Interview With XCOM Creator Julian Gollop
We speak to the mind behind so many classic strategy games, including XCOM.
Published on Aug 30, 2011
Carving a niche for yourself for any considerable length of time in the videogames industry isn’t easy, but Julian Gollop is one notable exception.
From humble origins on the BBC Micro, inspired by a love of traditional board and strategy war games, Julian made turn-based strategy accessible and, most importantly, fun, with the likes of Rebelstar, Laser Squad and the massively successful X-COM series. Given the extent of his career, we thought we should start at the beginning.
What would you say were the roots behind your twin passions for board games and computer strategy games?
Largely thanks to my father, I think. Ever since I was a young child we used to play all kinds of games – board games, card games, chess. Actually, we as a whole family played games a lot, especially at Christmas time – we didn’t watch many films.
Because I liked relatively complex strategy games, when home computers came along I immediately saw them as a very useful medium for playing these kind of games – the kind of games we went on to make.
What prompted you to start designing your own games?
As soon as I started playing games I was making them. Around the age of 14 I started getting into more complex games – Dungeons & Dragons, SPI board games, Avalon Hill… a lot of stuff. From then on I developed a big interest in strategy games.
Islandia was one Gollop's first games.
You devised Time Lords and Islandia while you were still at school. How did the publishing deal with Red Shift come about, and what do you remember about creating those two titles?
I got involved through a friend of mine, who was involved in a group of war-gamers in Harlow. Red Shift was set up by a guy who was a miniatures war-gamer to create computer games. Time Lords and Islandia were programmed by a school friend of mine, Andy Greene, who later on worked with us at Mythos Games.
He had a BBC Micro, I had nothing apart from some game designs, so we combined the two Time Lords started out as a pen-and-paper game. You had to generate the universe using dice, and it required a game master to run the game.
It was clearly really designed for a computer game naturally, so that was my first computer game design, I guess. For Islandia I wanted to do a game with some very basic resource management economics and, for some reason, naval combat. I had the idea of randomly generating a map with islands, apart from the central island where the four players start.
It must have been pretty exciting for you. How did they
do sales-wise? Though we’re guessing you probably didn’t see huge piles of royalties come flooding your way…
It was a tiny amount, actually. Not being too business-savvy at that stage, the owner of the company took most of the money, of course. I didn’t really know how well they were selling relative to other games; there weren’t any sales charts that I knew about at the time. I don’t think I ever saw them on sale in any shops.
We sold them at various computer shows and mail order, and I guess some were sold through distribution channels to various independent shops. I was pretty excited when they were published – when we actually had a physical product printed with the instruction manual, and the cassette tapes were manufactured.
So where did the idea of action points, which we first see in Islandia, and an important feature of most of your subsequent games, come from? Does the concept derive directly from strategy board games?
Yes, from board games. Many of the SPI games used concepts like this. They had fairly sophisticated things like simultaneous movement and trying to simulate the cost of different actions. They didn’t really work well as board games. In fact, they would have worked better as computer games.
How did you get into programming games yourself?
I bought a ZX81 from a friend at school for £25 and started to learn programming. I was quite amazed by it, in fact. I could really appreciate the power of these machines, even though it was a ZX81 with 1K of memory, chunky characters and no graphics processing to speak of. I
then bought a ZX Spectrum and started programming Nebula. It
wasn’t a bad little game, I programmed it relatively quickly in BASIC, and it did pretty well.
Rebelstar Raiders is, except for X-COM, one of Julian's most adored titles.
You were obviously quite a sci-fi fan from an early age. Time Lords is clearly influenced by Doctor Who, and do we detect a bit of a Star Wars vibe with Rebelstar Raiders?
I can’t deny a certain influence, it’s true. But still, I would probably say that the main influence was some of the science-fiction board games I’d played. Game Designers Workshop had a game called Snapshot. Even though I’d never actually played the game, I did read the rules.
I think that game had the concept of ‘snap-shots’ and ‘aimed shots‘, which is a concept I used in subsequent games, of course. Rebelstar Raiders turned out to be very popular with friends and people who had bought it, despite being just a two-player game.
At what point did you decide that you wanted to pursue game design as a full-time profession?
Immediately after I left school, I think, or at least in my final year. Once I’d got my hands on a ZX81 I realised this was the future and never looked back, apart from a minor diversion at college. Although I didn’t do a lot of studying, I managed to complete two computer games while I was at college: Chaos and Rebelstar.
What were the inspirations behind Chaos?
Chaos was actually based on a board game I made in 1982, inspired by a game by Games Workshop called Warlock, which I remember some kids at school playing, although they wouldn’t ever let me play it. So I thought, ‘Screw them, I’ll make my own magic game and it’ll be better than their game anyway!’ So I made this board game in 1982, and a preliminary version was programmed by Andy Greene on the BBC B.
Then I decided to do an adaptation for the ZX Spectrum. It still had a lot in common with the board game, although it had some new elements – the idea of casting creatures as illusions, for example. In a way, Chaos is one of my favourites of all the games I’ve ever made.
I’m not quite sure why, but it was a good, fast-playing, fun game; you could play with up to eight human players or a mixture of human and computer opponents; and it was certainly chaotic with that many people! But yes, I liked the game a lot when I was making it and playing it.
It may not look like much, but Chaos's multiplayer mode was compulsive.
What do you remember about working with Games Workshop?
I wasn’t the one involved in directly negotiating with them; it was mostly former Red Shift guys who had decided to do some games for Games Workshop. We actually did some adaptations of some of their board games including Battlecars, which I programmed the car designer for, and Talisman, which was done by another colleague of mine.
Chaos was an original game, of course, although I don’t think I ever told them it was inspired by Warlock. I think they pulled out of publishing computer games after a short while, although at one point they did want me to do a game based on their Judge Dredd board game.
I wanted to do a strategy game where you controlled a squad of judges and sent them to crimes, with a tactical combat sequence where you dealt with the perps, but they didn’t like it, and opted for some sideways-scrolling platform game with Judge Dredd on his bike, which was awful.
Why do you think Chaos was so appealing to players?
It has an interesting balance of randomness and strategy. You don’t know what spells your opponents have, there’s a certain amount of randomness in whether you can cast a certain spell or not, and at the same time you’ve got to think tactically depending on what spells you’ve got and what you’ve managed to cast.
You could say it’s a nice blend of tactics and chaos… I get a number of requests each year from people wanting to do a remake of Chaos, and I say, ‘Yeah, go ahead. Not a problem.’
After Chaos you created Rebelstar and Rebelstar 2 for Firebird, which updated the turn-based tactical scenarios seen in Rebelstar Raiders. What do you remember about creating it?
I did Rebelstar at college, entirely on my own. Again, it was originally just a two-player game. I took it to Telecomsoft, because they had an office in New Oxford Street, very close to where I was living in Islington. They liked it, but they said they wanted single-player, so I went back and spent a few weeks working on the single-player version.
I had no idea how to do it: I had to invent a path-finding algorithm and I knew nothing about such things, so I had to come up with something from scratch. But it works. You had the single-player version on one side of the tape and the two-player version on the other side.
They published it, and it sold pretty well, even though they decided to put it on their Firebird label – I was hoping they were going to put it on their more expensive label. The royalty was a pittance – I think I got ten pence a copy – but it sold tens of thousands. I can’t remember the exact figures, but it did sell a lot. I bought a nice shiny red guitar and dotted around for a bit spending some money, so yeah, it was cool.
Rebelstar 2's was the first of the Gollop brothers games to be built with a new code.
How did you find the jump between programming relatively simple single-screen titles in BASIC (Nebula and Rebelstar Raiders) and the more complex, multi-scrolling Rebelstar games in assembly language?
Pretty natural really. I didn’t have too much of a problem with it. Chaos was the first assembly language game I did and Rebelstar was the second. Although I do remember debugging was something of an involved process: I often had to print out the Spectrum source code on long rolls of printer paper and go through the code line by line, which was a very reliable method, by the way.
I was relying on dual microdrives with a small set of microdrive cartridges, which I have to say proved 100 per cent reliable through the whole development. Amazing, really.
The Rebelstar games were your first foray into the ‘squad-based tactics’ genre in which you later made a big name for yourself. What attracted you to making them, and why do you think you continued to be interested in producing this style of game through to X-COM and Laser Squad Nemesis?
Again, it goes back to board game roots. One I was playing called Sniper stands out as being quite influential. Also some miniatures games we were playing in our games group in Harlow were actually closer to Rebelstar Raiders and Rebelstar.
The tactical decisions in the game, like whether to use a quick snap shot or higher AP aimed shot are interesting, as they require the player to make decisions which sometimes might be obvious, or sometimes might be a fine balance between risk and reward.
Laser Squad is generally seen as your ‘classic’ 8-bit era squad tactics game. What improvements or changes did you try to make to the already successful Rebelstar formula for that game?
One of the main things was to create a scenario-based system where we had multiple scenarios because we wanted to release expansion kits to add some configurations to your squad, so you got to choose weapons and arms before going into battle, improve the AI, and have something of a continuous story.
We introduced a line-of-sight system in Laser Squad, so you had hidden movement of the enemies, according to line-of-sight rules, which Rebelstar didn’t have. It was a number of innovations, really. I was keen on innovating and exploring the basic turn-based tactical combat system.
My brother Nick joined me at this point, and we set up Target Games. We did so many versions of this game – Spectrum, Commodore 64, Amstrad CPC. We did separate disk-based versions of all these, and there was a PC version by Krisalis Software, who also did the Amiga version.
So it was a major step up, I guess. I got most of the Spectrum version done and Nick was working on the Commodore 64 version – the Amstrad version used the Spectrum code because it was the same processor, of course. The Commodore 64 version was a bit trickier because the hardware was a lot more complicated.
Your games always seemed to achieve a level of critical acclaim in publications like Crash and Sinclair User that must have pleased you at the time. Was the critical response satisfying or a good source of motivation to create better and more complex games?
It was very inspiring, I guess because I was doing stuff nobody else was really doing, which helped. I was making games I wanted to play. It was important to me that the games would be something I would be actually interested in playing at the end of making them, so I was pleased that other people liked the games as well. You could say I was pleasantly surprised.
Lords Of Chaos was more RPG led than any of Julian's games before.
Lords Of Chaos was possibly your most complex game up until that point, from a design point of view. How did you set about updating the core ideas behind Chaos for this game?
I’d set up Target Games with a friend of mine. He left, so me and Nick decided we would create a new name for the company – Mythos Games. It was just the two of us, so I was still programming and designing, of course. Lords Of Chaos was a bit more role-playing oriented – you had a sort of wizard creator and you chose spell levels and basic characteristics.
The idea was as you played through the mission scenarios you gained experience points and improved your character. The problem with Lords
Of Chaos was that it was a much slower, more time-consuming
game compared to Chaos, so the immediate fun factor was a little bit lost. It was better as a single-player game.
We actually created maps that could be multiplayer and specific missions that were single-player only, which had much more puzzle-like elements to them.
It had some neat ideas: you could buff up your creatures with potions to increase their attack, defence and speed, and there were other useful ones like invisibility potions, for example. You could have creatures riding other creatures, such as horses and gryphons, and had flying creatures and ground-based ones.
Each spell had eight levels so a level one spell would summon one creature, but if you had a level eight gold dragon spell you could summon up eight gold dragons and your opponent would probably be quaking in their boots! Although having said that, the mana cost would probably have been prohibitively expensive…
As a multiplayer game it probably didn’t work so well. Because of the hidden movement system, you weren’t supposed to see what the other players were doing. I remember playing four-player games of Lords Of Chaos… it took hours. You had to be very patient.
Laser Squad had a lot of innovative features not seen in strategy games before.
UFO: Enemy Unknown began life as Laser Squad 2, so what prompted the change in setting and what were the influences behind it?
Well, at the beginning it was still just Nick and myself. We were working on a very basic demo of Laser Squad 2 – it had isometric graphics and the environments were correctly 3D modelled so your shots could go up and down, or left and right. But it was still just a two-player tactical combat system. We decided once we had this demo that we needed to find a better publisher.
We had a few candidates: one was Domark, another was MicroProse. MicroProse was the company we really wanted to publish the game, because of Civilization and Railroad Tycoon – to us they were really the premier publisher of strategy games in the world at the time.
We took the demo to MicroProse in Chipping Sodbury, and showed it to a guy called Pete Moreland [MicroProse‘s head of development]. He showed it to a bunch of other people there – fortunately enough people at MicroProse were familiar with Laser Squad – and Pete came back and said to me and Nick that he liked it, but he wanted a ‘bigger’ game.
I had to ask him exactly what he meant, and it turned out he wanted something a bit more epic in terms of scale and scope, like Civilization, rather than just short tactical skirmishes.
I thought, well, yes we could do that. In fact it was a very good idea. He even suggested the theme of UFOs. I went away and thought about it, did some research on UFOs, and the more I looked at it the more I thought this was cool.
I remember going out and buying a video tape of the old Gerry Anderson UFO TV series, and the thing that inspired me from that was the idea of having a worldwide organisation that was set up to stop this alien menace.
But I wanted it to be a bit more firmly rooted in UFO mythology. The other thing that was influential was a book by Bob Lazar, where he describes his experience of working on recovered UFOs in Area 51, and how the US government were allegedly trying to reverse-engineer captured flying saucer technology.
That, of course, was a major influence on the whole cycle of shooting down UFOs, investigating crash sites and capturing alien technology in X-COM.
Pete liked our ideas, so we did a very rough design document – about 12 pages long, which was the longest I’d ever done at the time – and started work on the project. Nick and I were doing the design and programming, and we had two artists assigned from MicroProse working on the game, so basically it was just four of us. John Broomhall did the music for us, in the last two months.
And, of course, the game was your big break into the profitable PC market…
It was a much bigger project than we had done before in terms of scope and ambition. It was seriously ambitious. We knew we had to move onto the PC, particularly for the American market. I remember MicroProse asked us if we could program PC games, and we replied, ‘Yes, of course,’ even though we’d never actually programmed anything on the PC at that point!
The actual demo of Laser Squad 2 was done on the Atari ST; believe it or not, we had originally programmed Lords Of Chaos on the ST as well – we’d given up, of course, on the 8-bit machines at this stage. But the PC was the main games platform in the US, and it was becoming that way in most of Europe, so we clearly wanted to develop for it.
We also programmed an Amiga version; Nick did the conversion, but it was quite tough because the Amiga wasn’t quite as fast as PCs were becoming at that time.
UFO: Enemy Unknown was a precursor to the XCOM series.
UFO: Enemy Unknown was far and away your biggest selling title. Did its popularity surprise you?
Well yes, it was phenomenally successful, and my career has gone downhill ever since! [laughs] I’m still staggered at the number of people who have played this game. When I moved to Bulgaria a few years ago I was astounded – most colleagues have played it, and I get endless requests from Russian magazines for interviews, so it was hugely popular in Russia, even though I probably never saw any royalties from these countries! And, of course, it was very popular in the US.
Why do you think it proved to be so popular with the games-playing public?
There were some fortunate coincidences. The first series of The X-Files had aired in America in ’94, just before the game had released. PC strategy games were quite a significant genre in those days, and I suppose the game as a whole had this expansive scope to it – we had this interplay between strategic and tactical missions.
Although the tactical missions are pseudo-randomly generated, it did fulfil the objective we’d been given, which was to make a ‘big’ game. When you look at some of its parts they don’t seem to add up to much, but when you put it all together it creates a very interesting multi-level game where what you do at the micro level has an impact at the strategic level, and in turn at the tactical level.
So when you’re on tactical missions you are thinking about which aliens or weapons you can get for research, and on a strategic level you are thinking about fighting terror sites in one part of the world and maintaining relationships with governments in others. For me it’s almost like an ideal game design in that sense.
In some respects we were forced down that route because we simply couldn’t create a vast amount of content with just a handful of people, so we had to leverage what we could out of the game system. As a MicroProse game I think it fit quite well into their catalogue at the time.
Were you satisfied with X-COM: Terror From The Deep given that MicroProse gave you such a short development period to create a sequel?
Well, they dragged their heels about it for some bizarre reason, I don’t know why. Eventually they agreed to do a sequel, and they wanted us to do it in six months. We said, ‘Well, we can’t really do anything meaningful in six months except just do new graphics and locations for the existing game.’
They were quite insistent about it so we said, ‘Okay, why don’t you take our code and do the sequel in six months, and we’ll work on the third game and spend longer about it.’ So we actually didn’t have much involvement at all with the first sequel.
Needless to say they took 12 months instead of six, even though their team size was massive compared to ours. I didn’t really play it that much, to be honest. The graphics were quite impressive, but I think they made a mistake trying to expand the scope of the game by making the missions bigger and longer.
XCOM: Apocalypse was built around an idea for a Judge Dredd game Gollop had years earlier.
X-COM: Apocalypse marked the last time you had personal involvement with the series. Why did you decide to go for a SimCity-style environment over the world map of the previous games?
Basically we didn’t really want to do another X-COM game at the time, because after three years working on the game we wanted to do something different, which was, of course, very silly of us because we should have really capitalised on what we had.
The original idea for Apocalypse was somewhat going back to the Judge Dredd concept I had many years before, which was having a city that was a living, breathing entity with different factions and corporations with economic relationships to each other, and populated by traffic and people moving around.
So we took that idea and put it in the X-COM universe, but this time the story was about multi-dimensional beings attacking the city. In retrospect it was entirely the wrong way to go, but nonetheless it was an interesting game with a number of innovations.
What do you remember of the development process for Apocalypse, and do you think the game lived up to the high standards and expectations of fans of UFO: Enemy Unknown?
We tried to expand the team, so we had level designers contracted in, mostly friends or friends of friends, so it didn’t work quite so well as a development process. The artwork was done by MicroProse, but we had many frictions with them because we didn’t like the art they were doing, and in fact it still didn’t end up looking very nice in my opinion. Apart from the guy who designed the vehicles, who did a very good job – the aliens, city and buildings didn’t look that good.
It was our first attempt at doing a real-time combat system. In fact, we were extravagantly ambitious in providing both a real-time and turn-based version – something that didn’t quite work – and the game was probably over-complicated in many respects, and was a real pain to play.
The thing that probably caused the most headaches was that the maps had multiple levels, which were somewhat confusing. So we kind of made some mistakes with that game, although it still sold well and was profitable.
Do you have any regrets about your subsequent sale of the X-COM property to MicroProse?
Well we didn’t really have much of an option because the actual intellectual property rights were somewhat ambiguous. Our lawyers told us that if it came to a court battle MicroProse would probably win; their lawyers were clearly telling them that if it came to a court battle, we would win.
They wanted us to do a deal where we would sign over any rights that we might have in return for some cash plus a high royalty on X-COM: Apocalypse. They more or less insisted on it, otherwise they were threatening to cancel the Apocalypse project, so there was a lot of bluff involved. We thought we may as well do it and afterwards, go and find some other publisher.
I do remember going to E3 in 1999 and MicroProse had a huge display for X-COM: Alliance, with giant tubes with alien foetuses and guys dressed up as aliens walking around, but when I went up to try and play the game they didn’t really have anything playable. They were clearly having problems getting the engine to work properly.
It was a squad-based game, four people in each squad, and it looked good, but it was kind of a tragic demo in a way – the playability wasn’t there. It was later cancelled, of course.
Magic & Mayhem introduced an RPG mechanic long before Baldur's Gate popularised it.
With Magic & Mayhem you updated some of the ideas and themes from Chaos and Lords Of Chaos for a contemporary audience. How successful do you think you were with bringing those ideas to PC gamers?
The idea was to take some concepts from the original Chaos, with a wizard casting spells and summoning lots of creatures, and make a real-time strategy role-playing game. It was partially successful, and we had some arguments with Virgin about the role-playing aspect.
We wanted to make it a little bit more involved, but they kept telling us that role-playing games didn’t sell. Baldur’s Gate hadn’t yet come out when we had this argument…
We wanted something much more RPG-oriented where you had a number of characters with more involved attributes and equipment, while they wanted more focus on real-time strategy. But I think the basic system was very good.
The multiplayer mode worked very well, in fact, and I enjoyed playing it. The single-player game was a bit less successful – it was our first real attempt at more sophisticated level design, the AI of your companions wasn’t good, and it had a number of frustrations for the player.
It was designed first as a multiplayer game and we sort of retrofitted the single-player experience, so it didn’t work so well unfortunately, for me.
What can you tell us about the cancelled The Dreamland Chronicles: Freedom Ridge project for Virgin, which was rumoured, in spirit, to be a full 3D version of your original X-COM game?
Yes, it was designed as a sort of remake of X-COM for PC and PlayStation 2, and it was looking very promising actually. We were using a lot of new technology, including the Havok physics engine, which was very new at that stage. At the time we were one of the very few companies that were using it.
It was quite an ambitious project – the closest thing I can relate it to is probably Valkyria Chronicles on the PS3. We had a third-person camera view behind your character with a bar representing your Action Points, which went down as you moved.
When you went into shooting mode it went into a first-person view and you could select snap shots or aimed shots, which altered the size of an aiming circle on screen. So you did the shooting from that view, and went back to the third-person view to move your characters. In fact, when I first played Valkyria Chronicles it was quite eerie because it was a very similar system to what we had with Dreamland.
We also had an interesting destructible terrain system with lots of physics, so you could blow holes in buildings with a rocket launcher and see all the brickwork fly around, then move through the gaps, it was quite advanced for its time. Unfortunately Virgin got taken over by Interplay, who in turn got taken over by Titus Interactive.
Titus had no interest in what we were doing – they were only after Interplay’s assets, and they cancelled the project. But because we had a four-game deal with Virgin and had only done one game for them – Magic & Mayhem – we had no choice but to wind up the company at that point.
With the introduction of tactical RPGs, Gollop looks to these titles for new inspiration.
You revisited two of your old franchises with Laser Squad Nemesis for the PC and Rebelstar: Tactical Command for the Game Boy Advance. Did you enjoy going back to your roots with these games and updating them for modern audiences?
Well we wanted to wind back the scale. Laser Squad Nemesis was done by myself, Nick and another ex-colleague from Mythos. We programmed the first version in a year, and the game is still going. It started as a play-by-email system, and we later added a simple web-based interface, which gives you a list of the games you’re currently playing and allows you to download turns.
It was important that it was a sort of asynchronous turn-based/real-time game and I think it worked quite well. Rebelstar: Tactical Command was a game that just me and Nick worked on, which came about via a contact we had with Namco, who were thinking about setting up a studio in the UK.
That didn’t happen, but it turned out that they had a spare slot for a GBA game and asked if we could fill it, so we said yes. It was quite tough because we had to produce the game in eight months, and we’d never worked on the GBA before.
It had some rather awkward and unique challenges, and because it was just myself and Nick we had to find people to do artwork, but what we managed to do in eight months was not bad actually.
We were going to ask if we’d ever see another tactical strategy game from you. How did Ghost Recon: Shadow Wars on the 3DS all come about?
I started working as a game designer at Ubisoft in Sofia in November 2006. I quickly became a producer, then sometime in late 2007 I pitched the idea of doing an X-COM-style game using one of Ubisoft’s franchises.
They said they had a Tom Clancy’s Ghost Recon game due out that year so I said, ‘Okay, we’ll do a ‘Ghost Recon Tactics’.’ It was originally going to be for the DS, but last year we decided to try and get it approved for the new Nintendo handheld, which we didn’t know much about at the time.
Fire Emblem is a title that Julian claims as recent examples of excellent tactical RPGs.
What other strategy games have you admired or been inspired by over the years?
Most of the recent stuff that’s inspired me has all been Japanese. I remember playing Vandal Hearts on the PlayStation and thinking, ‘Wow, this is the kind of game I would have liked to have done if I’d had the chance.’ I enjoyed Final Fantasy Tactics, Advance Wars and Fire Emblem on the Game Boy Advance.
In fact, when I came across Advance Wars back in 2001 I was amazed
that anybody could make a turn-based strategy game in this day and age, so kudos to Nintendo for doing it so successfully.
Advance Wars reminded us of History Line 1914-18 on the Amiga. Did you ever play that? Great game…
No, but I did play Battle Isle, which I think was similar…
When you look back at your games, are there any that you’re particularly fond or proud of?
Well, X-COM – the original, of course – and probably Laser Squad Nemesis. I’ve enjoyed playing that immensely with many people.
Lastly, have you enjoyed working with Nick over the years? Any sibling squabbles, or has it always been a case of brotherly love with the odd digital alien autopsy to be coded by the morning?
[laughs] Well he has always been pretty good. He’s a very good programmer, and is obviously familiar with the kind of games I’m interested in. So by and large I’d say we’ve worked together pretty well.