Recognizing Fun Through Elevator Pitches

Once upon a time, I spent three months writing an IF game called Five Gods Exiled. It was a spectacular failure.

Five Gods Exiled was a Frankensteinian attempt to combine the core mechanics of Arkham Horror with the procedural generation of ZAngband and execute the whole thing in Inform 7, overlaid with an epic science fantasy narrative.

This is one of my favorite failures because I learned so much from working on it. I could probably write ten articles inspired by that game (everything from “Best Practices in Procedurally Generated Landscaping” to “Why I Love Lists And You Should Too”), but the primary reason it never saw daylight was that it wasn’t any fun to play.

It was a whole lot of fun to build. But it wasn’t any fun to play.

You need to know whether or not your vision for a game is fun. If it isn’t fun, then it’s not worth making. (Or, if the goal isn’t “fun” – engaging, or worthwhile, or effective, or whatever you’re aiming for. Not all games are “fun”, but every game design has some goal that equates to “the player will not regret having spent time playing this game.”)

This is not to say that every fun game has its Most Important Fun Part from the beginning. Experimentation is real and valuable. At a Boston Postmortem talk, Joe Mirabello (Tower of Guns developer) mentioned that his game changed dramatically because of a well-received bug in a playtesting build. It does happen!

…but it’s a risk. If you don’t have a plan for “here is what makes my game fun”, or you don’t prototype your Fun Element to make sure it really is fun – then you risk spending X months (or years!) on a game that is, in the end, Not Fun.

This is why formal game development cherishes rapid prototyping. The idea behind rapid prototyping is to create the smallest, fastest version of some element of your game, and then rebuild it and rebuild it until you have a good version. Such as the important element of Why Is This Fun.

When I started working on Five Gods Exiled, I knew about rapid prototyping. I knew about the importance of finding the fun. But what I didn’t know was how to break the game down into components. I saw the game as one complex melange, and I trusted that, if I built the whole thing, it would be fun.

I love making games, and I talk with my friends about my current projects. I wound up trying to explain this project repeatedly (at different stages of development) to one of my friends. It was a two-to-five minute spiel, and every time I finished, she frowned and shook her head. “I don’t understand what you’re trying to do.” “That sounds really confusing.” “What exactly is the player doing?”

Eventually, I would just wave my hands and say, “Look, it’ll be awesome when it’s done.” And that should have been a giant warning sign.

I didn’t have an elevator pitch because I didn’t know where the fun in the game was. I explained the game as a collection of features instead of a single clear vision. I thought that a bunch of good elements had to add up to a good game. And I polished the heck out of that turd.

Now, I should note that not everyone is good at elevator pitches. If you’re bad at elevator pitches, that doesn’t necessarily mean that you have a bad game idea. It may just mean that you’re bad at elevator pitches.

Fast forward to 2014, when I wrote the first draft of a fantasy novel. I knew it was good, but my ability to explain it was very, very bad. Just as bad as my explanation of Five Gods Exiled, in fact (though the overall reception was far more positive).

At Arisia 2015, I had the fortune of a practice pitch session with renowned author NK Jemisin. I had a ten-minute block to pitch my novel and hear her feedback.

I walked in the door. She leaned back in her chair, looked me straight in the eye, and said, “Pitch me your novel.”

I said, “Uh.” Then I rambled for five minutes or so about the main characters, the worldbuilding, the start of the plot, and the themes.

She listened patiently until I trailed off. She asked a few clarifying questions and I scrambled through my answers. She eventually said, “You’re going somewhere good, but you have no hook whatsoever.”

“I know,” I said sheepishly.

“Look. What do you think is the most interesting thing about your novel?”

I told her.

Jemisin said, “Here’s your hook: ‘Finch’s friends and family think she’s dead. They’re probably right.'”

I probably swore (and apologized), but I don’t remember for sure. I just remember how obvious it was once she said it, and how bewildered I was that I hadn’t seen in a year of telling people about my novel. I walked out of that room with my head reeling and two minutes to spare.

Back in the world of game design, Jemisin’s advice still holds true. “What is the most interesting thing about this novel?” equates to “What is the most fun thing about this game?” Having an answer to that question means that you can prune a vague design vision into something concrete, clean, and – most importantly – testable.

If you know which part of your game is supposed to be fun, then you can prototype that aspect without building the rest of the game around it. You can build the fastest possible version, show it to some players, and verify that, yes, This Was Fun. And if you’re wrong, and it isn’t fun, then you can back up and try something else.

Should your answer be “The environment is procedurally generated” – which was, inevitably, my answer for Five Gods Exiled – then you have a problem. That answer belongs to a tech demo, not a game. I don’t regret those three months (I still steal tech from it, after all!) but I wouldn’t choose to repeat them. Writing elevator pitches for my games, even just in my notes, ensures I won’t do so.


Thank you to everyone supporting Sibyl Moon through Patreon!
If this post was interesting, useful, or entertaining, please consider becoming a patron.


Bookmark the permalink.


  1. Wow there are a lot of lessons there. I’m glad to see there’s someone other than me who deplores the recent (~4 years) spate of games that substitute “procedural generation!!!1!” for actual creative effort. Procedural generation is a tool. It’s perfectly possible to use that tool to produce something fun, but experience suggests it’s not easy. Much more often you’ll just end up with something which is “okay, I guess, but why would I play it again?” It’s better than the huge string of multiplayer-only RTSs from last decade, but… not much. At least Starcraft II put an end to that noise.

    It’s a good point though. You’re writing the darn thing. You know literally everything about it. If you can’t figure out what about it is fun, that might be a sign. A bunch of interesting elements, jammed together, is not necessarily fun. To be fair, it might be! Friedmann Friese’s Power Grid is 3-4 separate games, united only by the fact that you have to spend the same money to play them, and the victory condition is set up to reward winning at all of them (this last clause being Friese’s genius). This worked pretty well, but almost by accident. If I tried to give an elevator pitch for Power Grid, I couldn’t do it. Is it auctioning? Is it market manipulation? Is it building an empire across a map? I don’t know either. (I actually don’t like Power Grid, but all my friends do.) Compare Friese’s new game 504, for which the elevator pitch actually is this exact phenomenon: every world has a different combination of these 9 systems. One determines victory, one determines income, one “yields additional flavor”. Make the best of the world you’re in today.

    But all the neat elements in the world jammed together don’t necessarily make a good game, just a good tech demo. You might still need to write a game. You definitely need to know whether you still need to write a game, and if you can’t put your figure on “this is the game, right here”, that’s a hint that you haven’t written it yet.

  2. Because I have Spirit Island on the brain, I thought about its early days as I read this. In particular:

    “The idea behind rapid prototyping is to create the smallest, fastest version of some element of your game, and then rebuild it and rebuild it until you have a good version. Such as the important element of Why Is This Fun.”

    The very first version of Spirit Island had no spirits – or rather, no spirit-specific systems. Elements didn’t exist (though I knew I wanted them eventually), innate Powers didn’t exist, even unique Powers didn’t exist – I had each player grab 2 Minor Powers at random as their starting cards. The core question was “is playing these power cards to fight off the Invaders fun?” It turned out the answer was “yes”; people really wanted to play more despite wild power imbalances, very crude information presentation, and several mechanics in desperate need of more baking time.

    (Then, along with refining + dropping existing things as I went along, I layered other things atop that core – sometimes preplanned, like Elements, sometimes things I discovered along the way, like Fear.)

  3. I’ve found elevator pitches a lot simpler since I understood that “high concept” is pretty much the opposite of “highbrow”. I always think of “Snakes on a Plane”. You instantly know it’ll be stupid and funny and claustrophobic with some scary snakes and probably cheap gags like jump scares and snake jokes. And of course, the line about “I’m sick of these mother-f—ing snakes on this mother-f—ing plane!” The shorter your description, the better the pitch. But I also find a great pitch doesn’t mean the game itself is good… but that’s another story :)

    Oh, and it’s fun (and surprisingly useful) to summarise great works of literature. “A Hobbit throws away a ring (eventually).” “The evil wizard Voldemort is killed (for reals this time).” etc.

Leave a Reply to Eric R Cancel reply

Your email address will not be published. Required fields are marked *