Inside Emily Short's Philosophy — Make the Solving Itself Part of the Story

Introduction

Emily Short is a writer of interactive fiction — text adventures, in the older phrasing. She is known for Galatea (2000), which consists almost entirely of a conversation, and Counterfeit Monkey (2012), in which you transform objects by manipulating the letters of their names; she also wrote much of the Inform 7 documentation and its example games alongside Graham Nelson (Wikipedia).

The reason I want her on the desk is that she has stated her position on puzzle construction plainly — "I've ranted before about the value of a systematic mechanic for IF, rather than a hodgepodge of unrelated puzzles" — and, unusually, has published the working notes behind her own designs for over twenty years. The volume of primary material here is in a different league from most designers.

This is not a review of her games. It is a study of the person, built from what she has written and said on the record.

Background — a hobbyist who ended up in the industry

Her own account of the starting point is a diary. "I have a little diary from when I'm six or seven years old that recounts my attempts to solve a Scott Adams game on my VIC-20" (NME, 2023). Scott Adams wrote the first commercial text adventures in the late 1970s; the VIC-20 was Commodore's home machine. The habit of writing down what you failed to solve is already there.

In the same interview she says what caught her: "I think the constrained world of a text adventure where everything is part of a puzzle really hooked into my child brain" (NME, 2023). A world in which everything is part of a puzzle — I read the seed of her later design thinking in that phrase.

Her route to professional work was slow. "I was a complete hobbyist; I wasn't in the games industry. And I wasn't expecting to be" (NME, 2023). Her 2000s output was released free. In 2011 she co-founded Little TextPeople with Richard Evans; Linden Lab acquired it the following year. Spirit AI from 2016, creative director at Failbetter Games from 2020, departing in 2024 (Wikipedia).

The same interview contains one short remark about how she was treated once inside: "There's a certain amount of...just, you know, being overtly taken less seriously than your male colleagues" (NME, 2023). She does not dwell on it. She states it and moves on.

Philosophy — the solving is the story

The most consistent claim across the sources is that a puzzle should not be an obstacle placed outside the story. To HuffPost she put it flatly: "Solving these puzzles doesn't just unlock more story; solving the puzzles is part of the story" (HuffPost). In the same piece: "I prefer to see puzzles that contribute in a meaningful way to the themes of the story."

The claim is already present in 2007: "I enjoy writing the kind of game where the player becomes increasingly skilled at applying the rules of the game-world, rather than facing off against a bunch of unrelated single puzzles" (Game Developer, 2007). The target is fluency in a world's rules, not a run of unrelated brainteasers.

She also treats player attention as a currency. "The game should stay fun for as long as it takes to play; no aspect should take more of the player's attention than it deserves" (ibid.). And on rewards: "Every puzzle should have some reward, and a complicated or multi-stage puzzle should provide some minor rewards for partial solutions" (ibid.). Nearly twenty years on, that instruction still applies directly to daily-puzzle design.

Her observation about unsolvable puzzles is sharper still: "The more time a player spends in the presence of an unsolvable puzzle...the more importance he tends to attach to that problem" (ibid.). Time spent stuck inflates a puzzle's perceived importance — from which, I read, the author must never let an unimportant puzzle block anyone for long.

Obsessions — defining fairness before building

Her making-of for Counterfeit Monkey is worth any puzzle author's time. She lists three conditions for a productive systematic mechanic: it should suggest many interesting permutations that can be introduced gradually; it should be implementable through general rules of the world model rather than a heap of special cases; and it should have clear, expressible failure outcomes (her blog, 2013).

More striking, she fixed the shape of a "fair" puzzle in advance, to three types: you know what to change but not what to change it into; you know what to create but not what from; or you are told a general category with multiple makeable candidates. If a puzzle fell into none of them, she reasoned, it was probably not going to be fair (ibid.).

The definition functions as a cutting tool. A translation wand, foreign-language transformations, a turntable for anagramming — all dropped. "Some were bad because they were hard to clue to the player, or because they had exactly one use in game" (ibid.). Single-use is, on its own, grounds for rejection.

She also insists on making structure visible. Of the dependency chart she published in 2020: "The Counterfeit Monkey puzzle chart was one of my major design aids when I was building the thing" (her blog, 2020). The dependencies lived on paper, not in her head.

Failure — how she took the end of Versu

In February 2014 Linden Lab stopped supporting Versu — the social-simulation narrative engine she had been building with Richard Evans. Her own post is level: "What this means for me: I am no longer an employee of the Lab" (her blog, 2014-02-20).

What follows is more telling. "I am currently trying to see whether I can get back the IP for Versu and the apps that we built that Linden did not release, including Blood and Laurels", and "I may not be able to make that happen, but it would mean a lot to me to be able to do so" (ibid.). The reason she gives is not her own work but the people who had begun using the Versu toolset in academic settings and needed to keep using it.

A month later, the framing is settlement rather than grievance. "What remains to do is to accept the outcome with a good grace and plan future work that might draw on what we learned while not impinging Linden's IP" (Game Developer, 2014). Of the unreleased Blood and Laurels: "I was very excited about it, both as a work of mad science and as a story. I'd watch people play and they'd sort of swear to themselves when they got to the first big twist. It was gratifying" (ibid.).

She closes with thanks rather than blame: "I remain hugely grateful to Rod Humble and to Linden for picking us up when they did, and for giving us the run they gave us" (ibid.). Blood & Laurels did reach iOS that June (Inara Pey, 2014-06-12). Part of the wish came true.

Dilemmas — the four-body problem

The clearest account of her own conflict is a post from August 2011, about the project that became Counterfeit Monkey: "It's an obnoxiously emo process. I'm fighting the damn thing because the integrity of the story demands a clearer, a more assured resolution than I want to give it" (her blog, 2011).

She then rules out craft as the problem: "It needs to finish on a note I'm having a hard time reaching, not as a matter of technique, but as a matter of personal assent" (ibid.). She was blocked not on skill but on whether she could agree with the ending.

The same post contains my favourite sentence of hers: "It seems the three-body problem here is in fact a four-body problem: what the player wants, what the characters want, what the author wants, what the work itself demands" (ibid.). Player, characters, author — and what the work itself demands. Whether you count that fourth body changes a designer's posture considerably.

There is a structural dilemma too. In 2008 she wrote that "IF very often uses puzzles as a gating mechanism to control the player's access to parts of the story", and noted that in her own Floatpoint she had removed the blocking puzzles (her blog, 2008). The person arguing that puzzles should be part of the story is simultaneously wary of puzzles working as turnstiles in front of it.

Influences — the ones she credits

The earliest influence is the Scott Adams game on the VIC-20 (NME, 2023). On where her work comes from now, she answered briefly in 2007: "I draw a lot of inspiration from other people's work and from the feedback I receive on my games" (Game Developer, 2007). Naming other people's games and criticism of her own in the same breath is characteristic.

The same interview has a line on the closeness of making and playing: "If anything, playing IF is most like writing IF: there's a reason the community has so many members who do both" (ibid.). In interactive fiction the membrane between reader and author is thin, and she treats that as a resource.

Technically, Graham Nelson's Inform 7 is the ground she works on. She contributed to its documentation and examples, publishing Bronze and Damnatio Memoriae as annotated Inform 7 sample games (Wikipedia, Game Developer, 2007). Shipping the work and publishing how it was made are not separate activities for her.

Kizuki's reading

What follows is my own reading. I read her as someone who loved puzzles so much that she never stopped doubting them. The child who wrote in a diary about a game she couldn't solve was caught by "a world where everything is part of a puzzle", and grew into the designer most wary of puzzles becoming turnstiles in front of a story. The person who said in 2007 that every puzzle should carry a reward wrote in 2008 that she had removed the blocking puzzles from her own game. Not a reversal — the same question, pressed further.

I also have a habit of pointing at contradictions. In 2008 she wrote that sandbox-style emergent play is "really, really hard — nearly impossible — to render well in text" (her blog, 2008). Three years later she co-founded a company building Versu, a simulation-driven narrative engine. Strictly it is not a contradiction — her 2008 scepticism was about emergent puzzle solutions, not social simulation as such. But the fact stands that she staked a professional career on the territory she had called nearly impossible. That, I think, is the core of her.

And then the four-body problem: player, characters, author, and the work itself. Most design writing stops at the first, or at best the third. An author who counts the fourth will always suffer against a deadline. That she called the process "obnoxiously emo" and still refused to put the fourth body down is, to my reading, the most craftsmanlike thing about her.

Closing — where to start

The fastest way in is Counterfeit Monkey (2012). Its systematic mechanic — words as tools — lets you check by hand whether her three categories of fair puzzle actually hold up. Read it alongside the 2013 making-of (her blog) and the 2020 dependency chart (her blog) and the design opens straight up.

If conversation design is your interest, start with Galatea (2000). And her prose is itself the best material: emshort.blog is over twenty years of accumulated design notes, and there is nothing else quite like it for putting puzzle design into words.

This site has also covered Andrew Plotkin and Jon Ingold from the same interactive-fiction tradition. Reading the three of them together — three different answers to where puzzle and story should be joined — is worth the time.

Sources

Primary sources used in this article:

・Game Developer (formerly Gamasutra), "Inside Interactive Fiction: An Interview with Emily Short", Jim Munroe, 10 April 2007

・emshort.blog, "Emergent Puzzle Solutions", 13 February 2008

・emshort.blog, "Alexandra Leaving", 16 August 2011

・emshort.blog, "Making of Counterfeit Monkey: Puzzles and Toys", 24 January 2013

・emshort.blog, "Post-Linden", 20 February 2014

・Game Developer, "The end of Versu: Emily Short looks back", Christian Nutt, 14 March 2014

・emshort.blog, "Counterfeit Monkey Puzzle Chart", 20 January 2020

・NME, "Emily Short deserves her flowers" (Boss Level), Charlotte Krol, 14 June 2023

・HuffPost, "A Book Itself Is a Little Machine: Emily Short's Interactive Fiction" (interview)

Secondary references:

・Wikipedia, "Emily Short" (career dates)

・Inara Pey, "Versu's Blood & Laurels arrives in the App Store", 12 June 2014