Grand Thieves & Tomb Raiders

by Rebecca Levene

  In EMI’s world, entertainment arrived in three-minute consumable chunks, and it saw computer games in the same way. So Elite, years in the making and taking months to play, presented by students from the geekier end of an intellectual world, seemed to these confident media men like a category error. It was a fantastic technical demonstration, they told Bell and Braben, but games needed three lives and a ten-minute playing time. They needed a score, and objectives, and an appeal to casual users. They needed to be like arcade games.

  It was a dispiriting moment for the two students. Working on their own, looking only to each other for assurance, all they had was their own instinct that Elite was worth playing. Were they wrong? But they had other industry contacts through Bell’s earlier games, and so they tried the nearest of those: Acornsoft.

  The company could hardly have been more different from Thorn EMI. Hermann Hauser and Chris Curry were frantically busy with the success of the BBC Micro, and in any case were not really games people, so David Johnson-Davies had been given a free hand. When the two developers took him their game – Braben later recalled that he was working out of an office at the end of ‘a valley of bins’ – Johnson-Davies was enthusiastic, as astonished by the graphics as everyone else, but also mesmerised by the breadth of the game. He had some reservations: he thought the scale should be reined in to make the universe look like a handful of vast galaxies, rather than an unconquerable mathematical formula. And the ability to trade narcotics gave him pause – that there was profit to be had from drug-dealing seemed the wrong lesson for the nation’s educational computer to be teaching.

  But he didn’t hesitate to sign them up. Their relationship with Acornsoft felt natural: ‘The fact that Acorn and Acornsoft were within an easy cycle ride – walking distance even – did help’, recalls Braben. The advance was one thousand pounds each and he used some of his to buy a genuine BBC Micro.

  For a further few months the pair debugged and tweaked – the galaxies were cut back, but the narcotics stayed in – until the product was bulletproof. During this time, Acornsoft published a Spitfire simulator by Geoff Crammond called Aviator, which used similar, but much more spartan, wireframe graphics. The pair appreciated it but its small environment, featuring three buildings and suspension bridge, didn’t compare with Elite. There was still nothing else like it on the market. They remained nervous that something soon would be, though, and by summer 1984 they were ready to release the product. And then Acornsoft made them wait.

  Johnson-Davies could see that Elite was big, perhaps huge, and wanted to create a marketing buzz. This game used unheard-of technology and could take over players’ lives for weeks. If he managed this well, he could charge twice the going rate.

  His ploy was to make the sense of depth and quality visible throughout the product. It was only a few years since games had been sold with leaflets in plastic bags, and the form now was not markedly more sophisticated: a cassette box with a paper inlay, like albums at the time. Acornsoft had already differentiated itself by packaging its software in large cardboard boxes containing a plastic moulded berth for the tape or disc, and an A5 leaflet of instructions, but for Elite, Johnson-Davies planned unprecedented luxury: the game would have a thick, illustrated manual, written in the style of a pilot’s guide within the fictional universe. The BBC Micro was portrayed in a hand-drawn picture as part of the console of the spacecraft, and the instructions were rich with mythmaking about the vast world to be explored.

  Some details were real features of the game – the police, the pirates, the asteroid-mining lasers – and some were inventions that hinted at a greater universe. Players were left to wonder if there really were vast generation ships, or dredgers that ate other craft, and rumours that someone had found one would circulate playgrounds and magazines for years afterwards. And there was more: a novella written by renowned fantasy author Robert Holdstock. His story gave character to an already well-defined universe: a first for a game, and tremendously effective. The package was topped off with a poster identifying the most common ships, an aide-memoire to help untangle the complex controls, and an entry card for a competition open only to players who managed to reach ‘Elite’ status.

  Such lavish packaging took time to put together, but Johnson-Davies was also holding back for an autumn release and the vital Christmas sales. In the meantime, Bell and Braben were busying themselves with thoughts of a sequel. They toyed with the idea of the player taking a role in the military, so that rather than playing free form, they would have a role in a team. One of the pair’s first jobs was to tackle the slightly confusing radar system they had developed – in Elite, it showed the battle from two planes, which the player had to co-ordinate in their head. Braben and Bell tried a revised version with a squashed 3D map of the space around the craft, and the game became instantly more playable. In a sense, it was unfortunate: they felt almost an obligation to show it to the overwhelmed Johnson-Davies, who was two weeks from going to press.

  Despite the enormous disruption it would cause, he agreed that the new radar should go in, but the work to make this happen landed right back on the young developers’ shoulders. They stayed up late implementing, testing, debugging, and finally reproducing screenshots for the manual, but they hit their mark, and were ready for Johnson-Davies’ planned launch in September 1984. The press were invited to Thorpe Park, then promoting a science fiction ride called the Black Hole, where they watched Bell and Braben, each a boffinish figure in a shirt and tie, launch a ship into space on a giant projection screen.

  The response was rapturous. Every magazine and newspaper that covered the launch glowed with praise. Some missed its scope – they had only an afternoon to play a game that takes an age, after all – but the graphical leap was applauded, and the sight of eight galaxies of hundreds of planets made the scale clear, even if the variety on offer wasn’t quite understood.

  And consumers burned with anticipation. In an era well before television review shows or YouTube promos, the descriptions of this incredible game seemed tantalising, even too good to be true. Could these screenshots be real? Did these space ships really fly out of the screen, or was there some trick?

  The game became the nation’s best-seller as soon as it was released. Johnson-Davies was right to risk the higher price point – the size of the package made buying it feel very special. As it happened, Acornsoft had neglected to take into account the extra space folded paper needs, and the box was a couple of millimetres too small for the content. The distributors managed to squeeze it in, so when consumers ripped the polythene off, they found the box literally bursting with goodies.

  There’s a popular story that the game sold as many copies as there were BBC Micros in the world: around 150,000 of each. In fact this comparison takes a generous view of the timeline – by the time the sales of Elite had reached this figure, the BBC Micro was well into its lifecycle, during which it sold 1.5 million units. But the tone of the story is right. Children without their own BBC Micros did buy copies of the game to play in school at lunchtime, and the competition cards that were packed into the game became much sought-after – players with pirated copies of Elite often wound up buying the genuine article to acquire one. To enter the competition required reaching Elite status, which meant hundreds of hours of gameplay. As sacks full of cards arrived at the publisher, it was clear the game wasn’t simply a success at the till, it was using up millions of hours of British leisure time.

  It was certainly Acorn’s flagship game. If a home user, especially a games player, had a BBC Micro, it was assumed that they also had Elite. For a year, until conversions appeared on other machines, gamers with the dowdier, teacher-friendly computer could hold their head high in the playground.

  Almost by accident, Elite advanced the professionalisation of the games industry in Britain. While negotiating with Johnson-Davies, Bell and Braben had retained the rights to release Elite on computers other than the BBC Micro. In the modern games industry, Acorn might have
tried to make the game exclusive, to boost sales of its computers. But it was a hardware manufacturer trying to meet demand, and its thoughts were about product and production. It was only long after the contract was signed that Chris Curry thought of Elite as a means to promote the Micro.

  ‘The thing that really brought home the importance of games in the BBC computer time was when David Braben designed a watershed game: Elite,’ Curry says. ‘It gave you this wonderful combination of manual dexterity, trading and planning and fighting, which all needed fast graphics, and the BBC computer really was the only one around that could do it properly . . . it really didn’t work anything like as well on anything else.’

  However it came about, the two developers found themselves in the happy position of owning the rights to publish the country’s best-selling game on its most widely owned computers. ‘The BBC Micro was not the biggest market at the time,’ says Braben, ‘but it meant we held on to the rights to the game – something that proved very wise!’

  Elite became one of the first games to be sold to British publishers via representation. Jacqui Lyons was a literary agent, acting for authors of books, and in radio, television and film, when this new industry began to enter the public consciousness. ‘Computers fascinated me, though I couldn’t program,’ she says. ‘I recognised it as a completely new form of entertainment which was bound to grow as the public became more computer literate. I thought computer authorship was an extension of authorship.’

  She negotiated on behalf of Bell and Braben with an industry hungry for ZX Spectrum and Commodore 64 versions of their masterwork, and eventually rights were sold to Telecomsoft, to publish under its Firebird label. The pair hadn’t done badly out of the Acorn deal, but this was a different order of income for the 20-year-olds. ‘I had a six-figure income and was thoroughly enjoying it,’ says Braben.

  The conversion work introduced them to other programmers, as well. They brought in Jez San, who had built a ‘Programmer’s Development System’ that allowed a Commodore 64 to be programmed remotely from a BBC Micro. Meanwhile, a young man called Peter Molyneux tried, and failed, to win the contract to convert the game to the ZX Spectrum.

  Although Bell and Braben both stayed on at Cambridge, their world had changed. There were plenty of games that earned their writers a wage while at school or college. There were even some writers with fast cars and businesses. But Elite was a tidal wave that carried everyone with it. ‘When I went to university, writing games was my hobby and the university was my work,’ says Braben. ‘By the time I left, the university was my hobby and writing games my work.’

  It’s hard to overstate the importance of Elite in British gaming history. It changed the expectations, and probably the economics, of being a games developer. The list of its innovations is a catalogue of game-design touchstones, some of which took decades to reproduce: open-world gameplay, freeform objectives, optional missions, wanted levels, player rankings.

  Elite was the product of its authors’ vision. Under the closer direction of a publisher, it might have been released earlier, with less ambition. It’s hard to believe that anyone other than hobbyist developers could have created it at all: the incredible tricks that they relied upon, the inspired innovations, and the time they took would all be impossible under commission.

  There was a case where Acornsoft came close to demanding impossibly pioneering work, though. When it secured a sponsorship deal with a Formula Three team, David Johnson-Davies approached Geoff Crammond, and asked him to do what he could to make a simulation of it. And what he could do was revolutionary.

  The racing genre was just starting to develop a form in the industry. Arcade games like Pole Position used various tricks to give the illusion of movement with a full-colour screen. A flat black road would be shown disappearing towards a vanishing point on the horizon between green fields and a blue sky. Small background details would give the feeling of movement, while video tricks would swing the ‘point’ of the road from side to side to give the impression of turning corners. It was in no way a simulation, or even particularly realistic, but it did a good job of evoking a speedy 3D racetrack.

  The Acornsoft deal had secured them the services of David Hunt, the younger brother of the seventies Formula 1 World Champion James Hunt. In the hands of some publishers, Hunt’s endorsement might have meant a photograph slapped onto the box of a standard racer. But Acornsoft offered Crammond a working relationship with a genuine racing team based at Silverstone, and a trip around the track as Hunt’s passenger. Crammond quit his job at Marconi and went to work on his new project: Revs.

  It was a perfect project to bring Crammond’s skills, stubbornness and tenacious finesse to the fore. Unlike any racer before it, it was designed like a proper simulation – it had a three-dimensional racetrack, complete with banked corners, undulations and genuine bends that you could see in the distance. Whereas previous racers simply couldn’t accurately show the road beyond a corner, Revs included accurate S-bends, a first-person view of spinning off the tarmac and even, if you wanted to experiment, driving backwards around the track.

  To have managed this with the same wireframe that Crammond employed in Aviator would have been an achievement, but he insisted on making the graphics full, solid colour. ‘I felt that I wanted it to visually stand up against the arcade games,’ he says. ‘Doing it as a simulation meant I would be using a 3D mapped track, so I did realise that it wouldn’t be easy to get the graphics to cope.’

  Working full-time was useful: ‘I was able to experiment with all sorts of ideas to get the graphics fast enough.’ Some were incredibly ahead of their time, such as his innovative self-modifying code, which meant that the program rewrote itself while the game was running to become more efficient. And he overcame the memory shortage by storing data as pixels on the screen: ordinarily this would have made the sky appear as a multi-coloured mess of spots, but he tricked the hardware into drawing all colours as blue for that part of the screen.

  What the player knew was that for the first time, their speed and racing line mattered, and perfecting these to shave seconds off a lap time was utterly addictive. The bumps in the road could throw your car, and leaving the track meant spinning into the grass, rather than bumping along the side or being shoved sideways back into the centre. These things are a given for any racer now, but controlling a car with speed and traction for the first time was a revelation.

  And David Hunt’s involvement did help, at least a little. Driving Crammond around Silverstone, he showed the importance of throttle to steering, and afterwards, how the brand-new tyres had worn away at a forty-five degree angle. He played the game, and gave feedback that encouraged Crammond to believe that his simulation had some realism. Modelling the car’s contact with the ground was a phenomenally complicated job for a slow processor, but Crammond refused to cut corners with his racer – players could even adjust the angle of the tail wing, and all of these features mattered. Revs elevated racing games from a mildly distracting toy to a potential obsession. Even if an observer marvelled at the graphics, they might miss the depth of the game they were watching.

  The scope and ingenuity that players could expect from their games were expanding. Arcade style titles continued to dominate the home computer markets by volume, but the games that attracted admiration were the ones that stood apart from the norm. Typically they showed some technical wonder that would draw an audience in, but then reward the player’s dedication with incredible breadth of scope. Like Elite, they were the antithesis of three-minute arcade play, and a labour of love for their makers.

  One such landmark game, which appeared soon after Revs, was Mike Singleton’s The Lords of Midnight for the ZX Spectrum. It was trumpeted by its manual as ‘the world’s first ever epic game’, and in those early days of unclaimed territory, this wasn’t outrageous hyperbole. The player’s eye view was of an ice-ridden landscape, with mountains, castles and forests stretching into the distance. The influence of Tolkien was
transparent, but the world was well depicted, and saw the player uncovering a fantasy plot of war and stealth, and engaging in alliance or battle with thirty independent characters. But the incredible, implausible innovation was that the landscape was not decorative – the trees and buildings were features of the game world that could be found and explored. The Lords of Midnight featured nearly four thousand detailed, connected locations. Even after Elite, the scope of the game was staggering.

  Like Bell and Braben, Singleton had used impressive tricks to hold an epic fantasy realm in a few tens of kilobytes of memory. The map was hand crafted, but each location only took the tiniest sliver of memory – the view was compiled from the details of neighbouring locations, near and far. The inhabitants were determined by the smallest possible unit of memory – literally a 1 or a 0. The details of who or what the player would encounter were inferred from the co-ordinates of the location.

  The Lords of Midnight was another home coding odyssey. As every spare byte of memory was scraped out of the machine, the code had to be broken into parts. Each ‘version’ of the game was stored in ten separate files, which had to be meticulously adjusted in sequence for any change. Singleton became very careful about backups.

  It took half a year of mostly full-time work to complete. Singleton had the backing of a publisher, but it was an agreement more than a commission – the project’s combination of fastidiousness and scale seems outside the capacity of conventional, project-managed development. As it was, Singleton’s painstaking months paid off: players and critics alike were consumed by the game. They had to be, if they were to make any headway – guiding the epic story to some kind of conclusion took an investment of weeks. Before launching his game, Singleton completed a test run, with a complete knowledge of the map and the winning strategy. It took him nine hours.