The Making Of: Archon
Archon’s mix of arcade combat and chesslike strategy ensured it stood out when released in 1983, and the game is well remembered today, still feeling unique and fresh after almost 25 years. “Before Archon, there were two basic styles of computer game: slow, deliberate, turn-based, cerebral games (chess, adventure games, or military strategy games); and ‘mindless’, fast-action, ‘twitch’ games, which were typically clones of coin-op arcade games,” begins Jon Freeman, explaining the thinking behind Archon’s multi-genre approach. “This seemed to us a false dichotomy. Why not mix the two elements, reward quick thinking, and balance strategy with action? This seems obvious now, but lots of people at the time weren’t at all sure it made sense.”
Convinced a combination of game styles was the way forward, Jon Freeman, Anne Westfall and Paul Reiche started developing the game, spurred on by having signed the first two contracts with fledgling software company Electronic Arts. “We wanted a game that required – and rewarded – thinking, but that wasn’t as ponderous as chess,” recalls Jon. “Adding arcade action – something you couldn’t do with an ordinary board game – would keep the game moving and unpredictable.”
Although keen to exploit the possibilities videogames afforded over board games, the board itself came first. “Its design was the intersection of an intended visceral reaction and sheer technical practicality,” explains Jon. “We wanted something fairly similar to a chessboard, but different, so people would have two immediate, contradictory reactions: ‘It’s a chessboard’ and ‘It’s not a chessboard’ – that is, ‘I get it’ and ‘It’s different’. Both of those feelings were important.”
The chess-like set-up gave players a frame of reference, but the rules, pieces, ever-changing board and combat elements were all new. The board became a nine-by-nine grid, governed by the limitations of the Atari 400/800’s builtin tiles and sprites – during development, ten squares per edge had crowded the screen, while eight or fewer merely provided a de facto chessboard or limited potential for tactics. The nine-by-nine board also created vertical and horizontal central lines that defined the game’s important squares: the power-points. Rather than capturing the opponent’s king, the aim in Archon became to control the five power-points. (Alternatively, obliterating the opposition leads to victory for those players preferring violence over strategy!).
“Variable squares – impractical and unheard of on cardboard, but simple for a computer – served as another means of differentiating Archon from conventional board games and keeping the game fluid,” adds Jon. Working out the colours took a few iterations, and the final design was a mix of aesthetics and tactical balance. “The starting points were instantly obvious – clearly, the dark side’s home square had to be black, and the light side’s had to be white,” says Jon. “Equally obviously, to get the kind of dynamic game we wanted, the three power-points across the centre of the board had to be variable squares.” And so, as the game was played, about half the board would cycle from light to dark. This wasn’t a mere aesthetic consideration – Jon and co made it so that during combat, light pieces were stronger when on lighter squares and dark pieces benefited from being on darker squares.
Combat soon became Archon’s defining feature, setting it apart from its contemporaries. Unlike in many traditional board games, there’s a direct element of risk in capturing a piece, since both pieces are transported to an arena to fight it out in a frantic one-on-one arcade game. “Keeping things from bogging down was an obsession for us,” says Jon, explaining the reasoning behind the combat component. The flexibility and ever-changing nature of the board, combined with the arcade battles, meant that memorising opening sequences, as in chess, was no longer the order of the day. “Doing that seemed to us the antithesis of ‘play’, and so we tried to make it impossible or pointless,” explains Jon. “The arcade battles rewarded quick thinking and fast reflexes, but setting up the battles – picking the site, conditions, pieces, timing and circumstance – involved strategy and deliberation.”
With 16 unique pieces (eight per team), the variety of match-ups is huge, ensuring replay value is high, and Jon reckons that “the fact the battles turned out to be interesting, fast-moving and exciting was crucial to the game’s success”. Like the board, the arenas aren’t static: barriers fade and reappear, to keep the action moving. “We were afraid that if barriers were static, players might ‘park’ their icons behind an obstacle and only occasionally pop out to fire,” explains Jon. Interestingly, double-kills are also possible if two pieces strike home while both are low on energy. “We considered ways to avoid this, but the alternatives were more work and felt arbitrary,” muses Jon. “Double-kills had been a part of the live chess game I’d been in, which was a major influence on Archon, and we liked the impact they had on the board – when both pieces were gone, it was time to rethink strategy!”
The pieces and weapons themselves arrived from days of brainstorming on the part of Paul and Jon, who scoured fantasy and mythology books, aiming to find things associated with light or darkness, and come up with plausible matches – that is, corresponding pieces. “From the outset, we tried to balance the sides and seem to have been pretty successful,” reckons Jon. “We had a sense of how the pieces should – and would – play, and the tweaking of pieces and weapons during play-testing was surprisingly minor. Most players developed a favourite side, but neither side has a clear edge.” A further strategic edge and additional fantasy flavour arrived in the form of each team’s key figure – a wizard or sorceress – who can cast spells, such as teleporting a piece or reviving a deceased character. “The choice of spells was pragmatic: they made strategic sense; they fit the game; and they were doable without generating a lot of extra work for our overworked programmer,” says Jon.
Perhaps surprisingly, one of Archon’s most important components was something of a last-minute addition. “Archon was designed as a two-player game,” says Jon. “At the last minute, EA demanded a single-player option. In retrospect, this was absolutely the right thing to do, but at the time, in the face of an implacable deadline, the uncertain results of so much unforeseen work terrified us!” In just one month, the team had to develop, from scratch, the entire strategic and tactical AI, and this feverish work resulted in several breakthroughs. “Before Archon, it was assumed computercontrolled enemies couldn’t out-duel a human player, and so for opposition, arcade games relied on swarms of stupid – even mechanical – baddies, fast-moving objects, unforgiving terrain and sharp limits on player movement,” explains Jon. “We were confident we could make a computer opponent that was good enough and fast enough in its ‘thinking’ to compete with human players.”
Anyone who played strategy games in the early Eighties will attest to how notoriously slow many of them were. Jon recalls that “an early and not terribly complicated SSI space war game took as long as 20-30 minutes per move for the computer to make up its mind” – and this was fairly typical. In part, poor programming was to blame, but Jon reckons the primary cause was the mistaken assumption that intelligent decision-making was necessarily slow. “For Archon, that premise was rejected out of hand,” he says. “Obviously, the arcade action on the battlefield had to be fast, but it was also clear that a slowdown on the strategy board would ruin the flow of the game.”
The team decided a decent, fast AI was preferable to a perfect, slow one; it was also important that the computer’s behaviour was not too predictable. “Streamlined logic and efficient programming – and a lot of thought and sweat – produced an acceptable move in only 1/60th of a second,” says Jon, proudly. “It wasn’t perfect, but it was good enough – and it was so fast we had to slow it down artificially by adding a visual delay to give the impression that the computer was ‘thinking’!”
Archon’s one-player mode then led to another of the game’s selling points, albeit a rather more curious one: the ‘no player’ attract mode. “This started out as a play-testing tool and ended up as one of Archon’s most popular features,” says Jon, tantalisingly. “Once EA decided Archon needed a one-player mode, it was obvious players would want freedom of choice; therefore, the computer had to be able to play either side. In the process of implementing this, Anne realised a computer player – an AI – that could play either side could, with a few minor code changes, play both sides. Suddenly, we had a game that could play itself.”
Initially, this was great from a testing standpoint – the team could observe and test the computer AI on both the board and the battlefield. “Having the computer play itself exposed twice as much of its behaviour compared to playing a human,” says Jon. “It also became an effective, efficient way to demo the game and, ultimately, a powerful sales tool. After Archon’s release, computer stores put an Atari in store windows, with the screen facing the street or the mall, and simply let the game run. Hey presto: instant crowds! As far as I know, Archon was the first ‘arcade’ (ie action) game that could play an entire game, start to finish, with no human intervention.”
Despite the hectic schedule as Archon’s development drew to a close, the project was a success. Arguably instrumental in EA’s successful debut, the game garnered plenty of publicity, favourable reviews and numerous awards. “Archon ended up on many top-ten lists and was named ‘game of the year’ in a couple of places. However, the computer-game market was smaller then, so Archon didn’t sell the millions of copies you’d expect today from a big PC or PS2 hit,” says Jon. That said, the game still sold well, and the C64 conversion ended up outperforming the Atari original, although the Amiga version fared relatively poorly.
Today, Jon spends much of his time licensing Freefall’s classic games on new platforms, and while he notes that “many negotiations drag on for a year and then mysteriously collapse, which is frustrating”, he’s hopeful regarding upcoming next-gen versions of Archon. In the meantime, the original game is there for everyone to play, and fond memories of creating it remain for Jon: “I have great memories of play-test sessions in our basement with Paul, Anne, myself and Robert Leyland, who programmed Murder On The Zinderneuf – our ‘other’ game with EA. Despite the 100-hour weeks, it was such an exciting time, and we were all so enthusiastic that I don’t think I have any bad memories of working on Archon. And at almost every game conference we attended, people told us Archon was their favourite game, or that it was the reason they got into the games industry – how could we not feel great about that?”
