Colossal Cave Adventure: How to Lose Yourself in a Game

It's pretty amazing how few people know about Colossal Cave Adventure (often called simply Adventure), considering how important it was for the history of video games. Aside from single-handedly originating an entire genre, the adventure game, it is also considered by many to be the first example of interactive fiction. Legend has it that every time a copy of Adventure made its way onto a computer network at a university or business in the late '70s, all work would cease for a week or two as the local computer nuts tried to finish the game. When it comes to this game's influence on the video game industry, I can't help comparing it to the Velvet Underground's influence on pop music.  Brian Eno once said, "The Velvet Underground didn’t sell many records, but everyone who bought one went out and started a band." So it was with Adventure, which despite its limited commercial impact, effectively started the careers of the genre's earliest and most well known designers, including Roberta Williams (of King's Quest fame), Dave Lebling (Zork), and Scott Adams (Adventureland). For those who are interested in the history of interactive fiction, I suggest The Digital Antiquarian, an incredibly detailed and well-written blog on the subject.

There are multiple versions of Adventure, the first of which was written solely by Will Crowther (left) in 1975. However, the version that I'm reviewing here, which is usually considered "canonical", is a modification of the original that was developed by Don Woods (right) in 1977. The different versions are distinguished by the maximum number of points that they allow the player to accumulate -- 350, in this case. Later developments included a 550-point version by Dave Platt and a 581-point version by Michael Goetz.

Adventure can be a challenging game to get into on first approach. There are no graphics, so players must interact with the game world entirely through a two-word parser. Most commands consist of a verb followed by a noun; for example, "open door" or "kill bear". There are only a small subset of verbs and nouns that the game recognizes, but importantly, the player does not know what these are ahead of time. Interactions that require more than two words are possible, but only if the game decides that more information is needed. Typing "kill bear" causes the game to respond, "with what?", at which point you would respond, "with pineapple", which might or might not cause the ursus to succumb to a Vitamin C overdose. Later text adventures, like Zork, would implement a four-word parser that made the use of prepositions more straightforward.

Navigation is another challenge. While the player usually navigates between rooms using eight standard cardinal directions, they aren't always reversible. For example, just because you went northeast to get from the Ancient Crypt to the Great Chasm, doesn't mean that going southwest will take you back. As such, mapping the connections between rooms is essential. There are many online software tools to facilitate adventure game mapping, but my personal favorite is Trizbort.

What really seems to bother modern gamers is that many of the puzzles in Adventure have solutions that are... unusual. Don't expect to get through this one just by opening doors with keys and killing dragons with swords, the game is going to throw some really weird stuff your way and you're just going to be expected to... you know, deal with it.

Such illogic is a novice design mistake, right? After all, how are players supposed to solve puzzles that don't have logical solutions? Well, not so fast.  I grew up with graphic adventure games, most of which were created in the mid-90s and beyond, that followed a strict design formula that prevented things like unexpected player deaths, illogical puzzle outcomes, and events that make the game unwinnable. I enjoyed playing many of these games and acknowledge many as classics, but after playing Adventure, I can't help feeling that such design restrictions come with a price. When undesirable outcomes are artificially prevented, the player loses freedom; that is, it makes the game more linear and thus more artificial. It's remarkable to me the lengths that some modern adventure game designers go to in order to create bizarre, even surrealist fantasy worlds, and yet present the player with nothing but conventional logic puzzles. For some reason, it's perfectly acceptable that the player would be confronted by one Rube Goldberg machine after another, yet it's totally out of bounds for fantasy worlds to be fantastic and unpredictable. It seems to me that in the conventional wisdom of adventure game design, the goal is not to explore, but to make the player feel smart for having finished the game.

Rubbish. The first time I played Adventure was a revelation. The difficulty was frustrating at first, but as soon as I learned to stop approaching the game as a notch for my belt, I began to see what Roberta Williams and Scott Adams had seen in it back in the '70s. Just like them, I learned to approach the game as something to live with -- I ruminated on dwarf management while falling asleep, jotted down possible means of killing dragons while eating my lunch, and in my free time typed unusual game words into the terminal to see if they would have any magical effect. Instead of trying to treat every interaction as a deductive puzzle to be solved, I began asking myself, "I wonder what would happen if I did this." And sometimes nothing interesting would happen (and sometimes the game would make fun of me for trying), but when something interesting did happen, it was a blast. The fact that some modern gamers seem to wish for an Adventure without man-handled dragons, "plughs", or twisty little passages is crazy to me.

And is it really a big deal if you have to consult a walkthrough for some puzzles? Even in the early days of adventure games, playing was a community experience, with people sharing secrets at the water cooler or over a local network. Roberta Williams even allowed players to call her at home for hints to her early graphic adventures. If a particular puzzle is taking so much time that it’s ruining the experience, look it up and move on with the adventure.

I recommend this game very highly, but only if you're prepared to spend some time with it. Early computer games were difficult in large part because of their scarcity — once you were done with it, there was no online store to provide another quick gaming fix. So games were designed to take a long time. We now live in a world at the opposite extreme, where any number of games are available immediately and at little-to-no cost to the player, and the idea of devoting a week to solving a handful of puzzles seems crazy. But it's much more satisfying than you might think, and all of that time you spend puzzling over that shadowy figure on the other side of the chasm, far from being wasted, is time you spend immersing yourself in the game world.