Review: Uncertainty in Games by Greg Costikyan

Book: Uncertainty in Games
Author: Greg Costikyan, game designer and academic
Year: 2013

TL;DR: A perhaps too brief look at games from the lens of player uncertainty – how uncertainty affects players and how we might categorize it different types. A design theory book on an advanced topic, but lacking some meat.


Quote from the Publisher: “In this concise and entertaining book, Costikyan, an award-winning game designer, argues that games require uncertainty to hold our interest, and that the struggle to master uncertainty is central to their appeal. Game designers, he suggests, can harness the idea of uncertainty to guide their work.” (Source)

Uncertainty in Games attempts to take a single topic – “uncertainty” – and explore its role and meaning in game design. Costikyan, the author, is a fairly acclaimed designer – his essay “I Have No Mouth and I Must Design” (PDF) is a key reading for anyone looking at the history and evolution of the field of game design. In addition, the book was published as part of the Playful Thinking series from MIT Press (like Art of Failure, another excellent book). All together Uncertainty in Games seemed to represent a text that’s both practical and theoretical, with a dense deep dive into a singular topic. However, I felt the book ultimately fell short of that promise.

The book is an analysis and catalogue of ways games create uncertainty in players, and why that is important to games. The end result is essentially giving designers a new lens (uncertainty) from which to look at and evaluate their games.

The introductory chapters (only a few pages each) simply devotes itself to defining uncertainty and wrestling with various definitions of “game” or “play”, having a kind of dialogue with some of the more famous philosophers like Callois) in this field. This is the most academic part of the book in terms of tone and content, but still easy enough to follow along as someone not entirely familiar with the source references. “Games thrive on uncertainty, whereas other interactive entities do their best to minimize it” (p. 15).

Chapter 4 starts really getting into the meat of the topic by doing an analysis of various types of games – digital and analog from very different genres – and looking at what role uncertainty plays in each. Chapter 5 follows up by taking this exploration of uncertainty and categorizing them into specific roles.

This reads strangely as though the chapters are in the wrong order – most authors would define and categorize the theory and then use examples to support it. The result is the same though.

Below are the different forms of uncertainty the Costikyan identifies:

  • performative uncertainty, such as in games of skill like Tetris
  • solvers uncertainty, as in uncertainty in the challenge of puzzle-solving. The author favors puzzle-games like Lemmings where the puzzles develop emergently from the game’s mechanics over authored/contextual puzzles like in point-n-click games.
  • unpredictability of opposing players, such as in Rock, Paper, Scissors or any given multiplayer game, or even in singleplayer games with robust enemy AI to simulate another player
  • uncertainty due to analytical complexity, such as Chess, or Dwarf Fortress, or any game where the sheer amount of systems and their interconnectivity is beyond the ability of a player to fully grasp
  • uncertainty due to randomness, such as in roguelikes or the roll of a die. One interesting callout here is that while a single throw of dice is extremely dependent on randomness, games that use many throws tends to temper that randomness with strategic choices, where the ultimate outcome of the game is largely relies on player skill
  • uncertainty due to hidden information, such as Poker where you do not know the other player’s cards. This kind of uncertainty tends to encourage exploration and testing of the system.
  • uncertainty of perception, such as in a Hidden Picture game, but also applied to games like Guitar Hero. There seems to be a heavy overlap between this type and performative uncertainty.
  • narrative anticipation, especially in games where players can upset the balance and allow an underdog to win at the last moment. This is as opposed to a game like Monopoly where you know the winner fairly early on and it takes a long time to actually resolve the game.
  • uncertainty of schedule, a fairly new type popularized by FarmVille, where the player cannot be entirely certain of their own schedule in returning to the game at important intervals
  • development anticipation, where players look forward to updates, patches, and post-release content

Uncertainty in Games fails to identify any kind of uncertainty we haven’t fully explored in games. Where do we go from here? What avenues have we yet to open up? With the exception of a nod at social games (the references to CityVille already unfortunately feeling outdated), there seem to be no new innovations in the way developers use uncertainty. The last chapter that covers games lacking uncertainty or the problem of excess uncertainty is too short to really dig into exception, contradictions, or problems arising from the gameplay element.

My writing marks up many of the pages with questions, often pointing out seeming contradictions or statements that didn’t seem quite supported by my own experience or the rest of the text. For example, the authors says that endless games like World of Warcraft have uncertainty not in the outcome, but in the path the player takes. And yet he follows along saying puzzles are not games because they have no uncertainty – if you know X then you can derive Y. But I don’t see why even a simple jigsaw puzzle can’t have uncertainty in the path in the player takes, in which piece they solve next, in how the final picture is assembled. In fact, Costikyan seems to contract himself when he covers “solvers uncertainty” (p. 25) in puzzles.

I am summarizing here, possibly unfairly, but this represents one of the many instances where I wanted to argue with the text and some of the grand statements made within it. I definitely had to resist the urge to nitpick the content in favor of focusing on the broader concepts.

I could not help but constantly compare the way Costikyan treats uncertainty to how Koster focuses on patterns in his book A Theory of Fun For Game Design. For Costikyan, uncertainty seems to be the key element in the design of a good game. For Koster, good game experiences come from experimenting with the game’s systems until they fully understand it (pattern matching), at which point the game because unfun. This period of play when the player is trying to understand the pattern – but doesn’t – is essential the period of uncertainty. These aren’t contradictory concepts but rather felt like two sides of the same coin.

However, Costikyan never really talks about a game once it’s been mastered, or the difference between the first time playing a game and repeated playthroughs of games intended to be played once. I think that’s due to a (my perception) focus on analog games for good examples of uncertainty, or games that rely on algorithmic complexity (Dwarf Fortress). A game like Uncharted also has uncertainty – especially narrative uncertainty – but after completing it, that uncertainty largely vanishes.

For a real deep dive into the topic, I’d expect the author to explore this situation and others that challenge or complicate the concept of uncertainty. Fairly or unfairly, I can’t help but compare this book to the other one I’ve read in the same series – Art of Failure by Jesper Juul. While I critiqued that one for being a bit meandering, it was extremely comprehensives, looking at the single topic from as many angle as possible. Ultimately that’s what’s missing from Uncertainty in Games and why it feels like it doesn’t go far enough.

The book itself is a quick read – 114 pages, plentiful references, and not bogged down in complicated terminology nor does it feel dry like a textbook. While I think the content does require at least some experience in game design (as a student studying the field or developer practicing in it), I found it very easy to follow along. I would recommend the book if the topic specifically interests you but don’t expect to come away with anything particularly ground-breaking.

1 Comment on “Review: Uncertainty in Games by Greg Costikyan

  1. I think there’s a great difference between puzzles that have only one solution (something I feel too many video games suffer from: you need to pick up these exact items in exactly this order and apply them as the designer intended – which means that if I play at all, I tend to end up googling the solution to get back to the fun bits) and jigsaw-type puzzles that allow you to apply a lot of different strategies and which let you feel achievement without ticking off every single item, should you get bored near the end.

    As for a game once it’s been mastered, that’s where, quite frequently, _play_ starts for me. The first time I want to get my bearings, get through the story, find a reasonably good strategy. In later play-throughs, I’m happy to try out characters/classes/strategies that are not optimal and see what cool things await me. You may not be surprised to learn that I tend to play on ‘easy’ settings: I occasionally like the challenge of getting things perfect, but most of the time, I find hyper-optimisation stressful rather than fun.

    It just occurs to me that linear puzzles – you must solve them to progress, and there’s only one path through them – are a form of hyper-optimisation in themselves.

Leave a Reply

Your email address will not be published.