A Beauty Cold and Austere postmortem

First, I am thrilled with how A Beauty Cold and Austere did in IFComp this year. I’m really, really happy with 7th place - and particularly with the game’s score of 7.89.

What follows is very spoiler-y. It’s also long.

Origin of ABCA

I’m rather new to modern IF. I had played half of the Infocom games as a teenager in the 80s, but I hadn’t played IF in at least a decade before this year, and even then it was only to check out Adventure for the first time and replay some of those old Infocom classics. The catalyst in January of 2017 was that I thought my son (at almost 9) was finally old enough to appreciate Zork. So we played through Zork 1 and 2 together. And I remembered why I had loved those games so much as a teenager.

I had had a vague sense over the years that IF had continued post-Infocom, and somehow I had heard of Curses!, so I thought I would check it out after finishing Zork 2 with my son. That led me to the IFDB, Inform 7, and this forum. And I loved Curses!. If a game that good was made in the early 90s, and the IF community had been going on ever since, how much had I missed in the past 25 years?

Probably unsurprisingly, I found myself wanting to write my own piece of IF, too. I had wanted to as a teenager playing Infocom games, and I had made a few feeble attempts, but they never went anywhere. Now, though, in my 40s, I felt I had the programming skills, self-confidence, and above all persistence that would enable me to actually complete a game. (My hat is off to those folks who have successfully written games in their teens - some of us develop a little slower!)

While playing Curses! I started forming a game in my head. The Klein bottle puzzle in Trinity had fascinated me as a teenager. It was the first time I had heard of Klein bottles or even anything from the branch of mathematics known as topology that Klein bottles are associated with. As an adult, I’ve been a math professor at a liberal arts college for over a decade now, and I spend a lot of my time thinking about how to explain mathematical concepts to undergraduates. My major professional train of thought (teaching mathematics) crashed into my playing of Curses! and my memory of the Klein bottle puzzle and produced, “I could make a game full of puzzles like the Klein bottle. Done right, it could be a teaching tool - a way to introduce mathematical concepts to people through a completely different medium.” I started working on such a game, while teaching myself Inform 7 and playing more modern IF to try to get a sense for how the genre had changed since Curses! was written.

I also searched the IFDB to see if anyone had created a game like the one I had in mind. The closest I could find were Peter D. Killworth’s GiantKiller and Harry Giles and Joey Jones’s Chinese Room. The former had all the math puzzles I was hoping for, but the puzzle structure was rather random, mathematically-speaking, and the game was aimed at a middle-school audience. The latter was very close to what I had in mind except for being about philosophy instead of mathematics.

I was afraid that I was missing something, though, and so I finally decided to reach out to someone in the modern IF community for help. That turned out to be a stroke of good fortune. The most prolific reviewer in the IF community not only knows the back catalog of IF games extremely well but is also a mathematician: Mathbrush! He was very encouraging about my project, said that he wasn’t aware of anything quite like what I was doing, and suggested several other games that had mathematical content or some kind of teaching component so that I could get a feel for what others had done. Later, he became my first serious beta tester, suggested another beta tester for me, and continued to be the person in the IF community I went to with questions as I worked on ABCA. ABCA would not be what it is without his help.

The Structure of ABCA

The continual challenge in writing A Beauty Cold and Austere was how to take concepts that are usually and concisely expressed in the static and often formidable language of mathematics and put them in an interactive, narrative form instead. With the Klein bottle puzzle in Trinity as my guide, I decided I needed to represent them somehow as concrete objects. There are some exceptions, but I tried to do this as much as possible.

I also realized early on that I didn’t want the puzzles to be haphazard, mathematically-speaking. The more I thought about this, the more I convinced myself that the game would feel more unified if the puzzle progression were to align with the actual development of mathematics. Also, since gating is something that happens naturally in puzzle-heavy IF anyway, why not put the gating to use with the theme by having it reinforce the sense of how mathematics developed intellectually? From there it was a quick step to having the history of mathematics - not just mathematical concepts - be part of the game’s content.

Some examples of this theme of the development of mathematical thought: When you first enter the dream, you have to create a number system before you can do anything else. You have to demonstrate some knowledge of algebra and geometry before getting to calculus. Zeno’s paradox is encountered early, but the solution to the associated puzzle can’t be available until you’ve learned limits, which in turn requires calculus.

The Zork area is out of place with that historical development. It was originally near the end of the game, as I had thought of it as a nod to the recent development of computer science as an outgrowth of mathematics. However, I moved the Zork area to about a third of the way through the game after feedback from beta testers, who were frustrated that they were collecting a bunch of books that they couldn’t do anything with. I still believe sacrificing internal structural cohesion for playability was a good move in this instance.


Designing a puzzle in A Beauty Cold and Austere generally went something like this: “I need to include X (where X is a mathematical subject) in the game. How can I make a concept associated with X (1) concrete, (2) interactive in a text-based format, and (3) interesting from a puzzle perspective?”

To give a better sense of that, I’ll talk about the Casino region. In the game it represents the concept of probability. Historically that works because the origins of probability can be traced to Fermat and Pascal working out some problems arising from gambling games. There are lots of examples of real-world probability problems, so making probability concrete wasn’t hard. Lots of games involve probability, so interactive wasn’t too bad, either. The difficulty with probability was the third principle: How to make it interesting? Most probability examples I could think of were either too easy (probability of success too high, or too easy to guess best strategy), too hard, or too hard to explain. A puzzle with random elements adds another twist because the player could be following the best strategy and not realize it because of an unlucky throw of the dice. The puzzle needed to be something that the player would want to keep playing even after losing. I finally took a classic problem in probability that was too hard but was easy to explain, made it easier, and turned it into a game. Thus, instead of having to guess the largest number out of 100 random numbers, you can win by guessing one of the two largest. What’s perhaps interesting and maybe surprising about that is that with the right strategy you have a better than 50% chance of winning.

Some of the puzzles succeed better than others. My personal favorite in its original form was the roller coaster puzzle. Almost immediately in testing, though, it became clear that that puzzle was just too complicated and underclued. The version of the roller coaster puzzle you see in the game is much more streamlined. (I suspect I got too excited: “I can have the player do this!” And “Let’s just add that little twist, too!” Note to self for future game-writing: If you think a puzzle is just wonderful, there’s a good chance that it’s too hard and complicated for the player.)

The first puzzle I came up with was the one with the sequences and series machine and the golden path. It was inspired by the beanstalk in Curses! - although the puzzle went through a few mutations before ending up in its final form.

The two puzzles that I had the most trouble with conceptually were the ones with the balance scales and the laser bike. For the former, I thought the game would be incomplete without some acknowledgement of algebra as a tool for solving equations. But I had already settled on the design principle of “no symbolic manipulation,” and it took me quite a while to come up with something that would represent the idea of solving an equation in physical form without using symbolic manipulation. Judging from reviewer feedback, the balance scale puzzle still needs better cluing. Hopefully the post-Comp version will improve on that without giving the whole puzzle away.

I won’t say too much about the laser bike puzzle, beyond mentioning that I really wanted the underlying mathematical idea in the game somehow. I went through two completely different ideas that I eventually discarded before settling on the laser bike.

The puzzle with Descartes started out as a throwaway joke. After a while, though, I began to be more and more bothered by the fact that the joke involves a logic error. In a game that takes mathematical thought seriously, that just wouldn’t do. The scene I came up with to fix that is perhaps my favorite scene in the whole game, although as a puzzle I think it’s not as strong as a lot of the others.

Fractals initially had a much more prominent role in the game. After I settled on the design principle of developing the mathematics historically, though, that meant they could only appear near the end.


There are a lot of things I like about GiantKiller, but I wanted more of a story than that game has. On the other hand, I had a lot of trouble coming up with a believable story that resulted in someone needing to solve a large number of mathematics-based puzzles. Since I’m a professor, it’s probably not a surprise that I soon gravitated toward someone needing to learn a lot of mathematics to pass a test. Still, even that wouldn’t give a believable reason to do a bunch of mathematics in physical form. Hence the dream.

A couple of reviewers have compared the game to Bill and Ted’s Excellent Adventure. In all the months I worked on ABCA, that comparison never even occurred to me!

For the record, the PC in ABCA is not based on anyone I actually know.

Why enter a game in IFComp that takes 5+ hours to play?

The short answer is that I didn’t write A Beauty Cold and Austere with IFComp in mind. Instead, I wrote a game because I had an idea that I wanted to make real, and then I went looking for a venue to share what I had made with others. As a first-time author and someone very new to the IF community, I thought I needed a venue where lots of people would at least try out my game. The games in IFComp seemed to be the most widely-played IF games, and so even with the two-hour playing limit for judging, I finally decided IFComp was the best way to get ABCA “out there.” See also this thread.


A shout-out to all my beta testers, as well as the folks who sent me bug reports during the competition. I’d like to single out a few folks, though, who were particularly helpful. I’ve already mentioned Mathbrush. Andrew Schultz gave ABCA a thorough beta test, and he helped me solve some coding problems with the game as well. David White gave ABCA an extremely thorough beta test; I think I had 18 transcripts from him by the time he was done. Thanks to all of you!

Also, thanks to the reviewers during the Comp for their helpful feedback. I know the reviews aren’t really aimed at the authors, but they were still helpful in getting a sense of what works well and what doesn’t.

Additional Random Thoughts

Writing A Beauty Cold and Austere was one of the most intellectually satisfying things I’ve ever done. Solving coding problems, thinking creatively and hard about getting mathematical ideas across, and trying to write language that would express those ideas exercised more parts of my brain on one task than perhaps anything I’ve ever done.

It has also been incredibly rewarding to watch people engage deeply with and enjoy something that I put so much effort into this year. Since I’m an academic, part of my professional time is spent writing mathematics papers. Few other people ever read them (usually only other mathematicians who are interested in the same niche things I am), much less comment on them. I know the IF community sometimes bemoans the lack of interest in parser-based IF, but I can guarantee you the audience for parser IF is gigantic compared to that for, say, “Asymptotic Moments of the Bottleneck Assignment Problem.”

Now that I’ve played a lot more modern IF, it strikes me how old-school ABCA is (it’s a massive puzzlefest!) and how much it mimics Curses! and Trinity in structure (in that it has a central location from which you can unlock alternate times and places by solving puzzles). I suppose that’s not surprising, given my IF background when I started writing the game.

I had originally conceived of ABCA as an experiment in mathematics education. I imagined my audience to be advanced high school students and older who are interested in exploring mathematical ideas in a non-conventional setting. However, I realized during a conversation with Andrew Schultz who I had actually written ABCA for: myself, at age 16. That teenager who was fascinated by the Klein bottle puzzle in Trinity would (I think - I hope) have loved what ABCA is trying to do.

Now what? I’m working on a post-Comp version of ABCA. I plan to incorporate some improvements to the game based on the various reviewer commentary I’ve gotten during the competition. There won’t be any major changes, though; I think significant further time spent on game development would be better put toward creating another game.

Speaking of which, while I would like to write another game, I also have a hard time mentally justifying spending as much time as I did on ABCA unless I can convince myself it’s got some professional value. So it’s likely that if I write another game it will also have a heavy math component. But… we’ll see. I’ve been inspired by some of the other games in this Comp, and that might take me in a different direction.

I also think I’m going to try to sound out the interest of the mathematics community in ABCA. There’s a subset of folks involved with the Mathematical Association of America who are always looking for new ways to explore mathematical ideas, and I suspect some of them will appreciate ABCA.

Finally, my favorite reviewer quote during IFComp was Jack Welch’s comment on the title: “Hopefully, not a love letter to Angela Merkel.” Ha!

If you’ve managed to make it this far, thanks for reading and for your interest in my game.

A challenge for people who’ve finished ABCA

There are references to Colossal Cave, Zork 1, Zork 2, Enchanter, Sorcerer, Spellbreaker, Hitchhiker’s Guide to the Galaxy, A Mind Forever Voyaging, Trinity, Curses!, and Aisle in the game. ABCA and this year’s Comp game Absence of Law also contain references to each other. Some of these references are admittedly a bit oblique. Can you find them all?


On the off-chance that anybody still has any questions remaining after reading all of this, I’ll be happy to answer them.


By far my favorite game in the competition, but a puzzlefest about mathematics is an easy sell. Very nicely done.

As someone who included some math and physics puzzles in one of my old games, yay for ABCA! I love both the puzzles and the game’s structure mirroring the development of mathematics.

Thanks, tmack and Sargent!

Sargent, which game is it? I’ve got a list of math puzzles in IF games going.

Losing Your Grip, in the 4th chapter. The puzzles aren’t that sophisticated, mathematically speaking, but there is a number line!

I’ll check it out. Thanks!