romero’s wife (another famous game developer) Brenda Romero is working. The Romero family is also working on a game designed by 11-year-old son Donovan, a quirky title dubbed Gun Man Taco Truck.

Romero released a brand new level for the original Doom earlier this year, a first in 21 years. That served its purpose of putting Romero back in the limelight and helping him hone his views on what it takes to make a fun level. That map, Tech Gone Bad, has had more than a million unique impressions.

Romero and Adrian Carmack cofounded id Software (along with Tom Hall, who is now at Glu Mobile, and John Carmack, who is at Oculus VR) in the early 1990s. The team churned out Commander Keen in 1990, Wolfenstein 3D in 1992, Doom in 1993, and Quake in 1996. Romero left in 1996 after the release of Quake, and Adrian Carmack (who is not related to John Carmack) left in 2005.

Here’s an edited transcript of our interview.

GamesBeat: You were there at the beginning of the multibillion-dollar first-person shooter market. How does it feel to be part of spawning a market that’s become so big?

John Romero: Multi-billion! That’s huge. We did not think that it would be this big. When we were making our games, we were just trying to make a cool game, and then make the next cool game that was better than the previous game. We didn’t think about them beyond what we were making. Then we’d make the new game.

Looking back, it makes sense. It was fun playing shooters, playing deathmatch. There were lots of people who wanted to play that kind of game, and also to make a game like it. I was excited to see other people making shooters, which is why we created the engine licensing business and license our tech out, so we could see more games like Heretic and Hexen from Raven. LucasArts made Dark Forces and Jedi Knight. Everybody was making shooters. It was great to see everyone doing that. Then you had tactical shooters and Battlefield. It just got massive. It’s great to see that the shooter is still the number one genre making money.

GamesBeat: It wouldn’t be that big if you guys didn’t make some of those decisions.

Romero: If we didn’t start with Hovertank One, which you probably don’t know. It was awful. Then we made Catacombs 3D, which was the first texture-mapped 3D shooter. It wasn’t very good either. A couple months before we started working on Wolfenstein–Wolfenstein is really where we nailed it, because we put guns in Wolfenstein. The other two games didn’t have that. That cemented it – not first-person fireballs, not first-person tank, just first-person shooter. Guns. That was the thing that made the difference in Wolfenstein.

GamesBeat: John has a hyperthymesia memory. He doesn’t forget anything. I don’t have that, which is why I’m looking at my questions on the laptop here. What do you think is mis-remembered about the games you worked on in those early days?

Romero: I’ve read some recent interviews with other people talking about stuff I’ve made. There’s a lot of wrong information, just from 20-plus years of time. People don’t write these things down. They remember them, and then they mis-remember them. I remember everything that happened, which is great for game design, because you always know if something’s been done in a game before. What I usually do is just write a post, or I contact the author if I know them and say, “Hey, this needs to change. This is what actually happened.” It just happens. People don’t have perfect memories.

GamesBeat: You’re working on Gunman Taco Truck. How’s that coming along?

Romero: It’s great. You probably have not heard of Gunman Taco Truck, but our nine-year-old–he’s now 11, but two years ago he came up with a great game idea. We decided that we should make it. It was such a good idea, and it would show Donovan that the idea of a game is actually more important than the discipline of creating a game.

If you start making a game that’s a bad idea, you’re wasting a lot of people’s time. You can be the best programmer in the world making the dumbest game idea and no one will know who you are. If you have a good level of programming and an amazing idea, you can make Minecraft. I wanted to show Donovan, “Yes, you need to learn a skill, but game design is where everything in a game comes from.” Put our money where our mouth is and say, “That’s a great idea. We’re going to make it and spend a lot of time and put a team behind this and do this professionally.”

We started doing it as a game jam. We did about a year’s worth of game jams. Now we’re working on it with one person full time, and I’m putting in as much work as I can. Gunman Taco Truck, you’re basically the last Mexican taco-truck driver in the United States. Seven nuclear bombs went off – this is all Donovan’s idea – because a scientist put a cup of coffee on the red button. Everything’s mutated, but you need to continue the family business. You need to get to Canada because they don’t have any taco trucks up there.

You start in San Diego and, in an FTL-like way, choose which city to drive toward. While you drive down the ward, there’s an action scene. Your taco truck has weapons on it. You blast everything in front of you away, and in a Ridiculous Fishing manner, it has the hovering meat and stuff you shoot. You pick that stuff up, because when you get to town there are all these survivors waiting for you to make tacos. That’s what you make tacos with, basically roadkill. You make roadkill tacos to get money to buy gas to finally make your way to the end. Along the way you’re collecting scrap metal and buying new trucks or upgrading your truck with scrap.

GamesBeat: How do you make this game the best that you can, but also stick to your 11-year-old’s design?

Romero: There’s a lot of really bad balance, because he doesn’t know how to balance stuff. We take the idea, what he wanted—he designed everything. We’re actually releasing the PDF of his design, with his drawings and his descriptions of what these things are. You see what it turned into in the game, this cool retro pixel art style. We take his stuff and we balance it correctly and ramp it and do all the things you need to get right in a game. But we start from his initial ideas and drawings and explanations.

GamesBeat: Your wife, Brenda, is also a game designer. How do you teach young kids how to design games?

Romero: I like to teach through programming. We’re teaching Donovan how to code in a language called Luo, which is really simple. Using Corona SDK, that lets us put stuff up on the screen quickly, so he doesn’t get bored doing a lot of calculations. In one line of code he can put something up on the screen, and in one other line he can move it around. We talk about, well, what are we going to do? What will we make today? And then we make it happen quickly, because Corona lets you do that.

It teaches him that there’s an idea and then there’s development. Game design happens in the journey to development. The initial idea is very important, but the development of that idea is what game design is all about. He learns how things change as you’re building them. The more experience you have doing that, the better you can pre-think your game up front.

I like to teach that way. Brenda likes to teach using board game mechanics. She has her classes create board games or small D&D combat encounters, very simple characters, just within a few hours. When you get down to the mechanics of it, or down to the programming and implementation of it, that’s where people get to understand what it’s like to develop. Getting them into that space is great for them to see whether they like it or not. It may be a future career for them, or just something they learn how to do and then move on.

GamesBeat: You have this game design family – you, your wife, your sons. What is that like? Do you ever turn games off in the house?

Romero: Nope. [Laughter] Especially where Donovan’s concerned. Games are the only thing he does. If he’s not playing games, he’s watching YouTubes of people playing games. Games are everywhere. They’re on our phones. Our daughter Avalon, his twin, she plays Geometry Dash all the time. She makes lots of levels, uploads levels. We think she might be a level designer. Our older daughter is becoming a writer and game designer. She does a ton of writing. She’s written a game that she wants to make this summer with Brenda. And our oldest son is 28. He’s in the industry. He just finished releasing the Gordon Ramsay Diner Dash game that just came out yesterday. He has about four games going at once, usually. We’re all coding away, designing away.

GamesBeat: You may have missed Brent Bushnell’s talk this morning, but he was talking about seeking random influence. He’d go to a dentists’ convention just to get inspired about something. You guys are completely immersed in games. What do you do if you want to find something more creative?

Romero: Go play indie games. Or just think a little. A lot of great design comes from looking at what’s out there and figuring out where the next step would be. It starts with really funny ideas. Going to a dentists’ convention, going and talking to people—a lot of our ideas just come from talking to people and seeing things that are off the wall. “That’s a system. We could build some mechanics around that.”

Ideas really do come from everywhere. Reading the news. That’s tons of information you can build designs around.

GamesBeat: Peter Moore talked about 10,000 hours. I think you must be around 100,000 or so over 20 years.

Romero: Yeah, it’s over that. It’s been 37 years. I think I got about 87,000 hours in the ‘80s.

GamesBeat: There’s no such thing as drowning in games for you, then.

Romero: There probably is. I don’t play every single thing out there. If it’s on the radar, if it’s something I know is good from various weird ways that I hear about things, then I know I should play it. Then I play it. But I don’t have this giant pile of stuff I absolutely have to get through, just because—You need space. You can’t just be constantly looking at new things all the time.

If you’re going to be a consumer, that’s great, because you could do that. But I almost never function as just a consumer. I’m always a designer, always thinking about what I’m playing. What kind of idea could I inspire? It’s pretty rare to play a game and not be thinking about how it was made or what they were thinking while they made it or where they got their ideas.

GamesBeat: With the new Doom coming out, I thought it’d be interesting to tell people more about the original Doom. What’s memorable about it? What kind of contrast do you see there?

Romero: This is a bit funny. When we were thinking about Doom at the end of 1992, we’d just finished making Spirit Destiny, the sequel to Wolfenstein. We were thinking about what we could do that was even better. We had a modified version of the Wolfenstein engine that we’d licensed to EA to make a game called ShadowCaster that Raven created. It was better than the Wolfenstein engine, but it wasn’t as cool as the Doom engine was going to be. Because we saw that it was not as interesting as we were hoping to make a game—there were too many dead ends in the architecture that we couldn’t get past. We needed to create a completely different engine with different capabilities that could look better.

John Carmack was thinking about the Doom engine for at least three months before the beginning of 1993. We were starting to think about the design of Doom during November and December of 1992. In January of 1993 we knew what we were going to make. We knew everything about the game, pretty much. We knew what we wanted to make.

We didn’t know what it would turn out to be, just because whenever you’re creating something, the process of that iterative development cycle brings the game out. It’s not what you call a waterfall, where you just design something and make it happen and magically it’s fun at the end. We’d do what you call an organic development method, where we come up with the core that we want to create and then everything grows from there. Then we’d have an idea of how much there should be over time, and we’d solidify that.

With Doom it was really a theme. We knew what the technology was going to be. We knew we’d put multiplayer in there. We knew it’d be fun. We put out a press release in January that said, “We’re starting to make the best game in the world.” We listed the reasons why. “This is going to be the biggest loss of productivity on earth.” You can look up that press release online. We put it out before we even started working on the game, which was crazy. Don’t ever do that. It’s not a good idea. But we knew we were going to do something great.

And that’s what we did. We made the game over one year. We put in multiplayer during the last three months. It was not a crunch mode game. It wasn’t like we lived in the office making it. I got married and went on a honeymoon in the middle. We stopped making Doom for about three or four weeks because we needed to make the Super Nintendo Wolfenstein. We stopped work on Doom, learned how to program the Super Nintendo, and ported Wolfenstein over in three weeks. Then we got back on Doom. Lots of things happened in the middle.

But we knew what were making. The only issues that came up in the development of Doom were—there was the development of the levels, what they were going to look like, because when we started making the game our levels looked like Wolfenstein. The same heights, bright lighting, 90-degree walls. And then I had to come up with a level design style that I called abstract level design. A rule at that time was, if we were creating an area in Doom, it needed to not look like it could have been made in Wolfenstein. I used that same rule for Quake. This room should not look like it could have been made in Doom. That was the minimum bar.

Creating the level design style was a problem for the first four months of making Doom. Then, in April, I came up with the way of creating levels, and that was it. The other part that was an issue, we didn’t have binary space partitions, which speeds up the calling of the world and the draw order of everything. We were using just lists that we would traverse. It was slow when I was building certain types of geometry in Doom. John needed to solve that problem. He was reading a white paper written by Bruce Naylor at Bell Labs. He had created this data structure, binary space partitions, that was made for 3D models, pulling back the polygons off a 3D model. John took that idea and made it work for a level. That made the levels extremely fast.

GamesBeat: I remember that Doom gave me nightmares. There’s one particular kind of scene that produced them, which was — I’d go into a dark place and see nothing but darkness. Then I’d hear this creature start growling, and this fireball comes out of the darkness right at me.

Romero: We were doing that while we were making the game. There was a point in Doom’s development, when the AI started getting in and we started getting sound effects in—the levels were so dark and scary. When we played through the game ourselves, while making our own levels and playing our own levels, we didn’t know how the enemies would react. It was as scary for us as it was for anybody who was playing it for the first time. We knew this game was going to get a reaction, because we were making it, and we were still reacting to it. We’d hear the sounds and get scared and think, “Oh, this is so perfect.”

We were very excited to see how the game’s AI was functioning in the geometry we were creating. By putting darkness in there, which also helps the game be creepy as the light gets darker down the hallway, the whole thing gets very scary. Scarier than any game that had come out before. Wolfenstein was all bright lighting. It really did make a big impact, with the way the AI worked and the darkness and the atmosphere.

GamesBeat: Fast forwarding to now, you did have a very horror-focused Doom 3 that came out in 2004. And now the style is much brighter again. There’s almost this ubiquitous orange. It’s a very different art style. I wonder what reaction you have on a high level to where they’ve gone with Doom.

Romero: I think it looks great. I haven’t actually played the new Doom, because I’m a Mac user. I don’t have a Windows machine. We’ll have one when we need to test a game, but we basically live on Macs. I don’t play console shooters because I don’t like the controls. If the Mac version comes out I’ll play it. Otherwise I’ll have to get a Windows computer that can play Witcher 3 and all the new stuff.

GamesBeat: The environments are lighter and much larger. And there’s a story overlaid on top of it – reproduced scenes and things you can find to tell you more about the backstory. That seems a bit at odds with the legacy of Doom, which is a pure action game. What do you think about that trend toward putting story on top of shooters?

Romero: Oh, I think it’s great. Games have been narratively driven forever. Adventure games started everything, this whole narrative. When we made our shooters, we weren’t spending time on story. We did have some screens in Wolfenstein that told the story of what you were going to be playing, but we didn’t do that in the game. It was outside the game. Doom had a story in the first couple pages of the manual. Otherwise it’s an implied narrative as you play through the game.

But I love games that have narrative them. Half-Life 2 was an example of games doing narrative in a great way. It doesn’t take control away from me. I can sit there and characters can talk and give me information and forward the story, but I can walk around and do whatever I want within that space, not feeling like my head’s being pushed toward some character to listen to them. I like that way of telling a story.

In the new Doom there’s a part where you have to go find a guy who’s dead, take his hand, and put it on a scanner. It’s a little sequence that you don’t have to watch – you see the ghost image of this action replaying from the past – and you don’t have to follow it if you don’t want to. But if you don’t you won’t see, “Oh, here’s the guy.” That sequence was taken from the Doom bible, which we wrote in November of 1992. The captain’s hand was a key for a lock back then. You had to rip the captain’s hand off to put it on the scanner. The Doom team was trying to be as faithful to everything in the original Doom as possible, so they put that in there as a reference to something in the original design doc that we never actually put in the game.

They did a great job. Single-player looks great. Lots of really good feedback from players saying that it felt like the real Doom. I know that they spent time on knowing how long it took to shoot at an enemy, how long it took it in the original one, and comparing the experience to make sure they were upholding the core values of Doom. I’m excited to play it when I can.

GamesBeat: It seems like game designers still have to strike some kind of balance there. There’s an expectation of nonstop action from gamers, but this need to be more mature and tell a story — I had that reaction to the God of War demo Sony had at the most recent E3. God of War is a hack-and-slash game, but this demo is one of the best examples I’ve ever seen of a father-son coming-of-age story.

Romero: Everything is telling a story now, yeah. It’s an interesting trend, if it doesn’t take away from the core values of the game. If what you’re doing is fun and you don’t feel like you’re sitting through QTEs all the time—if you like QTEs, you’ll love most games. But if you like where you’re in control and you’re experiencing a narrative that is not overbearing, not changing how you want to play the game, people are very accepting of story in pretty much any genre now.

GamesBeat: We’re at the 20th anniversary of Quake, and it’s being remade as well. One of the most fun moments at the Bethesda conference was when they showed a black DOS screen, and then they created a directory and typed in the letters one by one to execute Quake.

Romero: I never saw that presentation, but that’s a great idea. Quake did a lot of things. It’s funny, if you go back. Wolfenstein was, “Here’s a first-person shooter. This is what kind of game it is.” Doom was the much better-designed first-person shooter that had all this other stuff in it. You could mod it. It had multiplayer deathmatch. There were tournaments back then. It was the rise of LAN parties and all that.

When Quake came out, now mouselook is in games. You can use the mouse to look around. Internet multiplayer got huge. Clans created esports, which is massive now. Full 3D now. This is back before video cards with GPUs and 3D APIs. Quake did a ton. It reflects more of the modern types of games than Doom does. You could look at a game now and say, “That traces back to Quake.” Some things trace back to Doom, but really, Quake came from Doom.

The 20th anniversary is pretty cool, seeing that the franchise is still going on with the fifth one. They’ll probably just call it Quake. Or Quake Champions, right? I heard a little while ago that they’re planning on doing a reboot of the original Quake, with the H. P. Lovecraft designs. The whole franchise changed multiple times over its lifetime. But it sounds like Quake Champions is going to be the new thing, the multiplayer.

GamesBeat: They haven’t described it very much. They just showed one trailer.

Romero: It’s going to be an Overwatch, LawBreakers type of thing, possibly. Coming from Quake Live, I would expect something competitive like that. But anyway, it’s great. 20 years and the game still survives. Those were the three big franchises I worked on. It’s cool to see that all of them are still alive today, with Wolfenstein: New Order being the last Wolfenstein not long ago. My guess is they’ll keep on making sequels.

GamesBeat: People sometimes say shooters are kind of exhausted, but you’re working on a new shooter, Blackroom. Can you talk more about that?

Romero: Since Daikatana, which shipped in 2000, I made a—what is it, Red Faction for the N-Gage in 2003? But that was just a port of someone else’s game, something to do with our mobile game company. We were putting out a bunch of mobile games. Shooter-wise, I haven’t actually made a shooter since Daikatana.

I figured, if I’m going to make another shooter, I need to do something that’s more innovative. There has to be a good reason for it. It needs to be a better reason than just a story that people haven’t heard. It needs to be multifaceted. Blackroom’s design is more enabling of modding. It takes into account the modding culture, and also the speedrunning culture. It puts them all in a narrative and a fiction that I haven’t seen before in a game.

I think it’ll make an exciting experience in lots of ways. It’ll allow a roller coaster of design. It’ll allow very surprising gameplay, very scary and very funny gameplay. It’ll be interesting to see how turns out how we’re making it. Right now it’s an interesting idea to build a game in. But I’m not pre-thinking the entire thing. I want to see what comes out of it as we’re making it.

GamesBeat: There’s this thing called a Boxel.

Romero: In the game you’re a chief engineer of a company that creates holographic simulations. You have a piece of equipment on your wrist called a Boxel, this box that controls the hoxel tech, the holographic technology in this thing called a Blackroom. With the Boxel you can activate the UI. The UI is in the space, in the world. It lets players see the construction of game worlds through a holographic lens. You’re in a holographic sim, so it has a special shader that makes it look a little different. The game lets you use the Boxel UI to visualize what this world is made up of, the holographic world.

You’re trying to solve some issues. This new type of technology went online that’s starting to mess up the simulations. You’re trying to find out where it’s messed up and debug it while you’re in the simulation. What you find out is there’s a lot of really crazy stuff happening while you’re in there. There’s a technology called predictive memory, for very advanced AI, that creates AI from past memories. In the game, your memories are read and AI is constructed around your memories. Then it tries to predict, in the future, the evolution of those entities, so they’re not frozen in time with the knowledge they have. They can go forward.

The story of the game is solving a really bad feedback loop of predictive memory tech when your memories are represented in a messed-up way. But then that becomes the new memory and it feeds back and gets worse. It has an interesting idea. The monsters are crazy. The settings are all over the place.

GamesBeat: You could be working on a Battlefield-sized team with 100 people, and yet you’re making this with a very small team. Why did you go that route?

Romero: I like small teams. Teams that are small, that have very experienced members, can get a lot more done than a big hierarchy. When you have a small team, everybody is very responsible for the entire game, as opposed to a game team that could be 40-plus people and everyone can think, “Well, maybe that guy will take care of it.” All the layers of management and meetings and the time that gets eaten up with big teams—in a small team you’re in the same room and you can just look over and ask questions. You get stuff done quickly. It’s a more fun atmosphere. You’re hopefully with friends, instead of having a gigantic team.

I worked on an MMORPG for four years. I had 100 people working on this game every single day. That’s so many people. It’s just crazy. And it’s not as much fun. I don’t like directing games. I like making games. I wanted to be on a small team so I could make this game, not just tell people how to make it.

GamesBeat: You’re working on this with Adrian Carmack, one of the original id staffers.

Romero: He’s the guy who set the look of first-person shooters at the very beginning. It’s great. He’s still a great artist. He’s the same Adrian. He’s very quiet. You don’t want to interview him. He doesn’t want to talk much. He just wants to do his art. Same guy, same crazy ideas, same excellent ability for art.

GamesBeat: Do you think of the id team as kind of like a band, where everyone’s gone on to do their things over the decades? John Carmack is at Oculus. Tom Hall just finished this Gordon Ramsay game.

Romero: Tom and I co-founded three companies together. Tom and I really stuck together for a while after id. We co-founded Ion Storm. Ion Storm lived for eight years. Right after we left Ion Storm we co-founded MonkeyStone Games, and then later—MonkeyStone stopped in 2005. Six years later, when I started Loot Drop, he came back and worked with me at Loot Drop for a couple of years. Tom and I keep in touch. Adrian and I keep in touch.

Adrian stopped being in games after Doom 3. But he’s excited to get back in games. Everybody else has been in games the whole time.

GamesBeat: And now you’re making this game in Ireland. What’s that been like?

Romero: It’s great. I moved to Ireland from the U.S. I’ve spent my entire life in the U.S., other than a little bit of time in Europe back in the ‘80s. I’ve lived so many places in the U.S. There’s not really a place where I’d say, “I want to move there really bad.”

My wife is of Irish descent. We came to Ireland and I’d never been there before. I was just hugely impressed by how beautiful the country was, the friendliness of the people, and the emerging game dev scene there, from Galway to Dublin. There were so many game developers in Ireland that we thought this could be a great emerging place we could feel more of a sense of community than, say, Silicon Valley, where nobody can talk about the games they’re working on.

We do lots of game jams and meetups and monthly meetings in Dublin. Lots of conferences. It’s very alive with game devs. We’re really excited to be there. The country has six million people in it and 12 million come to visit every year. Tourism is massive.

GamesBeat: Is it the next big gaming epicenter?

Romero: Hopefully it is. Especially Galway. Dublin already has several companies. [One investor from] Culver City has invested in a company called Digit in Dublin. There’s money coming from the U.S. into Ireland. We’re also working on doing the same thing.

GamesBeat: It’s surprising to look at Dallas and the economy that’s built up around games there, because of id.

Romero: There wasn’t really anything down there when we started, yeah. Apogee was down there, and we went there because Apogee was there. Then a lot of companies sprang up around Dallas.

Question: You talked about teaching game design through practice, but working on your own games, you prefer small teams with experienced people. What do you recommend to people who are finishing their studies and looking for someplace to grow as game developers?

Romero: There’s a lot of game programs graduating thousands of students now. Of course you’re graduating to get a job. Small studios like ours, because we’ve been doing this so long, we want to stay small, but we want to stay super capable. We only hire senior people. But lots of other studios hire juniors, especially the big studios like EA and Ubisoft. They have to have these large teams and you can’t pay a large team full of seniors. You need to hire juniors and mentor them. They can come up through a big structure. That’s usually the best place, unless you start your own indie studio at night. That’s the best way into the industry.

In Ireland, in schools, they have a thing called TY, the transition year. Students take a whole year off and go work. They do this two years before graduating secondary school. And then the colleges also have work placement, where they have students interning in a company where they hope to work. We have two interns at our company right now, and they’re great. But they’re still learning.

When we start getting to full production and hiring for our team, we usually hire people that are very experienced. If I have to, I’ll relocate them from other countries. There aren’t going to be a lot of people, but it’s important to have people who are good at making games. We had that issue on Daikatana, where nobody on the team had ever made a game before.

Question: What did you feel about id Software moving on without you, especially since it’s gone on to become such a big company?

Romero: Quake came out June 22 of 1996. In November of 1995 we had a big meeting about the future direction of Quake. In that meeting, I basically saw that the team was burned out. We treated the company in the way we needed to treat it, because we didn’t know the tech was going to take so long. I knew at that point — when we decided to not do a really new style of game, but to fall back on Doom-style weaponry and goals – that game design was kind of dead there for me. I decided I would leave right after Quake.

Just a couple of months later, I was talking to Tom Hall, who was working over at 3D Realms, and I told him, “After Quake’s done, let’s start our next company.” And so Tom said, “All right. I’m working on Prey. Whenever you’re ready we’ll go.” When I finished mastering Quake and uploaded it to the net and got the commercial release done, I started talking to publishers about starting Ion Storm. One of the publishers called back to id and told them I was leaving. The next day they said, “Time for your resignation.”

I was fine with it because I was leaving anyway. I’d already had nine months of, “I’ll be leaving all this stuff here. All the awards, all the IP and all that stuff.” Because of creative differences. I still would have done that anyway. If I could go back and change anything I wouldn’t change that, other than I would have changed how we were developing so it wouldn’t have happened.

If we could have split the company into two teams – one team using the Doom technology and one team creating new technology, no matter how long it took – we could have kept making games and been excited and refreshed. When it came time to use the new engine we could have switched over and started making new stuff that way, instead of creating this game we thought was going to be released at some point, but could never get released because the technology took so long to release that people burned out. That created the permanent fracture.

When I left in August of 1996, within six months half the company had left. They were left with six or seven people. And then they built up after that to go and create sequels to the games we’d made.

GamesBeat: It’s almost like a property that resembles Star Wars. If you tell a filmmaker, “Hey, you have a chance to work on Star Wars,” it doesn’t matter that the people who originally made Star Wars aren’t there. The same seems to hold true for Doom and Quake.

Romero: The idea of those games is what they’re excited about. They’re excited to create a new version of that. They’re not thinking, usually, if they’re going to work on Doom, that they’ll make something completely not like Doom. They can’t wait to work on Doom. It’s always trying to make the best version of that idea.

Question: With all of your experience in first-person games, what do you think about VR at this stage? And what do you play yourself these days?

Romero: The best VR games, we have yet to see them. Games in VR need to be made only for VR. Ports of current games aren’t going to be the killer app for VR. As has been demonstrated by people working in VR, porting these 3D game experiences, shooters and so on—people are getting sick playing them. There’s a disconnect there. They’re finding that stationary games are really what works. Mouselook with your head is good, but when there’s motion of your entity as well, people start to feel queasy.

The best programmers in the whole world are working on VR to make it happen. The art form is going to be specifically made for VR. It’ll be different than games that we play on consoles and PC today. I don’t know what those games will be. I think they’ll have to go beyond being stationary and doing something. But I don’t know what that will be. We’ll have to develop in it for a while to evolve that art form.

As far as games I play, I’m going to be playing Firewatch soon. I really want to play that. I played the Beginner’s Guide recently, which I thought was incredible, like the Stanley Parable. I played World of Warcraft every day for five years, at least six hours a day. Even longer on weekends. Thousands and thousands of hours in WoW, with all 10 characters and full sets of everything. I love MMOs. I love Minecraft.

I do love games I can spend hours and hours with, but I also love games where I can just spend a few minutes. Like Drop Seven, which nobody here has heard of, probably. Drop Seven is a really great game. It’s still out there. You can still download it.

Some things I hold on to and keep playing. The original Doom, there are still deathmatch tournaments going on. I was just in a double elimination tournament last month in Doom. Three years ago, at GDC, there was a full week of me just playing Doom all day long in public with a bunch of computers. People were deathmatching and the winner would play me. One person beat me by one point. But normally nobody beats me. Whenever we have these tournaments, the winner is the person who gets to play me at the end.

FTL is one of my favorite games. I loved FTL. I listen to the music all the time. That’s true of a lot of games I’ve spent a lot of time with, like Duke Nukem 3D. I listened to the soundtrack for Duke Nukem for 10 years, probably, every day. Loved that music.

GamesBeat: Can you take lasting psychological damage that way?

Romero: It’s funny. The music was made in loops, and people operate in loops really well. Think about the various designs on Facebook that are all about that gratification loop. The music of Duke also has a lot of really good loops in it. It’s easy to make long songs and you copy-paste stuff, but Bobby Prince and Lee Jackson, they did a really great job making that music. I wouldn’t listen to it all day long. But every day I’d listen to the whole album.

Question: There was a feature online published recently comparing a Doom level with a Final Fantasy XIII scenario, which was just a long corridor with cutscenes down it. What do you think about how level design has evolved since your early days?

Romero: The funny thing is that they took the worst example of a level to illustrate a difference. In Call of Duty, for a while there were a lot of on-rails parts of it. That little line going like this—basically you’re in a jeep and you’re watching a movie as you go through the scene. But yeah, level design got much more simplified, because they’re making these big games with tons of art in them, and if you have too much geometry to explore, that’s a lot more art and a lot of QA to make sure you’re not going to get stuck in a bad location. Modern games will put collision walls up everywhere to keep the player from going out there and causing issues, to keep them in the corridor going toward the exit.

There are still people buying tons of Call of Duty, so you can’t say it’s a bad thing when people are buying the hell out of it. But I do think there should be a variety. I don’t think that should be the overriding design style. I like games where I can explore. World of Warcraft is a great example. I can’t rocket jump, but I can run up the sides of mountains and slide back down. It lets me go all over the place, even if I can’t get over a mountain to someplace I’m not supposed to be in. The game lets me explore, and that’s the kind of thing I like.

I like to make levels in games that let the player explore, and then reward them with hidden stuff — the way Half-Life hid the lambdas all over the place, where all the cool secret areas were. I like explorative design. It makes me feel like I’m playing Zelda or something. That’s what interests me.

GamesBeat: And we’ll be back again for the 20th anniversary of Blackroom.

Romero: That’s right.

Disclosure: The organizers of Gamelab paid my way to Barcelona. Our coverage remains objective.