Skip to main content

Perfect Dark: the oral history of an N64 classic

In far sight.

It was never meant to take as long as it did. As far as the GoldenEye team were concerned, Perfect Dark should have come out a year or two after their seminal console first-person shooter, a quick-fire follow-up to one of the greatest games ever made. But it wasn't long before trouble knocked on the door of Rare's countryside farmhouse in Twycross. First, Martin Hollis, the genius programmer who led the GoldenEye team to stardom on the Nintendo 64, left the company at which he had become a legend. His acrimonious exit set off a chain reaction that led to the Free Radicals - Dr. David Doak, Karl Hilton, Steve Ellis and Graeme Norgate - walking out soon after to form their own studio. Those who remained were left to pick up the pieces. Struggling to cram a game bursting at the seams with ambition into the Nintendo 64's tiny memory limit, the developers of Perfect Dark achieved what once looked impossible: the highest-rated Rare game of all time.

20 years after Perfect Dark was released, I asked 10 of its chief creators, as well as then Nintendo of America producer Ken Lobb (Rare co-founders Tim and Chris Stamper could not be reached), what it was like to make. Perfect Dark was a visual feast when it launched in May 2000, pushing the N64 so hard it required the console's 4MB expansion pak for everything but a stripped down multiplayer. Its Blade Runner-influenced graphics and sound, its super cool spy heroine Joanna Dark, and its X-Files-inspired story won plaudits from N64 owners and reviewers - despite the shocking framerate. Though it seemed Rare could do no wrong in the '90s, with N64 mega hit after mega hit seemingly effortlessly spewed forth from the company's secretive headquarters in the English countryside, developing Perfect Dark was anything but easy.

"Turmoil is the only way I can describe it," says Duncan Botwood, a game designer who was part of the original GoldenEye team and who stayed with Perfect Dark until the bitter end. "There was turmoil everywhere."

Here, in the Perfect Dark developers' own words, is the behind the scenes story of how Rare's troubled shooter became a masterpiece.

Watch on YouTube

In the summer of 1997, before GoldenEye 007 had even launched on the N64, Martin Hollis and David Doak began plotting what they would do next. The team wanted to do something different, something of their own making. Talk of a Bond sequel, a Tomorrow Never Dies game, emerged, but, so the story goes, Nintendo was outbid for the licence by Electronic Arts.

The GoldenEye team collecting an AIAS award in 1998. Image credit David Doak.

MARTIN HOLLIS (team lead): From my desk, I had a call come in from Simon Farmer, head of production at Rare, to ask if we were interested in doing a sequel. You know, straight up. We thought about this for a day or two, and we replied to him to say no, and that was the last we ever heard of doing a Bond sequel. I'm surprised in retrospect because Nintendo made so much money from the game you would have thought they would have put more pressure or at least made more encouraging noises towards Rare to try and persuade them to do a sequel in the same line so they could have a similarly financially successful second product.

But after myself and the team saying no, I didn't hear anything more about it, and they respected our choice to make a different style of game.

DAVID DOAK (lead designer): The team went to the set of Tomorrow Never Dies, although by the time we went to the set, we had decided we didn't want to do it. It was where the Harry Potter World place is now. There's a shanty town thing in Tomorrow Never Dies. Brosnan and Michelle Yeoh were there that day. We didn't get to meet them. We were just some scruffy oiks standing off-set. I remember very clearly, they were shooting this scene while we were there. It got to five o'clock or something, and they were trying to get some sequence done, and it wasn't getting done. We were watching it, and at some point some guy who was some kind of lighting representative or something, just said okay, right we're closing the set, because it was going home time. We were like, fucking hell, games don't work like that! You go home when it's done. It was them just going, it's union rules. Time to go. The thing that closed the set on that day was not a director. It was some guy saying, my guys have to go home.

KEN LOBB (Rare producer at Nintendo of America): When GoldenEye 007 finished, after much celebration, and a great many multiplayer sessions, thoughts of what would be next began. Nintendo would typically let Rare do what they wanted, but in this case, there was a short period of time when the sequel was assumed. We went and read the script (at the London offices of some lawyer, literally reading through the script, sitting in a conference room, with no ability to take notes!). The script was good, but at the same time, it felt like the team wanted more. More control of the story, more ability to just build their idea of a perfect shooter. One which likely wasn't going to fit perfectly into anyone else's script.

Meanwhile, the negotiations had started, and because GoldenEye 007 was already doing very well, and our friends at Danjaq and Eon Productions were of course convinced the games' success was all about the power of the Bond licence... Although that clearly helped a lot, it's worth noting GoldenEye 007 sold far more copies in the second and third years than it had at the time of this negotiation. In my opinion, the game benefited from the licence for the first few months; then the success was all about word of mouth.

Anyway, the opening bid from the licence holders was far more than was paid as a minimum guarantee for GoldenEye 007. We knew of another bidder (though we didn't know who at the time), so that led to a very quick and easy conversation with Tim and Chris Stamper, and Martin Hollis. Do you want to do Tomorrow Never Dies? Nope, they would rather make their own game as a spiritual sequel. Sorry Mr. Bond, but your time at Rare was over.

HOLLIS: GoldenEye was three years long. You know, it was a very intense development, and we spent a lot of time in the Bond universe. And really, we just had enough of that, and we wanted to stretch our legs and try out some fresh ideas.

Perfect Dark began life. Its world was fleshed out, its story outlined, its main character settled upon.

HOLLIS: Well, really it was me who wanted to have a woman as the star because of various spy characters I knew. I thought women were underused in video games.

Joanna Dark concept art by Brett Jones. 'I wanted somebody more human than this Lara Croft character we had seen all the time. And that was the idea behind Joanna Dark, to make a female James Bond character who had all of his skills and resources, but was female and very capable.'

BRETT JONES (animator): The idea was to do something that was the antithesis of Lara Croft. Although she was incredibly successful, she was a bit two dimensional. We wanted a female heroine with a bit more pizzazz and snap to her. Dr. Doak came up with Joanna Dark, which is from Joan of Arc, Jeanne d'Arc being the French, so that's where the name comes from.

HOLLIS: It's a pun, which is always a glorious thing. It's a cross-lingual pun, which is perhaps the most painful possible! All I could really say that we got from the idea of Joan of Arc was that she was fairly young. She was a woman. And she was warlike. She was involved in military violence. And that was really all we took from Joan of Arc.

But it's one of the many ingredients. Probably more than 20 different characters helped illuminate and give support to this character.

DUNCAN BOTWOOD (game designer): The name was a massive stickler. Just the name of your protagonist... so many discussions going around and around. Oh my god. And then eventually it was like, yeah, some pun on Joan of Arc. Fucking ship it. It was like, yep, I am minimally offended by that choice. So we went with that.

DOAK: The obvious thing is to do sci-fi, really, because that allows you to make stuff up and to do all the crazy stuff. In our office, Martin and I had a La Femme Nikita poster on the wall. If you look at something every day, it probably worms its way into your head.

Okay, let's make it something that isn't Bond, but it's going to be some kind of operative. For a while, Covert Ops was the working title, which embraces that. From very early on we were interested in having a female character, not just because of the Nikita thing, but also some sci-fi stuff I liked had female characters. There's a Robert Heinlein book called Friday, which has got a female assassin operative character on the front of it. If you spoke to Karl, he said Blade Runner. Flying cars, all that kind of stuff.

KARL HILTON (lead artist): The DataDyne office, which has got the big X on the roof, and had got buildings through the window with the moon behind them - that was my tribute to Blade Runner. That was the Tyrell building in the distance - the angled building with the sun behind it in Blade Runner. I had the moon behind it instead.

Perfect Dark was initially developed in Rare's first location at Manor Farmhouse in Twycross, Leicestershire. Image credit Steve Malpass.

DOAK: Covert Ops became Alien Intelligence when we decided we were going to have aliens in the story. Trying to name it was hilarious. In the end, Martin and I had a random word mixer. It had a database of 200 words, and it just used to run and spit out names. And we'd go, oh, we like that one. The test of a name was, if you printed it out on a piece of paper and stuck it to the wall and you didn't hate it in two days' time, then maybe that was okay. Perfect Dark came out of that because dark and perfect were two of the words that were in there.

LOBB: I loved the setting - cyberpunk meets "the Greys" - loved the idea of Joanna Dark as the heroine, and really separating the game from feeling like a Bond-like. But what I liked best was how the team were starting the design; they were thinking about how should the gameplay change as the first conversation, and then how that would that fit into a Bond game, and as it didn't fit well, what world would this "super spy meets aliens and cool weapons" really fit into?

The secondary fire for the weapons. Yes, that was an early discussion. As were the bots. And more depth to the multiplayer. Co-op. All of it. The first several months were filled with crazy aspirations of what they could do with a bigger cart, what if it could be online. A bit overwhelming, and a bit unfocused, but it was also inspiring to see the team drive so hard early on to see how far they could go. They knew beating GoldenEye 007 wasn't going to be easy...

Hollis and co spent much of late 1997 and 1998 working on building Perfect Dark and improving on what had been done with GoldenEye. Rare bosses Tim and Chris Stamper largely left the team to get on with making the game, but already development challenges presented themselves.

HOLLIS: The original plan was that we would be able to knock out a sequel in fairly short order using the GoldenEye engine unchanged. That very well-meaning plan melted away over the course of time, under the influence of the ambition of the team. Every single little nice idea that we wanted to put in, and this could be a bit better, that could be a bit better...

By the end of the project, I understand the software had been about 50 per cent rewritten. So it was a gradual process. It was supposed to be maybe a one year project, maybe 18 months in the beginning, and it just extended and extended as the game went on. The team was allowed by management to put back the delivery.

David Doak at his desk in Rare. Image credit David Doak.

DOAK: The technical limitations on the N64 were a problem because we'd done everything we could think of on GoldenEye to get everything out of the machine that we possibly could. And then it was like, okay, now we have to give yay much more. So it was very much in our minds from the beginning.

For me, a big thing was trying to make the AI better. Certainly story AI, that had been the thing I had done the most work on. I was doing the scripting of it, which was using the scripting language Mark had created. I would want the feature. He would implement the feature. We would riff on it and make it better.

We really liked the idea of dynamic cover, which wasn't in GoldenEye. The GoldenEye guards didn't really know about cover. That's why we ended up with floating crates. Floating crates are there because they're interesting from a gameplay point of view because they make dynamic cover and the guards knew about them.

An early addition to the engine were the light glows. We probably overused them. They're turned up to 11 all throughout the game. But they weren't there before and they added a lot visually. We made it so you could shoot them out. Steve did the stuff for making the levels look darker, and suddenly that became a feature. So you get levels like the DataDyne car park, which is built to allow you to bugger around with the lights. It's on a grid and there are lots of little rooms.

I guess there were a whole bunch of things about the way GoldenEye worked that we had wanted to improve for a long time - probably from the beginning of '97. There were things we couldn't change because they would have been big changes in the game. And we were still trying to bang all the content in. And so there was a shopping list of stuff.

I remember certainly from then there was also this kind of thing: okay, we're going to try and add a whole bunch of new stuff, which is going to mean that the game ends up running outside the memory footprint of the console, but I'm sure we will be able to get it back in there again later. And that never happened. That's why it ended up using the memory pak.

The Perfect Dark developers talk in glowing terms about the freedom they were afforded to make the game. With Martin Hollis as the bridge between Chris and Tim Stamper and the team, the rank and file were shielded from production pressure. In fact, Perfect Dark didn't have much in the way of a schedule at all.

Concept art for the FarSight weapon. Weapon animator Chris Darling says: 'I'd read some of the spacecraft in Babylon 5 were based on sea creatures. I thought that was quite a good idea to carry through for the Maian weapons to make them unique. So they kind of looked like whales, for want of a better phrase.' Image credit Chris Darling.

HOLLIS: I guess my real influence was to try and set a mood and a culture on the team so people knew if they had a cool idea for something they wanted to do, they only really needed to do it. They didn't need to seek anybody's permission.

For example, very early on, in DataDyne, it looks like mahogany, but the polished material is actually two layers of material. Karl Hilton was the artist on the level, and he just wanted to do this material and he kind of developed it on his own. It didn't use any programming, and then we wrapped it into the engine and supported it better so it would work in a more sophisticated way. And he got his desire, which was to have all these shiny surfaces.

It wasn't really from a document, from somebody saying it should work this way. It was from one of the workers at their desk who decided this would be a cool feature and I want to implement it, and they didn't have to go through any management at all.

BOTWOOD: Classical Rare development at its best is Tim and Chris, and Simon and Mark Betteridge [ex-Rare studio head] being this kind of wall between you and Nintendo, and Nintendo's feedback. They protected you from being pulled this way and that, and trusted you to your own devices. And that was Rare at its best.

And they were very experienced, the Stampers. They had made a lot of games. They knew what Nintendo liked. They had good eyes on the market. So they gave us good feedback. And that feedback came to Martin, and from Martin through to us. Because of the team size, it was manageable. But this is an indie team in today's standards. How did we make a game, let alone a 3D shooting game in like, less than five years? It was two years and seven months for Perfect Dark and GoldenEye, roughly. I look back on that and think, how the fuck did we do that?

DOAK: Looking back on it, I think it was an amazing place. The stuff Rare did, particularly the N64, kept Nintendo in business. It was a powerhouse. Without the Rare catalogue, Nintendo might not be in business now. Also, at Rare, we weren't competing with the rest of the world. We were competing with the other teams at Rare. It was a hotbed of creativity. Tim and Chris did a really good job of insulating the creativity and the production and development from the usual bullshit that is out there, but we kind of paid a price for that as well, I suppose.

LOBB: Nintendo and Rare had a mutually respectful relationship, especially in those years. Rare was very much free to decide what they wanted to build, without much direct contact with Japan. I was also going to Japan quite often as well back then, with the focus being to report on how all the games in development were progressing. In the case of Perfect Dark, Japan didn't really have much experience with first-person shooters, so there was even more freedom with GoldenEye 007 and then Perfect Dark than with some of the genres Nintendo holds so much expertise around.

What helped, once GoldenEye 007 was near finished, some of the Mario Club [the internal debugging team] testers in Japan fell in love with the game, and they became proponents for Perfect Dark early on as well. At a high level, Japan's feeling for Perfect Dark was that the game looked like the right game for the GoldenEye team to make, and that it wouldn't likely sell that much in Japan but would be great for GoldenEye fans.

The freedom the GoldenEye team enjoyed was common across the various teams at Rare, but the developer was not without its quirks. It was secretive, even with its own employees, and its secluded nature in the English countryside meant there was little distraction. In the 90s, most of Rare's developers were young men, either fresh out of university or in their mid to late 20s or early 30s, who weren't married or had children. Rare, its developers say, was a place where many put in enormous amounts of hours.

HOLLIS: I regularly did 80 hour weeks. My record was 120 hours on one of my timesheets. And the reason for that was two things. The management of Rare encouraged people to work hard on the projects. And the second thing was the people who were working on the projects were ambitious. And the teams were ambitious. And you knew that you could get more into your game if you were working at that pace. I mean, of course, people got damaged, and I'm sure I damaged myself as well. But then on the other hand, I can't really say that I've got any regrets about it, because the opportunity was manifest into a great success.

The bosses led from in front of the army. They did incredible hours as well. But Rare, in the run up to the sale, was very much a hard work shop. There was no respect given to anyone who did nine to five, I'm afraid to say.

JONES: It was definitely a work work work, work some more, do some work, work a bit more, are you still at work? Yeah. Then do some more work.

Put it this way: your bonus was calculated in some respects by the amount of time you put on the game. I put in 6000 hours overtime. I was really low down on the overtime list.

You've got no idea how the game's going to go. You just don't know. And it was the culture. It just was the culture. It really was. When you think about it, you've got 10 to 12 young people with no outside commitments whatsoever. There's no young families or anything like that. It's just the job. And we're out in Twycross. It's in the middle of nowhere. There really isn't much in the way of anything else to take your attention.

I would get in about a quarter to nine. I was living quite close by at the time. I would leave at nine or 10 o'clock. Sometimes that was seven days a week. We were there at weekends as well. It wasn't all three years, but it was certainly over a years' worth.

BOTWOOD: This is after the transition from bedroom coding into studio coding, where some of the habits of bedroom coding had been carried over. Rare obviously benefited from hiring people fresh out of university with bad or esoteric working habits, but they allowed them to focus on the things they enjoyed doing.

It wasn't explicitly said, you must stay late. It was just, we do it. Crunch was a thing that we didn't even call crunch then. If you wanted to work long hours, you were supported. I was supported. It wasn't unrewarded. You were paid very well. You were promoted if you were dedicated, and by dedicated I mean, do you stay long hours and work yourself hard?

DOAK: The back end of GoldenEye, it was not unusual to do 100 hour weeks. To do a 100 hour week, you need to go in every day, seven days of the week and do 14 hours. That's how you get to 100. When you go home, you go to sleep. And when you get up, you go back to work. If you get there at eight, you stay there until midnight. You have a little bit of downtime, because there was lunchtime, and you would chill a bit more in the evenings. But even during the day, at Rare, the rule was, nobody listens to music during the day. Because Tim and Chris were worried people would be slacking off. Which was a shame, because people weren't slacking off.

It produced amazing things. I think anyone who was at Rare in that period of time who is honest about it will say it did permanent damage to people. We didn't realise it at the time. But you can't do that.

Brett Jones, who created the characters of Perfect Dark, on Maian alien Elvis: 'When he began life he wore a Union Jack waistcoat, like Tim Brooke-Taylor from The Goodies. He was much sillier to begin with. He wasn't so much of a hero he ended up being. This became a United States waistcoat because they wanted him to be not so much an Anglophile, rather an Earthophile. That's why he's got those blue suede shoes on.'

GRANT KIRKHOPE (composer): When you got there, this was the norm. Everybody did it. Those people who started Rare absolutely made it what it was. The work ethic was incredible. The Stamper family sometimes come in for a little bit of knocking from some of the staff. I know what they went through to make that company. They had nothing. And I love that story where people come from absolutely nothing with no money at all, to being completely successful, just because they would not give in. Absolute dedication.

When people joined, they saw what the bosses had been doing and they just did it with them. It wasn't like the bosses were sat in their office, feet up. It wasn't like that. They were in the pits. They were at the coal face with everybody else.

For someone like me, who had been sat on the dole for umpteen years, it was startling. I was like, oh my god, this is how you make something special. The reason Rare was special in that period of time was because that whole ethic just bled to everybody.

At the time, Rare rewarded staff with a sometimes spectacular bonus scheme that would pay royalties based on the success of a game and an individual's contribution to it. The Perfect Dark bonus scheme was even better than GoldenEye's, and many of its developers admit it motivated them throughout the production. In the end, off the back of over two million sales, the Perfect Dark royalty payment to its developers was significant.

JONES: A house and a car! Thank you very much! Oh yeah! I had a mortgage for six months. I got a new pushbike for GoldenEye.

KIRKHOPE: The bonuses were quite spectacular. They really were. They'd buy you a house. The head of the game would allocate your contribution to a percentage, and you'd get that cut from the money. The head of the game would know what you did. It wasn't just based on hours. It would be based on, was your contribution a worthy contribution?

The royalty rate started at 17 cents a cartridge on GoldenEye. That would go to the team. Then it got to 50 cents a cartridge. Then it got to a dollar a cartridge. When games sold a few million copies, like they did back then, it was a few million dollars between the team, and the team was like 15 or 16 people. So people started to get a lot of bread.

HILTON: We were all young guys fresh out of university. Some of us had girlfriends. I had one girlfriend who was in Nottingham. We didn't have wives and families, generally. We didn't have kids. And Rare was a phenomenally entrepreneurial environment. I turned up there from university, they'd finished Donkey Kong Country, and I remember hearing they'd all got paid bonuses of thousands of pounds, which as an impoverished student sounded incredible. They'd all been taken to somewhere nice on holiday as well for a week - taken off to some Caribbean island to say thank you for the work. So you felt like you had landed in this paradise where you were being paid to make video games, and if you made a good video game, you might actually get quite a decent chunk of money out of it. So it was like, right. This is my chance here, so I had better get on with it.

While technical challenges had emerged, Perfect Dark's development was about to suffer its first earthquake. Behind the scenes, Martin Hollis planned to leave Rare. He left in September 1998.

HOLLIS: I didn't choose the timing. I would have preferred to stay to finish Perfect Dark. But I did choose to leave Rare because I wanted to go onto new avenues and try different things. And because I was on a four-year contract and my four-year contract came near its end, they wanted me to sign another four-year contract and I refused. So I was pushed out by that contractual system.

Martin Hollis and David Doak at E3.

Simply put, Perfect Dark wasn't different enough for me, and I could see myself ending up on a rail where I'm doing one first-person shooter after another after another after another, and that's not what I wanted.

I think about it this way: if I cared about money more than other things, then I would have stayed at Rare. What they were offering me was very generous financially.

I don't think there's anything they could have done because it was everything about the company and what it was doing. I wanted to be able to go on holiday for a long time and travel the world. I wanted to be able to set up my own company. There was no way they could have delivered me the things I wanted. That wasn't in their power to give.

I was the most senior on the team. I guess if there had been an org chart made, it would have shown me reporting to both Tim and Chris Stamper. But it didn't feel like I was reporting to them. I was good friends with Chris. I had a very warm relationship with Chris, and a good working relationship with Tim. That ended when they found out I wanted to leave.

I think they had a gradual recognition from little hints here and there. In the end they said, we're gonna put you on gardening leave - I got that, but that was later. The gardening leave was a technicality, really, so they can get rid of you when they want to get rid of you. I was suspended for two weeks, while I thought about my future and whether I was going to say yes or no to this new contract. Maybe it was one week. I had rather too much time to think over what I knew already and what I'd already decided. So I just played some video games.

It was really my heart that was made up. The heart drove the mind for that decision. They [the Stampers] were very angry and hurt. Anybody leaving their company, they would take it personally. The reason they take it personally is that they've invested so much of their own personality into the making of that company. From that position, it's like it's a personal insult.

From what I hear, or from what I don't hear, they haven't really spoken to anyone from Rare since they left the company some time after Microsoft bought it. I'd be happy to speak to them. From my point of view, I think they did a good job and they deserve respect for that.

It was very weird. It was like somebody bringing up your child and you're not allowed to be there. It was very difficult for me. Mainly I coped by not looking at it and by not being in the scene and not knowing what's going on. I kept my head in the sand.

Hollis' exit affected some more than others.

LOBB: I was really happy for Martin, and sad, bummed, a tiny bit mad, a tiny tad of envy (he was going to travel around to parts unknown, Southeast Asia was where he wound up going). But foremost, I was happy for him. The reality was that Martin was my main contact on the team. I spoke with most of the team on every visit, but when I wasn't visiting, my calls into GoldenEye 007 and the beginning of Perfect Dark, were always with Martin. Play the latest build, call Martin. Have a crazy idea, call Martin. He is still a great friend, and I forgave him instantly, but there was still that... "you're leaving?" Yea, that feeling.

Martin Hollis' desk at Rare in 1998. The monitor at the bottom has Perfect Dark on it.

DOAK: The seeds for all of that were in us finishing up GoldenEye. All of us on that team had a love hate relationship with Rare. The team was almost entirely above the average age of people at Rare. We'd all been recruited from university. The wave of people who set up Rare, when Tim and Chris branched out from Ashby Computers and Graphics, and Ultimate, were mostly young guys from that area. In that transition to making 3D games, certainly on the coding side it was necessary to hire people who had higher level qualifications in maths, because in order to write that stuff, you needed to know it. I was doing a PhD in Oxford.

It was a bit of a culture shock going to work at Rare, because it was essentially a family company. And Tim and Chris ran it like a family company. Before I went there I had a college lectureship at Oxford University. I used to teach. I used to go and do all kinds of other things with my life, which weren't just spending 24 hours a day in Twycross. You get pretty pissed off with it after a while. But on the other hand, we were all addicted to making games. So it was this really strange thing where you were doing your dream job and it was a lot of fun, but it was also quite annoying.

Particularly the kind of presenteeism thing. That was infuriating. At Rare, the idea that people would go on holiday or go home was just discouraged. Very much the message from management was, if you keep your head down and work hard and make some nice games then you'll get some money and you can go and buy a sports car. That was the entire incentivisation. Keep your nose clean, and then, if you're good, you'll get some money.

Employee retention incentives was giving people share options. There were three classes of Rare citizenship. Top tier, you were actually in the family, or you were very close to them. Second tier was, you've done a good job, help us make some games, we're giving you some share options, which by the way you will lose if you leave. So, golden handcuffs. And then the third thing was all the people who were hoping to get to tier two. I was the 'hoping to get to tier two' stage, and after GoldenEye they did a - it was a bit like the mafia - It was like, who gets made? Who are the made men, now? They'd do a round of these share options. The list came out and I wasn't on the list. So it was like, well, how come I'm not on the list?

The tipping point for me was they then retroactively made up this rule, which was, yes, it's people who made an outstanding contribution to the company, but you have to have been here for over three years. I was like, well, fuck you.

They'd hired me as a system administrator, so I had much more access to the rest of Rare than anyone else. All through the time we were making GoldenEye, even when I was totally on the team of GoldenEye, I was still fielding technical questions or support questions across the company.

Very much the message we had from Tim and Chris was, you've done really well, you've laid a golden egg, can you guys lay another golden egg? I wanted to do more. And I think Martin wanted to do more.

There was a discussion with Tim and Chris, which Martin facilitated because he was able to talk to them, which was, well this team has also produced things like exporters and technology which are used in other games within Rare, we're more like a mini-company within Rare, can we not be held to the same draconian, everyone has to be here 24 hours a day thing, because we actually have other lives? I think we asked for some reasonable things. Tim and Chris considered it and then said no, and pretty much made Martin's position untenable. So Martin left.

I knew why he was leaving. Tim and Chris, their message to us was, well you see what happens when somebody isn't playing by the rules. We all had these incredibly restrictive covenants in our employment contracts. So Martin was immediately placed on gardening leave. He couldn't go and work for anyone else. The irony was Nintendo then said, why don't you come to Seattle and help us make the GameCube? That was really funny.

I never really had any problem with it, but then they put Mark in charge, probably because I was seen as more of a troublemaker. I just had enough. For me, it was a shit result. We'd lost the person who was the team lead because of them not being accommodating. So I pretty much decided I was going to leave then.

The seeds of Free Radical Design were planted. Halfway through the development of Perfect Dark, four members of the original GoldenEye team left to form their own, rival studio, making a first-person shooter of their own. Each founding member has their own reason for leaving.

DOAK: I didn't really have a plan. My plan was to go and get my life back somehow. And the reason we ended up with Free Radical was because of Steve then saying to me, what the hell are you doing? You're going to leave?! Steve and I decided we were going to do that, and then we persuaded the others to be involved. And that's when we left.

That was a hard decision. I would have ideally had some of the others involved as well. Particularly Duncan and Mark. It was a hard thing to walk away from that game because I put quite a lot into it, but I wasn't going to finish it on the terms that had been offered.

I don't think Tim and Chris are bad guys. They ran that place the way they understood would work. I think they were probably blindsided by people like us. I was at college with David Cameron. Lots of people I knew were off earning fortunes doing weird city jobs. To be in the position where someone's going, keep your head down for a few years and you might be able to buy a Ferrari, it was like, well, yeah, but that's not what I'm interested in. Money's great, but I'd be interested in helping direct this company, look into the future of the company, and we think our working practices are not good.

Karl Hilton working on Perfect Dark in 1998. Image credit David Doak.

HILTON: It was one of the hardest decisions of my life. I agonised over it. I really did. On the one hand I had done very well at Rare. I'd been there for four-and-a-half years and we'd done GoldenEye and Perfect Dark was looking awesome. I liked the team. It was a great place to be. I'd got quite senior. I was the art lead. And Rare had been very good to me in terms of salary and bonuses. I remember talking to my dad, and my dad was like, it's a great job, what are you thinking? You should be staying.

But on the other hand, we'd learnt a lot. But you get to that point where you think, well, they're doing very well. And I could stay here and carry on doing fine and making games for them. But, I could go and do it for myself, too. I wouldn't have ever left Rare to go and join another company. That would have made no sense at all. We left to set up our own company, because at that point we felt we knew enough about video games that we could go and do it for ourselves.

The Stampers had worked incredibly hard, and they had set up Ultimate Play the Game, and then it had become Rare. We had seen how their careers had worked. They were the ones who had Ferraris in the car park. And you're looking at it going, hmm. I want one of those.

So, it was a mixture of ambition and naivety. We thought, actually, there's an opportunity here to go off and do something for ourselves. I had confidence in the other people, so I thought, I'd like to go do that. It was not a straightforward decision. With hindsight it was the right decision, but it could easily have gone wrong.

STEVE ELLIS (programmer): We were young. We were really, I guess, naive. We believed that we had done everything that made GoldenEye the success that it was. We didn't really credit the management or Nintendo with any part of that, which in retrospect seems ridiculous. But we felt we'd done it all ourselves, and we thought that other people had got most of the benefits from it. We thought, if we're going to do that, then why don't we do it for ourselves completely? And then we're in control of our own direction, and we get the rewards from it.

It sounded doable. It sounded easier than it was. We talked about it for really quite a while among ourselves. And really, we talked ourselves into it.

I would say that money was part of it. But I would also, in hindsight, say that I don't think that the management did anything unreasonable. I think they treated us well. And I think that we were wrong at the time to think that we were being exploited. I think that's just the naivety of being young.

It seemed like we were doing all the work, and they were getting all the money, and we didn't recognise the whole other side of the business, and the risk and hassle involved in the non-development part of bringing a game to market. So I think they were good actually. I think they paid us well and rewarded us well with the bonuses. I can't really complain in retrospect, although I did at the time.

You know, it was a difficult decision, because we were walking away from fairly secure jobs, decent pay and good bonuses. It was a big decision to make. And I think if we hadn't been as young as we were, then we probably wouldn't have made the same decision. We were all young, most of us were single and didn't have any family to worry about having to keep the money coming in. So we took a risk. It worked out.

Shigeru Miyamoto became a Perfect Dark multiplayer character head after the Nintendo legend was scanned at an E3 party by Brett Jones. Ken Lobb recalls chatting to Miyamoto about Perfect Dark: 'I remember one fun conversation around Counter-Op, the idea of being the enemy, but really limited by their lifebar/weapon balance was something he thought was interesting, and he had some cool ideas there.'

DOAK: I remember sitting in an office with Tim and Chris, and Tim telling me I would fail. My exit meeting was that, was being given a last chance to recant, and Tim saying to me, 'when Chris and I set something up, it wasn't as hard as it is now. And I think you'll find it's going to be really hard and you won't succeed.'

At the time I probably really didn't like the guys.

ELLIS: They wanted to know why. I was probably one of the first to go. So at that time they didn't realise it was a whole load of us who were going to be leaving. It was the first day back after Christmas in '98. I went in in the morning and told them I was leaving. The way that they handled leavers was, you're out the door immediately. I got in at nine o'clock and by 10 I was on my way home. That was their process. I ended up talking to Mark on the phone that afternoon because obviously, there were a whole load of loose ends with bits of work that I had unfinished that he needed to know about. So we quietly sorted that out among ourselves, just so as to not sabotage the project.

HILTON: No, we didn't have a leaving party. I've spoken to Tim and Chris since. In fact I met Tim not long ago at the Unity conference and we had a great chat and catch-up. At the time we were not popular for the decision. But I think everyone understands why people do it. And certainly no hard feelings at all.

BOTWOOD: It was gut-wrenching. The people I'd spent the last three-and-a-half, four years, five years almost of my working life, who were my first professional friends, half of them were going. And Martin had already gone, and that was a blow. And then the other guys just left. And we had to somehow keep going.

First Hollis, then Doak, Hilton, Ellis and Norgate. Five of the original GoldenEye team had left Perfect Dark in the middle of development, and it sent shockwaves throughout the company. Would Perfect Dark be cancelled? Could it be saved? Rare stuck with it, putting Mark Edmonds, described by colleagues as a genius programmer, in charge. He was faced with handling a recruitment drive and getting Perfect Dark into shape.

JONES: Mark Edmonds became team lead because he was so familiar with the project. And, you know, he's three parts genius. It became his baby. He took on a lot of responsibility for actually getting the thing working and you know, because he's a genius programmer who can just sort of see mathematics.

BOTWOOD: It was a big blow because Martin had enormous respect with the management and on the team as well. I was in awe of his intelligence for a lot of my time there. And I still am. Martin is one of the smartest people I know. The other smartest person I know is Mark Edmonds. I have a lot of respect for both of them.

STEVE MALPASS (designer): Mark Edmonds was the father figure. He had a very calm, reassuring presence. You knew you were in good hands with someone like him.

MARK EDMONDS (programmer): Martin had been in charge of everything. I was just a programmer. After that, because I knew the most about the engine and the code, I guess I got stuck with the job of being in charge.

It definitely had an impact. Four other people left to form Free Radical. That was basically half the team left. The overall effect was a huge shock. Everyone was thinking, where do we go from here? Can we actually keep going with the project? How do we keep going? That caused a massive reset, because we had to get more people on the team to make sure we were still going in the right direction.

If more people had left, then that would have been really serious. At the time we didn't know exactly who was going to leave. It's possible the whole game could have ended at that point.

There were significantly long work hours put in after they left, just to get the whole game completed.

LOBB: I felt that the game finishing was never in doubt, as the foundation was already there when they left; and honestly, they left the game in a good place, and with a remaining core team, specifically so that it could finish. It wasn't like, 'this is impossible, we are out.' It felt to me that they got it mostly out of pre-production, then felt that if they were going to go form a company, that was the right time to go for it.

There was an original plan for late '99, but we didn't really count on that. When Martin, and then the Free Radical members left, then early 2000 became the plan. I know that from the team's perspective, the game slipped. Meanwhile, GoldenEye 007 just kept selling, so from my perspective, the reality of May 2000 was just fine.

With the Perfect Dark team halved, it was necessary to staff up the team. Rare designer Chris Tilston and his team, who had been working on a project that never came to fruition, were brought on board. Staff were hired. Others from elsewhere within Rare, such as designer Steve Malpass and composer Grant Kirkhope, were parachuted in.

MALPASS: Perfect Dark needed some help. They were slipping. So I got put on that team for a number of months during the summer of 1999.

KIRKHOPE: It was nice to get back to a bit darker music. I'm a massive sci-fi fan, so it was right up my street, really. GoldenEye was based on the Bond theme. Perfect Dark was my chance to do my own thing. In my head I was thinking, Blade Runner and the X-Files. It was really big at the time, and I loved the X-Files. I tried to get my own whistling sound instrument into the music.

I wanted to make sure I created the right atmosphere. Perfect Dark has got a particular atmosphere. I would say this, but I think the music does give Perfect Dark a different atmosphere. It feels Blade Runnery. There's that tune called Chicago Stealth. It's my favourite one in the game. I really felt that summed up the rainy, steamy, hot city that you'd see in Blade Runner. So I wanted to make sure it was dark. I feel like I managed it in Perfect Dark. I feel like I did give it an atmosphere maybe Perfect Dark Zero didn't have.

Watch on YouTube

BOTWOOD: It was a weird time. Management were kind of like, are they going to go as well? Do we trust them? But we have to ship the game, so we're going to build a team up around them. So what do we do? Do we put people we're confident in on the team? And then we see what happens?

I had a few tense meetings with Simon and Chris, when I wanted to just talk to them about something, but they assumed it was going to be, ah yes, Duncan's leaving. So they were both sitting there very formal. And I was like, hi, I wanted to talk about... why are you looking at me like that?

The project itself was in turmoil for a while. We had the shape of the project mapped out. It was effectively at alpha. We had a vertical slice, not that we knew what that was. We had the plot mapped out. We had most of the level ideas mapped out. Not necessarily physically, but we knew what they were going to be. So really, shipping it was going to be kind of just building it at that point.

Arguably, if the project had been any earlier, there would have been a stronger argument for canning it and doing something else, because it would not have been laid out in front of you to say, there is another year and a half of development here.

We were lucky in project terms that it was in the state that it was. We were unlucky to lose everybody, but they needed to do that for themselves.

At the time when all these decisions were made, I had glandular fever. I was off work for three months. I just wanted to know what was going on. And when I came back to work, turmoil is the only way I can describe it. There was turmoil everywhere, and it was a while before things settled down.

It was kind of obvious what the Free Radicals were going to work on, and we wanted to get Perfect Dark out in a timely fashion. So, just building up the team to expand the knowledge was sensible, so they did that.

I don't think we ever thought we were in direct competition, but we didn't necessarily want to be launching at the same time. We thought it would take them a while to set up a company and get a game out. Typically that would be like, well, we've got maybe a year on them. That gives everybody plenty of room.

We didn't wish them ill. Far from it. It helps everybody if we have a healthy video game industry, with lots of different companies to work at. If there's rivalry there, it's friendly rivalry. We stayed in touch. We gently ribbed each other if we bumped into them at E3. It wasn't like, I'm never talking to you again! All the stuff you do is shit! It's never that, because the industry is too small to suffer that kind of nonsense.

We had a plan. We had our production laid out for us. We had all of our information there. It was a case of adding people onto that, to do that work.

In spring 1999, not long after the Free Radicals had left, Rare moved down the road into bigger offices. Image credit Steve Malpass.

MALPASS: It wasn't in a great state. I remember the framerate being really poor, which for a first-person shooter, that's not ideal. They had memory issues as well.

Back then, Rare didn't even share engines. The team wrote their own engine and didn't share it with other teams. Perfect Dark was an evolution of the GoldenEye engine. They had super talented guys working on it, but still, they were hamstrung with the tech they had.

Grant Kirkhope, he was obviously always quite outspoken on pretty much everything, he was like, their engine is shit! Banjo's engine is much better!

There were the beginnings of a few levels. The first level, that was in and working, but the framerate was terrible. That was the first time I'd seen it. You never really saw anything of the games early, you know. The Stampers obviously had access to all the teams. It didn't look the best.

Edmonds and this second life Perfect Dark team realised the game had become too ambitious to run on a retail Nintendo 64. They had no choice but to require the console's expansion PAK for most of the game to work. This was never the original intention, and it was a significant blow to Perfect Dark's potential success.

HOLLIS: We didn't plan to use the ram pak from the beginning. It was simply the accretion of all the features that were added to the engine, to the levels of the game, meant that it didn't really work on the conventional size of N64. So the decision to actually use that was made fairly late in development, to have that as a required thing.

HILTON: It was discussed, and we all agreed that absolutely no way should we do that. That was one of the things we were working on - was to make sure it didn't happen. So when it came out and it needed it, that was a surprise to us. The dev kits had the extra memory. But it was always planned we didn't want to go down that route. Whether we'd have achieved it or not, who knows? Maybe we wouldn't have managed it either. But it wasn't something we wanted to happen, and we were going to do everything we could to stop it.

We thought it would limit the sales. It's a cartridge where you have to buy another cartridge to get the thing to run. Ultimately, it didn't sell anywhere near as well as GoldenEye. There are a whole host of reasons for that, but certainly one of them is, if you have to buy a game and you have to buy an expansion to even get the game to run, it's clearly going to hurt you.

Perfect Dark's infamous Anal Land sign confused fans for years. 'We were like, oh fuck,' says Chris Darling. 'The general viewing public was never meant to see it. It was always a bit of an in-joke with us. It was one of the first things people spotted about the game.'

BOTWOOD: It was brutal in engineering terms. We ran away with ourselves. We didn't keep well enough of an eye on the technical requirements, and we didn't shrink the game down enough to get it to stay on cart.

We let the textures get out of control a bit, and the geometry, and when it came to try and ship on the actual console, it wouldn't run. And at that point, it was quite late. We took a decision to ship with the RAM pack. I was not privy to those discussions, and I can only imagine how painful they were. But that undoubtedly impacted our sales because it was like, you can play this game if you buy this other thing.

But we did at least manage to ship the multiplayer mode outside of the RAM pack. So the RAM pack kind of unlocked the single-player mode. There's definitely something there to play, and arguably it was going to be the more popular thing because at the time GoldenEye, it was rented for a weekend game between mates. And we were selling that sequel in a way.

EDMONDS: You can probably lay the blame for that on me. Just adding more stuff in. Just trying to keep everything going. I wasn't worrying about how much memory was being used. It was just a case of keeping the team moving forwards. The development consoles we had, they had more memory in than the actual consoles everyone had in their homes. We were using up this extra memory. Then, when it came to trying to get the game to run on a normal home console, we realised, okay, this might be a problem. It was just impossible. It was just lucky Nintendo had been working on this expansion pak. Definitely it decreased the total number of sales, because not everyone would have had one, which was a shame.

The N64 RAM expansion pak was pretty much required for Perfect Dark.

LOBB: I wanted them to use the 64DD and have online; and although that ended up not happening, I still wanted the Expansion Pak to ship. In fact, I needed it to ship for a few of the other games we were building at the time, including Donkey Kong 64, StarCraft, Star Wars Episode 1: Racer, and Star Wars: Rogue Squadron. Meanwhile on the Perfect Dark team, after they decided that the DD wasn't for them, they just kept adding and adding to their memory allocation, believing they could later compress/rewrite to fit into 4M [meaning no Expansion Pak]. Meanwhile, I was on the "don't worry about that yet" side... I had seen the 640x480 version, the widescreen, so many things which simply wouldn't work without 8M... and didn't want that to go. I was always hoping that they would just get to the point where there was no way to pull back memory.

Meanwhile, at NOA, the whole discussion around how to describe on the box that most of Perfect Dark wouldn't work without the Pak was the other side of this story. I really wanted to pack it in, but it had already shipped with Donkey Kong 64, and NOA felt that core users would have it when PD shipped in May 2000... but no-one wanted to ship this game as "Only works with the Pak". In the end, we did what was best for the game, and hopefully most players didn't have too much problem finding a Pak in order to play the whole game. If I could go back, the Pak would be free, but we can't go back. Of course, the 4K/60 version also plays really well on Xbox today within Rare Replay.

Even with the help of the N64's expansion pak, Perfect Dark's framerate is pretty awful. This outcome was also something the original developers actively tried to avoid.

Watch on YouTube

HOLLIS: The framerate, you know, it kind of bulges at the seams. The framerate doesn't perform as well as I would like. It's partly my responsibility and partly not my responsibility. I don't really want to blame anybody because it was the trajectory the team were on, to add more and more detail and more and more interest visually to the scene and more and more features to the AI, and so on and so forth. So what happens when you've already done all the optimisations that you can realistically do on the project before? Well, everything gets a little bit slower.

HILTON: When it came out, one of the disappointments was that the framerate, they hadn't managed to maintain it. I spent a lot of time planning out the levels to make sure we weren't drawing too much. We'd learnt an awful lot about the engine and the N64. I'd created a lot of environments and done a scripting language for them so they ran really well. The problem was, how you then set those things up with a number of characters and the amount of AI - that's when it really starts to stress everything.

In that final six, nine months of development, when most of us who had planned all those levels weren't there anymore, the new people came in and designed levels and put all the stuff in to make it play well, but they didn't understand some of the technology behind how we were trying to keep the framerate up. So I think that hurt it a lot.

I'm not trying to blame those people. It's just the way it is. You do a lot of work to make sure it runs better and looks better and is more high-res, and when the game finally comes out and it doesn't quite do what you want it to do, you're like, oh, that's a shame. I'm sure we would have had similar issues with the setup, too, but I think we would have been more conservative on the setup, just to try and keep that framerate up. It was one of our primary things - to get a smoother-running game than we had with GoldenEye, and I don't think that was the priority for Perfect Dark in the end.

BOTWOOD: We were hoping engine optimisations would get some frames back. And they did, to a degree, but then, it was never great. Another way to look at this is, GoldenEye had been successful in spite of the framerate, so it didn't seem to be that much of a problem. If it was serviceable and it worked, then okay. And you could play the game. It's just the framerate now would not survive any kind of review process. It would be unacceptable. We would not ship with that.

While the Perfect Dark team was largely shielded from the influence of Nintendo, there came a point when the Japanese company stepped in to pull a promising feature from the game. The developers had created a system that would allow players to scan a face into the game via the Game Boy camera, and then convert that into the face of a multiplayer character. Fearful of a nightmare PR situation hot on the heels of the Columbine High School shooting, Nintendo put its foot down.

EDMONDS: That was actually a really cool feature. It was using the Game Boy camera. Putting in all the menus, basically working out how to map a photo onto this polygonal square head. The feature was finished. Nintendo decided it was too risky to put this feature in, so from a publicity point of view, it was better not to include it, unfortunately. So we then had to cut it out. It was because of school shootings going on in the States. It would have been a bad situation if someone had taken someone else's photo and stuck it in the game, and that had somehow connected with real life. It would have been just something again that no-one else had done.

CHRIS DARLING (animator): There was a bit of a disagreement between Nintendo and us over that, in that Nintendo's press release said we couldn't make the feature work. The Stampers were a bit upset at that, because they liked to think there wasn't much we couldn't make work. And it was a functional part of the game at that point. We'd mucked about with it a bit. There had been a rash of school shootings. I think the concern was people would go into school, take photos of their teacher or something and then fantasise about shooting them, and then try to do it. It was a fun thing to have, to shoot your mates and stuff. As long as you're not a psychopath. But I understand completely why it was taken out. I don't think it really detracted from the game to remove it. So why not?

Perfect Dark launched in the US on 22nd May 2000 to spectacular reviews. It remains the highest-rated Rare game of all time. For those who stayed to finish the job, millions of sales meant life-changing bonuses. For others, mixed feelings.

HOLLIS: It's a strange situation when something you've worked on so hard to form is wrenched away from you before it's complete.

Mainly I've got very positive emotions towards Perfect Dark. I feel very warm towards the project. I think it's a very good game. I think we all had an opportunity at that company to produce games of that calibre, which was unusual, if not exceptional, and I know that every member of that team was an absolutely top notch talent. The quality of people we had on the project was extraordinary.

Watch on YouTube

DOAK: I was really busy at the time. I played through it on the Xbox version. That was the first time I played through the whole thing.

Mark did such great stuff with the multiplayer, with the bots and stuff, all the deployable weapons. There's a whole bunch of stuff in Perfect Dark which is really important and innovative in that space. By the time we were doing TimeSplitters 2 and Future Perfect, we were ripping it all off.

A lot of that design innovation will have been Duncan and Mark. It wasn't me. A couple of times in recent years I've gone to a Norwegian retro games event. You're signing people's boxes and carts. A lot of people bring me Perfect Dark to sign. Yeah, I'm proud and I feel completely justified in signing it. But I always say to people, you know I only did the easy part of this game? I did the part where you make up a whole bunch of ideas and say, wouldn't it be good? And then you piss off.

HILTON: I think it was good in terms of storytelling. We told an interesting story at a time when games weren't telling stories. For us it was a big thing that when we left and set up Free Radical, we wanted to do another, better storytelling game. And that ended up being Second Sight, which is personally of all the games I've made the one I'm most proud of. So in that sense it was a progenitor of a lot of good stuff. The artwork stands out quite well, in terms of what we managed to achieve on that console.

The multiplayer, it never really gets talked about, but it had everything GoldenEye had plus a lot more that we'd learnt. I remember playing some of the multiplayer levels not long before I left and just thinking, this is so much better than GoldenEye. For an N64 game, I think we achieved a lot. I think it was a good second game.

It was slightly bittersweet as well. It was like this child we'd abandoned on someone else's doorstep and walked away from. So we felt slightly guilty.

ELLIS: I played the game. I enjoyed it. I never really felt much ownership over that game, I guess because we didn't see it through to the end. When I look back, it's not a game that feels like my game, in the same way GoldenEye and the TimeSplitters games do.

HILTON: We were at the 2000 BAFTAS as Free Radical with TimeSplitters. Perfect Dark was up for some awards, and it won. It was quite odd, because it won an animation award, and James Cunliffe [ex-Perfect Dark animator] was with us at Free Radical at that point. And so someone from Rare went up to pick up the award, and yet all the people who had actually done the stuff were sitting at our table, but we weren't employed by Rare anymore. We were all cheering and clapping for James. There was obviously more work done on it by other people. So it's not that they weren't entitled to the award. But we were clearly entitled to it, too. But obviously that doesn't happen if you've left the company.

ELLIS: The 2000 BAFTAS was actually the next time that I saw the Stampers. It was fine, actually. Tim didn't really want to talk to us. He took it all more personally than Chris did. But Chris was quite nice. He said, are you proud of the game that you've made? Yeah, we were. I think he didn't really think that we'd done the wrong thing.

While Perfect Dark is considered by many to be a better game than GoldenEye, it has always lived in the shadow of its predecessor. It sold far fewer copies and it failed to break into the popular culture. But, 20 years later, its developers are intensely proud of it.

BOTWOOD: GoldenEye is more closely linked to popular culture outside of video games. If you didn't know video games very well at that time and you were going into a shop, you did know Bond. But the draw of the licence is important when you're thinking about GoldenEye and Perfect Dark. And Perfect Dark didn't have a licence. It was the sequel, sort of, by the same team, sort of. It's a gamer's game. It's not got that same cultural weight as something like GoldenEye. I think that contributed to it.

The RAM pack contributed to it being below the radar. Perfect Dark was about as successful as I thought it could be, without being blindly naive. I had my moments of being blindly naive, thinking this is going to be great! It's going to be amazing! It's going to be huge! It's like, well, maybe not. Perfect Dark as a name, as a brand, as a logo is very weird in comparison. It was not ever going to have that same kind of weight. Although we might have wished it.

EDMONDS: I was happy with it. I was very happy to have worked on it. Satisfied with how it turned out. I don't think at that time we could have done much more to make it any better. So I'm definitely satisfied with the work we put into it.

The whole James Bond licence, you can't really beat it, can you? So that adds to GoldenEye. But I do think Perfect Dark was a better game. I agree with the Metacritic rating, which is just slightly higher. Even if the framerate is slightly lower, I think it was a better game.

Concept art of a Maian soldier holding the Callisto NTG sub-machine gun. 'It was like a whale,' the gun's designer Chris Darling says. Image credit Chris Darling.

DARLING: It was great. The first time I saw the game in its final form, I was in Las Vegas on holiday with my family and girlfriend. We saw the demo running in Electronics Boutique in the window in the front. I was like, wow! That's the game I worked on! That was really something, to see it actually out there and in a shop window. It was the first time I'd actually seen something I'd worked on in a shop window in my career. That really blew me away. I was like, wow, look, there's a machine gun I animated! You sit there for yonks in Maya thinking, how is this bloody thing going to work? I tried to make all the different animations unique, and that's really not easy when you've got a bunch of firearms. I was just like, oh, do I cock it that way and just slam that in and do that? You try to make each one different. You don't want to bore the person playing. So I was very single-mindedly trying to get this thing to work. It was a completely different experience to see it in a shop window as a product. It was surreal. You don't ever think of it as a finished thing that's going into a shop when you're making it. It's quite an odd feeling to see something as a final product just on a shelf. It's a really surreal experience. It's just kind of mad, really.

LOBB: It's always different from the inside. If I think about the game, I think of all the multiplayer modes, the Laptop gun, the FarSight! The challenges! Counter-Ops. Co-op split screen in a story-based FPS! As with GoldenEye 007, I was proud of the incredible replay ability of the campaign. Also, of Joanna as a cool heroine at a time when there really weren't enough female protagonists.

But what I am most proud of is that great work this agile, churning team did to ship the impossible. As Martin has said in the past; early on, it was "we can do everything". It's not a bad way to start a sequel to a huge and growing hit like GoldenEye 007. They were constantly trying to live up to what was happening all around the world with GoldenEye 007, which was growing in popularity literally every month for most of the Perfect Dark dev cycle. Even with the team members leaving, that internal pressure to live up to GoldenEye 007 was constant. As a fan, a gamer, and a friend, they did it. That's what I am most proud of to be honest.

This article is dedicated to Jon Mummery, a member of the Perfect Dark team who sadly passed away before its 20 year anniversary.

Read this next