The best-known computer puzzle games of the 1980s were each built from a few rules that their designers could state in a sentence. A warehouseman can push a crate but never pull it. Seven shapes fall and must be packed without gaps. A boulder falls if nothing holds it up. A lemming walks until something stops it. The depth came from how the rules combined, and several of the designers later explained how they found those rules. Alexey Pajitnov, who wrote Tetris, said the hard part was the opposite of what players might expect: “you don’t need to be really smart to create really hard puzzles — it’s not correlated. You need to be really good to create easy puzzles!”
Fast facts
- Sokoban (1982): Hiroyuki Imabayashi, Thinking Rabbit, Japan. Push crates onto marked squares; you can never pull one back.
- Tetris (1984 by most accounts): Alexey Pajitnov, Moscow. Seven shapes, because seven “is very good number to remember”.
- Boulder Dash (1984): Peter Liepa, First Star Software. Rocks, dirt and jewels governed by one set of physics rules.
- Lemmings (1991): DMA Design, Dundee. Eight jobs to give out, built up “bit by bit rather than designed from the outset”.
- Pajitnov’s distinction: a classical puzzle has “the full information, which is set up once”; randomness and real time “wash out” the puzzle and turn it into a game.
- The reviewers’ rule: “very simple in concept and play, but also addictively challenging” (Computer Gaming World, 1988); “Accessibility alone determines whether a puzzle game lives or dies” (Edge, 2001).
Sokoban: one rule and no way back
Hiroyuki Imabayashi wrote the first Sokoban in BASIC in the spring of 1981, and in 1982 he founded a company, Thinking Rabbit, to sell it, according to his own account on the game’s official website. Soko-ban is Japanese for “warehouse keeper”. The player moves crates around a warehouse floor plan onto marked storage squares. Crates can only be pushed, never pulled, so one careless push into a corner makes a level impossible.
The first cassette version had 20 stages, the American magazine Computist reported in 1989, and it would not let a player take back a move. “You can only push boxes forward or horizontally, never can pull them behind nor can get back to your previous moves,” its reviewer wrote. A crate pushed by mistake, often thanks to a key’s auto-repeat, meant restarting the stage. The magazine blamed this “rather strict (rather stoic) policy” on Imabayashi, the game’s programmer, and said most Japanese players considered it a defect. Later versions gave way: the Famicom edition, with 150 stages, allowed unlimited moves back.
The game reached the United States in 1988, when Spectrum HoloByte published it alongside Tetris. In Computer Gaming World Roy Wagner reviewed the two together and found that they had the same quality: “Two of these are somewhat similar to non-computer shape puzzles, but totally new from anything currently on a computer. These games are very simple in concept and play, but also addictively challenging.” He compared Soko-Ban to the sliding fifteen-puzzle, which it “takes its start from … but carries it quite a bit farther”. “You are only able to push the crates. Walls, doorways, and other crates makes this process challenging. It is quite easy to push a crate into a position where it is trapped.” This version allowed one step back, “but only if you do it before making any other moves”. It had 50 levels, scored each one by the number of pushes and the time taken, and came with a construction option for building new screens. Wagner’s verdict was “pure and simple, very playable and mentally challenging”.
Tetris: few elements, and time pressure
Pajitnov took Tetris from pentominoes, a puzzle of twelve shapes made from five squares each. He told Edge in 2001 why he cut them down: “Five squares was too much because people can’t remember 12 shapes, so I simplified it down to four. There are seven of them, and seven is very good number to remember.” Wagner described what that did for the player in 1988: “After a little play it is rather easy to quickly rotate and position each piece and drop it into place. Your challenge then becomes the ever increasing speed level.” His verdict: “Tetris is deceptively simple and insidiously addictive.”
In 2007 Pajitnov gave Edge a longer account. He had started writing puzzles at the Moscow Computer Centre, which was “one of the small number of open organisations [in the USSR] which dealt with computers”. New machines arrived from abroad to be tried out, and “the best way to deal with a new computer is to try to write a small program and see how it works. A game was a very good example. And puzzles are some of the easiest things to program.” Tetris itself was “1,600 lines of code — 27 kilobytes, once compiled”.
He drew a line between a puzzle and a game. “A classical puzzle is something with the full information, which is set up once. But those games are random, so that you have variety in the task every time, and you don’t know the configuration. That gives great variety and interest to the game, but at the same time it doesn’t allow you to make the puzzle really deep and complicated. Realtime is one element that ‘washes out’ puzzles, and the other one is this randomness.” By that measure, “Tetris is not a pure puzzle game. It has a very strong arcade element.” Computers, he said, can “turn puzzles into games”: “According to public opinion, puzzles are dull and take too long to solve, but computers really increase the speed and dynamise them.”
Boulder Dash: rules first, then the reasons to play
Peter Liepa built Boulder Dash the same way, from the rules up. “The game’s original graphics were crude — just squares, circles and spaces — and what evolved was a cellular automaton that encoded the basic physics of what was to become Boulder Dash,” he told Retro Gamer in 2006. A cellular automaton is a grid in which every square follows the same few rules, so that complicated behaviour grows out of simple parts. Liepa filled the screen with random rock and dirt and dug through it with a joystick. “But this was a long way from being a game, I needed some way of keeping score, and so I added jewels. And I had to add risk, so now you died if you were hit by a rock.”
He did not design the caves square by square. “Randomness and determinism are used throughout the game,” he said. “The boulders and jewels fall predictably, but randomness is used to place them.” A random-number generator scattered the rocks, dirt and jewels at chosen densities, and Liepa laid walls, fireflies and butterflies over the result. He told Edge in 2010 that the themes of the caves often came out of those random layouts: “You’d take a theme like chasing, evading or destroying fireflies and butterflies, and make it a main element of a cave … There was no formula for this — it was just like jamming on a musical instrument, and the possibilities seemed endless.”
Lode Runner, the one Pajitnov admired
Asked by Edge which puzzle games he admired, Pajitnov named Doug Smith’s Lode Runner (1983), in which the player digs holes to trap guards while collecting gold. “I consider it a real puzzle game. That was my favourite for many, many years. It’s such a good combination of dynamic stuff, the fingerwork, and real planning and puzzle-solving. Every level was a real puzzle with its own mechanics, its own dynamics, its own kind of solution. At the same time, it was an arcade game to play.” His other choice was Pipe Dream, which Edge noted was originally known as Pipe Mania: “That’s the other absolutely classic, classic game. It will never die.”
Lemmings: skills added one at a time
Lemmings began as an animation, not a design. Dave Jones told Edge in 2001 that a DMA programmer had drawn “a bunch of little guys walking up a steep slope, and being blasted at the top by a big gun”, and that he “just thought – there’s a fun game in that”. The rules came later, and by trial. “The game was built up bit by bit rather than designed from the outset,” Jones said. “Because it was very unique, there was nothing to benchmark it against.” The animator Gary Timmons described the same process: “We all chipped in ideas for skills for the lemmings, and then I had to animate them to see which ones worked best.”
Eight jobs survived: Climber, Floater, Bomber, Blocker, Builder, Basher, Digger and Miner. Edge saw in the game “one of the most basic design principles: simple elements combining to produce complex interactions”. A Basher could cut through a pillar of rock and a Builder could bridge a pit, and it was “the vast combinations of these functions which made Lemmings”.
What keeps a player playing
Pajitnov described what drives a Tetris player as a loop with no story to it. “Every moment you feel you could do a little bit better, and lower this garbage by one line. And you do it and you discover that you’re higher than before. Also you have a very nice feeling when you fit pieces together, and it comes in a good rhythm.” Shooters, he said, “plan some other emotional picture”: an easy start, a challenge, a stable period, a harder challenge, and then a boss. “Tetris doesn’t have this overall picture. It just happens to be repeatable stuff with a certain rhythm.”
Two more of his rules came out of the same interview. The first was abstraction. Edge suggested that “most successful puzzle games are usually completely abstract in their presentation”, and Pajitnov agreed: a particular style, such as ancient Egypt, will “cut out a big part of the audience”, “but when it’s abstract, nobody gets annoyed.” The second was that players must feel themselves improving. His complaint about PopCap’s Bejeweled was that “no matter how good a player you are, your score is always about the same. You don’t have any feeling of your own progress.” His own answer was Hexic, which he designed for Microsoft.
Accessibility, 2001
Edge’s review of Minna De Puyo Puyo for the Game Boy Advance, which it gave eight out of ten, opened with the principle stated as a law: “Accessibility alone determines whether a puzzle game lives or dies”. Tetris, it said, “hits the spot by throwing in a mere handful of elements”, while a newer game, Hot Potato, “stumbles because it muddles the puzzle ethic by presenting six individual components for simultaneous juggling”. The Puyo Puyo series had “walked the fine line between complexity and manageability”: drop blobs in pairs, match four of one colour to clear them and send the leftovers to your opponent. “Ostensibly, that’s it. But dig a little deeper and soon you’ll be mastering the chain-reaction manoeuvres which lead to advanced Zen-like play.”
Sources
- Hiroyuki Imabayashi, “ごあいさつ” (Greeting), official Sokoban website, sokoban.jp (web, secondary): the first version in BASIC in spring 1981; Thinking Rabbit founded and Sokoban released in 1982. https://sokoban.jp/greeting.html
- Computist issue 70, 1989, Japanese software column, “Soko Ban”: the cassette version of 1982 with 20 stages, the no-undo rule, Imabayashi as programmer, the 150-stage Famicom version, Thinking Rabbit as distributor.
- Roy Wagner, “Puzzling Encounters: Two Titles from Spectrum HoloByte’s International Series”, Computer Gaming World issue 47, May 1988, pp. 42–43: Soko-Ban (designer credited as ASCII Corporation; 50 levels, the single step back, the construction option, the verdict) and Tetris (speed levels, the verdict).
- Edge issue 100, August 2001, making-of feature on Tetris: Pajitnov on pentominoes and seven shapes.
- “Puzzle quest”, Alexey Pajitnov interview at the GameCity festival, Nottingham, Edge issue 183, Christmas 2007, pp. 68–71: the Computer Centre, 1,600 lines and 27 kilobytes, classical puzzles and randomness, Tetris as an arcade game, motivation and rhythm, Lode Runner and Pipe Dream, abstraction, Bejeweled and Hexic, easy and hard puzzles.
- “The Making of: Boulder Dash”, Retro Gamer issue 27, 2006, pp. 24–27: Liepa on the cellular automaton, jewels and risk, randomness and determinism.
- “The Making of… Boulder Dash”, Edge issue 220, November 2010, p. 103: Liepa on random layouts and cave themes.
- “The making of… Lemmings”, Edge issue 94, February 2001, p. 136: Jones on the animation and on building the game “bit by bit”, Timmons on the skills, the eight professions and Edge’s comment on simple elements.
- For Tetris’s date and Doug Smith’s Lode Runner, see the Tetris and Lode Runner entries.
- Minna De Puyo Puyo review, Edge issue 105, Christmas 2001, p. 91 (Game Boy Advance; Sega, Sonic Team/Caret House; eight out of ten).