An essay on adapting books/stories into parser games

I’ve been wanting to do some advertising for my Spring Thing game, but there are not any standard channels for advertising parser games.

So I thought of writing a series of blog posts about game design and development that are also related to my game.

Today I want to talk about adaptations of existing stories into the parser format. This is a problematic area, with a lot of major pitfalls, as I’ve learned from experience.

First, I want to talk about the three most well-known adaptations of texts: The Hitchhiker’s Guide to the Galaxy, The Hobbit, and The Tempest. A fourth game, The King of Shreds and Patches, has also been very successful, but it was adapted from an RPG module, while this essay is focused on adapting linear texts.

The Hitchhiker’s Guide to the Galaxy

This game was Infocom’s second-best selling game, after Zork. It also has been released to the public by Douglas Adams (you can play it here). Unlike all other parser adaptations of famous books, this one was (at least theoretically) directly under the control of the author. This would allow the author to adapt the story to be more parser-friendly.

I say theoretically, because, as the Digital Antiquarian says: “The latter [Douglas Adams] mostly provided just the text for the direct path through the game, leaving Meretzky to deal with all of the side trips and the incorrect and crazy things the player might try as well as any of the boring bridging passages that Adams couldn’t be bothered about.”

Hitchhiker’s Guide to the Galaxy is also a very, very hard game. Almost no one could get out of the first general area (the house and the pub) without hints. It was on a timer and required you to perform some very, very specific actions.

However, players continue to remember it with fondness, likely because the writing was so good, and because, at the time, expectations about fairness were different. Andrew Schultz commented on the Digital Antiquarian’s description of the unfairness of the game, saying “It’s good to read a breakdown of why the game could be unfair. I remember my mother buying my sister and me the invisiclues early on (we were stuck in the dark! Explaining this would be a spoiler) and my feeling was that adults must’ve known a lot of stuff I didn’t to be able to solve these puzzles without a hint book (I was sure they did) and I was looking forward to growing up and understanding all this. Especially how I’d know about the ways you can be locked out of certain puzzles and how the game is mean about that.”

The Hobbit

The Hobbit, by Melbourne House, was a rare Tolkien-Estate-approved adaptation of the classic book.

It was crammed into a small amount of memory (around 48K) and featured both graphics (which would slowly fill in as you play) and independent characters (like Thorin, who is fond of singing for gold).

Like Hitchhiker, the Hobbit struggles with forcing a linear narrative into a non-linear game. The game is essentially set up in discrete chunks, each a major part of the game, and you are supposed to guess the right action in each chunk that advances the story. These chunks are still connected physically, though, so you can travel from the hobbit to Rivendell and back in just a few screens.

This is another horribly difficult game, especially given its randomization and independent characters whom you have to rely on. Even following a walkthrough, the game can become unplayable without any fault of your own.

It was very popular, however. Having recently played it on an emulator which was set at the original speed, it is thrilling to see each screen load in slowly, tracing out the boundaries of the graphics and filling them in, and seeing an old-feeling adaptation of a great book.

The Tempest

This game, playablehere, was by Graham Nelson, author of Inform and Curses!, among others. Graham had released the incredible Jigsaw in 1995 and won IFComp in 1996 with The Meteor, The Stone, and a Long Glass of Sherbet.

This game was entered in 1997’s IfComp under a pseudonym, where it took 25th out of 34.

It is a straight adaptation of Shakespeare’s The Tempest, written in iambic pentameter. All responses are in Shakespearean-esque language. For instance, here’s ‘You can’t go that direction’:

“The eight points curl there yet with cloud; but one,
Principal and most prickly, might bear all.”

and here’s ‘Press any key to begin’:

“Pray depress key, that albeit depress’d
Cheerly unlock this our trumpery-chest --”

It was innovative and poetic, and, yet again, completely unplayable. I don’t know anyone who’s been able to get past the first few scenes, and I have no concept of what the ending would be like.

Due to the difficulty of writing new Shakespearean text, and for fidelity, most of the game consists of ‘find the correct action to unlock more text’.

Other adaptations:

There have been other adaptations. Peter Nepstad did two excellent games based on Dunsany’s work. These, too, have you find the magic action that advances the story. It is similar to the two Manalive games, adaptations of a G.K. Chesterton novella.

There are few Twine adaptations. CMG did The Griffin and the Minor Canon, and I felt it was better than a parser game, but still suffered from the ‘do the right thing to unlock the next part of the story’.

My game: Sherlock Indomitable

This is usually the point where an author says, ‘Here’s how I overcame the problem’. Instead, this is the point where I will say, “Boy, now I understand why so many people had problems. Adaptations are so hard!”

For my intro comp/spring thing game Sherlock Indomitable, I chose to adapt two Sherlock Holmes mysteries (The Speckled Band and The Six Napoleons) into the same format I used for Color the Truth. Color the Truth has been my most successful game up to this point, with 44 ifdb ratings (only 2 parser games have had more ratings since 2014) and good reviews. Several of those reviews wanted to see more games in that clue-based format, and some mentioned Sherlock Holmes. I also got reviews saying my prose was weak, and I decided at the time to do a straight adaptation of Sherlock Holmes, using my clue-based format.

I was well aware of the pitfalls of adaptations, but I had a master plan. I would make the actual exploration in the game simple and linear and put all the ‘real action’ in the mind. Players would gather clues exactly the way it occurred in the written text, and then they would ‘link’ clues to each other (as in Color the Truth) to produce deductions. This part, not shown in the text, would make the player feel like they were really Sherlock Holmes, making brilliant deductions.

What I discovered is that adaptations are a horrible trash fire. Writing an adaptation is like spreading frosting on a hot cake, with the top of the cake ripping off and the frosting melting. Most of my time I’ve spent on Sherlock Indomitable has been trying to stop it from melting into a hot mess.

What is difficult about adaptations?

Books are not like parser games

In books, the question you always ask is, ‘What happens?’ The story is driven by current events that occur in sequence, with unexpected results to normal actions.

In most of the best parser games, the question is ‘What happened’? Anchorhead centers on uncovering a town’s past. Hadean Lands deals with investigating an accident on a ship. Galatea is puzzling out the history of an animate construct. Some have more action, like Counterfeit Monkey, but much of the best moments of that game involve piecing together mentally what happened to various people and objects, and learning about the two protagonists.

This is because parser games are very good at static descriptions, such as past events, and because the openness of the command structure invites tentative exploration and punishes fast events (one reason timed parser games never took off). Stories thrive on action, while parser games do not, in general.

Difficulties in adapting Holmes

Sherlock Holmes has been difficult to adapt because there are just so many characters. The Six Napoleons has Sherlock, Lestrade, an assistant, a Mr. Barnicot, a German manager, the Harding Brothers founder, his assistant, a thief, a Mr. Sanderson, Mrs. Hudson, a telegram boy, and on and on and on. Each of these people needs a description and some sort of reaction to conversation.

I was able to tone down much of the difficulties with conversation by restricting most characters to a TALK TO interaction. More advanced characters (Lestrade, Watson, and important players) can converse on almost any of the game’s 53 topics. This has required a lot of writing to put in their reactions.

Another difficulty is the linearity of the stories. Adaptations either have ‘guess the impossible combo to advance’ or ‘push the button to advance’: too much difficulty (like Hitchhiker and the Hobbit) or pure boredom.

This requires a lot of scene programming in Inform 7. That language is well adapted to scenes, but so many things can go wrong. When the vicious Dr. Roylott barrels into your apartment and bends a steel poker in half, what’s to stop the player from going outside and catching a train, leaving Dr. Roylott mid-sentence? It’s difficult to know how far to corral a player.

Expectations about Sherlock Holmes

Everyone has different expectations of Sherlock Holmes. I wanted this to be a cerebral game, with an emphasis on deduction and reasoning, almost more Poirot than Holmes. Many intro comp voters expressed dismay about this, preferring an active Holmes, using magnifying lenses and dusting for prints. It’s hard to find a balance.

It’s also difficult finding a place for Watson. Of course, Watson is only in the original novels to provide an excuse for the story format and to be a sort of reader-avatar. Many games have you play as Watson instead.

I discovered, though, that Doyle frequently describes Sherlock’s depressions, and mentions that the admiration and respect of Watson and Lestrade are deeply meaningful to him. I decided to personify Sherlock’s depression in his ‘house of the mind’ as a flooded basement whose waters creep ever higher up the stairs. Watson’s purpose, then, is to tame the depression. Once per story, they rise too high, and the player has to talk with Watson until you’ve impressed him enough, and subdued your depression enough, to move on.


The ideal story structures for books vs. parser games are so different that porting one to the other directly is less than satisfying. When more freedom in adaptation is allowed, one can be more successful. Toby’s Nose, by Chandler Groover, is an original Holmes story focusing entirely on Holmes’s dog Toby and uses a completely non-linear format. It’s one of the most brilliant games I’ve played and a major influence on my own writing.

I will never do another direct adaptation; the payoff is just not there for author or reader. The story, as originally described, must still be told, and so you just wait around for the next chunk to appear. Perhaps one day someone will adapt Spoon River Anthology; until then, I don’t expect parser adaptations of linear stories to experience great success.

I plan on continuing this post with at least one more on conversation styles in parser games, and perhaps more after that.


I believe this has been done:

In what ways do you think it would have been easier or harder to adapt a book to a choice based format?

@bg thanks for the link!

@dfabulich I think half of the difficulties are still there. There’s still an ‘ordained path’ that you have to unlock. I haven’t played many adaptations in choice based formats, but I think what inkle did with 80 Days was the best; they just took the voice and spirit of the original and made it very nonlinear.

Max Gladstone has done really well with his choicescript games, too, but he also isn’t doing straight adaptations of his novels. I’d be really interested in hearing more from him on how he approaches the two differently; he’s the real expert here.

There’s also something to be said about Fahrenheit 451 which doesn’t so much adapt the book as much as it complements it.

I seem to recall that Tower of the Elephant worked surprisingly well as a game while staying close to the original text. Probably because the original story was pretty game-like to begin with.

One of the most infamous adaptations is James Clavell’s Shogun by Dave Lebling.
Making use of much of Infocom’s impressive later technology, it was big and fancy, but woefully disappointing as interactive fiction.

Jimmy Maher said it best: “The player who has read the novel will find little interest or challenge in pantomiming her way
through a re-creation of same, while the player who hasn’t will have no idea whatsoever what’s expected of her at any given juncture.”

Jimmy Maher’s essay on Nine Princes in Amber is interesting–he describes it as something that looks like a “guess the next action to do what the protagonist did” game but really lets you choose your own path through the world. I haven’t played the game or read the book–has anyone tried this?)

(I just played Tower of the Elephant and did find it to be a “follow the path” thing–in particular there’s a while where you have to follow a character, at least in the path I chose, and then I got flummoxed by a puzzle where I felt like I wasn’t getting clues to the action though it made sense. Though this is just me being mediocre at puzzles, probably.)

1 Like

It is possible that I was lucky with Tower of the Elephant and just happened to hit upon the correct path and correct solutions. It was a long time ago and my memories are vague, but I played it without having read the original Robert E. Howard story, and when I later read it I actually preferred the game version.

Of course Melbourne House also published Beam Software’s Sherlock in 1984 although that is more inspired by rather than a direct adaptation of a Holmes story. NPC behaviour is improved over The Hobbit. The will react to things being said even if they are not the direct target of the conversation or statement.

Zenobi also did a Sherlock Holmes adventure in 1990.

Melbourne House also did Lord of the Rings and Shadows of Mordor both based on Lord of the Rings in 1986 and 87 - it was too big for one release. Lord of the Rings had switchable (in game via “become”) protagonists - Frodo, Sam, Merry and Pippin. I never got hold of Shadows of Mordor to play.

Level 9 did The Secret Diary of Adrian Mole Aged 13 3/4 and The Growing Pains of Adrian Mole which I never played.

I played the first of these. It wasn’t their usual parser type of game but rather a choose from multiple options all the way through. Actually very like Twine or Choice of Games style today!

If I remember correctly it had the original book plot as the central core, but deviated from it all the way through, depending on what you chose. It’s a game that definitely rewarded replaying with different choices.

Actually thinking about it a bit more I think it’s really interesting that a company renowned for parser games went for a CYOA style in this case. Presumably because they thought it was the best fit to the book.

It also seems likely to me that Level 9 was aiming the Adrian Mole games at a more “casual” audience, fans of the book that were unfamiliar with games and computers. Before their time if that was the case.

EDIT: Their adaptation of The Saga of Eric the Viking the year before is an interesting comparison. A traditional parser game, much less commercially successful and a lot less fun to play today.


If anybody’s interested in picking up this thread, I have a related tangent I’d like to throw out. It hit me a year or so ago that a naive model for “source material for games of the 1970s and 1980s” would heavily weight things like “being known enough to provide their own branding,” “being public domain,” and “offering lots of opportunities to kill or be killed.”

And so I ask: where were all of the Shakespeare games this model would predict? I know there are a few (Avon, and a Macbeth game or two, for instance) but there’s way way less than “this is literally the best-known author in the English language, and by the way you can go very deep into storytelling or just provide a ready backdrop to punch the French in the face” would seem to spark.

The best answers I can come up with:

In the mainframe/early home computer wave, I think these classic veins of material got crowded out by the crowds of game-developing nerds that were enthusiasts of one or both of these genres:

  • The sword-and-sorcery intersection spearheaded by the likes of Tolkien and TSR
  • 19th century adventure/fantasy, from Alice In Wonderland to Treasure Island.

Infocom’s tentative and decidedly wobbly experiences with direct adaptations aside, companies in the mid 1980s sprang up to focus exclusively on adaptations, like Windham and Trillium/Telarium. They, also, avoided anything older than the 19th century.

Presumably Shakespeare, Homer, and the like were just not considered cool by the alpha hackers of the 1970s and the marketing and editorial/acquisition execs they hired to run their companies in the 1980s. But where were the other opinions and fandoms? They must have been out there somewhere, right?

I recall Steven Levy’s Hackers making a pretty strong case that if they were Into It at MIT or Stanford, it propagated into early computing culture. By extension, it seems like the fixations and fandoms at other institutions must have had some sort of expression but it died out in just an electric generation or two like a (something-something-Conway’s Life-reference.)

So was there a cadre at University of Michigan which was big into Shakespeare? A clutch at Texas A&M which rabidly quoted Emily Dickinson at one another? A lab at Carnegie Mellon structured entirely around the Oresteia?

1 Like

I think it was just that “genre” was a more popular source than “literature”. I’m using “genre” in the condescending-and-negatively-defined sense of “not in the literature section of bookstores, and definitely not taught in literature classes” – SF/fantasy/horror/mysteries. (Not romance, for clearly sexist reasons. Although the visual novel movement turned that around in the 80s and 90s.)

Dickens and Tolstoy were just as neglected as Shakespeare and Homer, whereas Dracula and Sherlock Holmes games popped up by the dozens.


I’ll buy that, particularly for the earlier wave. Now that you say it, it’s obvious that the Scott Adams library is practically a genre inventory.

You are carrying:

A treasure cave game
A pirate game
A spy game
A spooky game
A different spooky game
A spaceship game

So my assumptions overweight literature and underweight genre for that early wave.

It still makes me wonder a bit about the 1984-era no seriously adapting books is the thing we most want to do publishers, how much was “we just think this stuff is cool and that stuff is boring” and how much was “the audience’s tastes are already defined by the last five years of genre gaming, please see the Adventure International back catalog.”


Another thing, when you think about Shakespeare plays that offer an opportunity to kill or be killed, is that the most famous ones are the tragedies where the protagonist dies at the end. The winner is someone who comes to clean up the mess like Fortinbras or Albany or Macduff. And while doomed protagonists aren’t totally unheard of, Lord of the Rings and Treasure Island and even Alice in Wonderland seem much more in line with the wild videogame fantasy of being given a task and then completing that task. (I believe this is the original of that occasionally-copied tweet.)

Something like Henry V would be more in line with the traditional videogame arc, but Henry V is less of a unique Shakespeare character. (There’s a couple of Agincourt board games.)


Funny you should mention Hal/Henry V, in a related discussion today that devolved into the overlooked potential for Shakespeare AIF I pointed out that there’s an entire play (H4P1) devoted to the idea that Hal is basically the Mad Party Fucker and that if you squint Taming of the Shrew is The Farmer’s Daughter. (Never mind the endless prospects for wall-to-wall Midsummer Night’s Dream-themed banging.)

(my bigger, not-just-text-adventures point on Shakespeare being under-exploited for basic fight-until-somebody-dies gaming is it’s not immediately obvious why a “two barbarians hack at each other until one is dead” simulator is much more marketable than a “Hamlet and Laertes duel and there’s a poisoned sword powerup” simulator. Although the “genre > literature” answer does play there as well.)


You forget “fitting into 16 KB”. Literature is full of fresh new ideas. Genre fiction is grounded in well-known tropes, and therefore offers better compression rates.


But this too would be compatible with Shakespeare, considering that many of his original ideas are worn tropes now.

1 Like

Shall I compare thee to a summer’s day?
…Better stick with a parrot. Fewer bytes.

I get your point but you can’t fit lavish descriptions of inheritance and breakfasts into those confined tape-loaded memories either. Didn’t stop Tolkien’s influence from looming large.


(okay who wants to collaborate on a for-reals-would-fit-in-a-16K-TRS-80 Adams-esque Hamlet?!)


One option is to comedically critique the ordained path, that’s what Ryan North has done with his Shakespeare adaptations. You have to be a very funny writer, but if you are, then there’s a lot to be gained from responding to the player choosing the source material’s path with bewilderment or sarcasm.