The Iterative and Incremental Nature of Game Design Dr. Lewis Pulsipher Copyright 2008 Lewis Pulsipher Teachgamedesign.blogspot.com Pulsiphergames.com pulsiphergamedesign.blogspot. Who am I Designed my own games while a teenager Began playing commercial wargames in 1963 Played the original Atari 2600 and have played some PC games heavily Designer of six commercially-published board wargames (most recently February ’06, reprint this spring) Active designer of games (playtesters solicited!) First to teach game design in the state as far as I know (Fall ’04). Recently taught nearly 100 beginners at Wake Tech Presently writing textbook(s) about how to design games, and how to teach people to design games March 23, 2016 What is today’s subject problem? Students, and even instructors, who don’t actually design games, don’t understand the process of design Hence “game design” classes often become “game production” classes, and the most important part of design (the last 20% that takes 80% of the time) is skipped And students think, when they have a working prototype, they’ve “succeeded” March 23, 2016 The last 20% "One of the hardest things when dealing with students is to convince them to work on small games and complete them, rather than working on a single huge sprawling mess that dies under its own weight. It's also hard to convey the 80/20 rule, that 20% of the work gets you the first 80% of the game... but getting that last 20% of the game (which is the polish factor) takes a lot of time after the game already feels like it 'should' be done, and pressing on when you're sick of working on the game already is what separates the developers from the wannabes." Ian Schrieber, http://teachingdesign.blogspot.com/ March 23, 2016 Delusions of “wannabe” designers “I designed a game” because – I thought of an idea • Ideas are “a dime a dozen” • Publishers don’t publish ideas – I talked about that idea with others • OK, not a bad thing to do – I made a prototype and my family loves it • Or at least, they said they did – It’s reasonable to work on just one game at a time – I tested the prototype with other people several times • Good, but how much, how many, how long? – I’ve made a beautiful submission version (non-electronic, usually) • Publishers want good gameplay • Looks don’t matter in prototypes All delusions March 23, 2016 What’s the “artificial” way? Lopsided treatment of the process in “video game design” books – Emphasis on design “up front” rather than as a continuing process – This derives largely from the expense (time and money) of making electronic prototypes – Give impression that virtually all the design is done before a prototype exists March 23, 2016 Artificial way 2 Yet this is like any form of planning: “no plan survives contact with the enemy”; and that “contact” is actually playing the game If the fundamental design is bad, there will be no saving the prototype; but good playtesting and modification can make a weak game pretty good Poor playtesting and modification will make even the best game weak March 23, 2016 Artificial way 3 Costs too much to make prototype on “speculation” – Electronic games require all the “rules” (specifications) before a prototype can be made – And the “rules” are much more complex, because the computer must understand them, not a human with human intelligence – So “real” playable prototypes (as opposed to prototypes to show concept) are very complex and time-consuming Studios spend much time and effort persuading publishers to pay them on the basis of a plan (description of a design) March 23, 2016 Artificial way 4 Industry habit of running out of time – Publishers make schedules and the “suits” don’t understand the iterative nature of the process – Electronic games are engineering problems: and no one can say with certainty how long is needed to solve an engineering problem – “You can’t gather nine pregnant women and have a baby in a month”—games (and especially game design) take time – So frequently, studio runs out of time and a poor game is published – “Patches” can do only so much to repair weak gameplay; they’re mainly for “bugs” – Yet we do see games that are mightily improved by the first patch, because of additional incremental modification March 23, 2016 What’s the “natural” way to design a game? Get ideas, figure out roughly how the game will work, and try it out almost immediately Work on the game for what seems like forever, testing and making modifications Finish (or nearly finish) the game, then approach a publisher on the basis of the actual game, not a plan for the game “Narbacular Drop” (= Portal) Even then, the published game will often be much different from the submitted game (Portal again?) Development studios that can take as much time as they like are doing it the “natural” way– e.g. Blizzard March 23, 2016 When we get serious what the “wannabes” should be doing We’ve tested it with many people until I’m SICK of it • • • • The last 20% of polish takes 80% of the time Plan-monitor-control (modify)-replan Iterative and incremental This is where the real designers are separated from the “wannabes” I actually tried to sell it to a publisher (some wannabes do try) Wannabes often self-publish It was published and I am being paid or have been paid for it March 23, 2016 The game design process Adams and Rollings talk three phases – Conception, – Elaboration, – Tuning (which is iterative and incremental) I’ve expressed the process a different way, in Data Flow Diagram terms March 23, 2016 March 23, 2016 Data Flow Diagram notes There are seven processes (activities), not in a particular order—several are going on at the same time Each process has sub-processes not shown here There are also data stores, data flows (arrows), and agents external to the system (rectangles) March 23, 2016 List of processes Conceive and refine ideas Play game in “mind’s eye”—thought experiments Conceive game, structure, framework Create and refine prototype Write notes-rules-software Solo playtest Playtest with others March 23, 2016 Prototypes—”testing is sovereign” To truly improve a game, you must have a playable prototype – Civilization I—Sid Meier programmed, Bruce Shelley tested, they talked, Sid modified, Bruce tested again, day after day – Sid said recently on slashdot, "My whole approach to making games revolves around first creating a solid prototype and then playing and improving the game over the course of the 2-3 year development cycle... until we think it's ready for prime time. My experience in this area helps me to know what to do and where to start. I definitely spend a lot of time playing the game before I let anyone else look at it." March 23, 2016 Prototypes – The sooner Firaxis got a playable version of Civ 4, the more they could learn The rules for a non-video game are the equivalent of the programming of a video game – Programming must be precise and is very time consuming (game engines may help some) – A playable set of rules can be much less precise, relying on the mind(s) of the designer(s) when they’re present, and (inevitably imprecise) notes March 23, 2016 Proof in publication How many games aren’t very good until the first patch? That’s because the studio did not take (or was not allowed) enough time to finish the incremental, iterative process of improvement March 23, 2016 Art vs. science Game design is 10% art, 90% science Art is something created by an individual, then presented to the public “as is” – There is no “testing” or “focus groups” Science is something subject to repeated testing – And almost all good games are thoroughly playtested – A sign of an “amateur” designer is insufficient testing But it IS a creative endeavor, not mechanical March 23, 2016 Teaching design: non-electronic much more efficient For teaching purposes, non-electronic games teach students much more about game design Prototypes are quickly produced – It’s much easier to produce the physical prototype, than to create the artwork for a video game Simplicity lays bare the real process of design – Novices tend to assume the computer will take care of problems and limitations that are really design problems and limitations It’s also much easier to change the non-video prototype to test different approaches—more flexibility = better results There’s time to do the long polishing required and to actually “finish” the game March 23, 2016 Learning to design So we can have a playable, testable non- video game much more quickly than a computer game of similar scope or subject Consequently, it’s much easier to learn game design with physical games than with video games! Kevin O’Gorman (American International U) has designed at least six published video games: agrees that non-video games must be used to teach game design March 23, 2016 What happens in Game Design classes? Game production vs. game design We don’t teach game design or programming (or minimally) in 3D modeling We don’t teach 3D modeling or design (or minimally) in programming So why teach graphics and programming in game design classes? March 23, 2016 Game Design becomes “Game Production” class But we do teach programming and art in “game production” classes – For example, groups required to produce: • Playable prototype with 3 levels in three weeks • CD Cover, CD Box Art, 11x17 Poster, and Docs (Design, manual) – 3 weeks is ludicrous for an electronic game, unless it is “castrated” – “levels”—already defines what the game is going to have and not have – But at least they’re making games, not memorizing from a book Of course, expecting groups of students to produce an electronic game in three weeks, or individuals to produce a boardgame in a week (including rules) is ridiculous “You can’t gather nine pregnant women and have a baby in a month”—game design takes time for reflection and experimentation This teaches students that they’re doing something mechanical and simple, rather than creative and complex March 23, 2016 Why are we doing this? Derives in part from fundamental contempt for games: “oh, games are just kid stuff, anyone can design them” Derives in part from ignorance—most people teaching “game design” haven’t done it in any meaningful sense, they’ve been fooled by the textbooks Students think that because they’re expert gamers, they can easily design good games – There’s little connection between the two skills – Many designers aren’t particularly good players, though it’s better if they are (they can find flaws themselves rather than rely on testers) March 23, 2016 Why are we doing this? – “This will be the best game ever”. (Sure) • Testing proves otherwise False notion that game design is a knowledge, not a skill—knowing, not doing – Memorization of information – “Delivery of content” – At one large school, online game design class taught by someone who has not been a game player—ridiculous, how can results be evaluated? – Ask yourself, would this make sense in a musical composition or sculpture course? March 23, 2016 Game design, orchestral composition, creative writing, other creative endeavors This is a creative endeavor, involving complex entities, not simple entities Would you expect people with no experience to produce good compositions, or good fiction, right off? Examples: – John Creasy, over 600 books published, rejected over 700 times to begin career – Jerry Pournelle (SF writer) says “be willing to throw away your first million words” – How many of Beethoven’s early works would be played, if not written by Beethoven? Very few March 23, 2016 Creative endeavors No wonder so many early video games were poor games—lack of experience of the designers! No wonder so few “big budget” games are designed by inexperienced people, and no wonder the big games tend to be designed by committee March 23, 2016 The stages of completion of a non-video game design Idea Notes about idea Detailed notes about idea Rough board/layout of pieces (if any) Detailed board/layout (if any) Prototype (pieces/cards added) Solo-played prototype Prototype played by others Full written rules (rarely done before others have played) "Settled" game Blind testing "Done" (but still subject to change, especially by manufacturer) March 23, 2016 The stages of completion of a video game design Idea Notes about idea Detailed notes about idea Game treatment “Rules” Computer Prototype (usually for show) Playable Prototype (usually new code) Development Testing “Done” March 23, 2016 Design vs. “development” “Development” has two meanings – In video games, it means writing the program – In non-video, development (often by a person other than the designer) sets the finishing touches on a game, but may include significant changes • This can include the bulk of the testing and modification – Development takes longer than design, in either case – Video game classes tend to entirely leave out “development” in the sense of testing and modification to put on “finishing touches” March 23, 2016 The designer’s game vs. the game that’s published Video games are often overseen by the publisher, who is paying the bills; so it is modified to suit as it is developed Non-video games are often unseen by the publisher until “done”; some publishers then modify them, often heavily March 23, 2016 March 23, 2016 My original intention for this talk How do I represent the changes in this short face-to-face format? Many of the changes that occur in a game are not immediately visible, or not visible at all—only noticeable through extended play We’re interested in gameplay, not appearance—functional changes Maps? Cards to pass around March 23, 2016 Example: the progress of a design . . . “Law & Chaos” Design constraint: I wanted a game that primarily used colored glass beads (“stones”)—elegant, visual effect – Likely to be abstract, then—not enough variety for anything “realistic” But how much variety can you get with one kind of piece (even chess has many kinds); how could I provide variety? – Introduce a random but somewhat controllable element – Dice undesirable to publishers nowadays – Why not use cards to change the rules (from Fluxx, CCG) March 23, 2016 “Law & Chaos” What to change? – Victory conditions (pattern of stones needed to win) – Capture methods (pattern needed to take an opposing piece) The game turned out to be best for three players (which is very rare), not something I looked for when I started And there are many, many variations now extant, changing other aspects of the game (the board, the placement/movement of pieces, etc.) Examples (if there’s time) March 23, 2016 Some Web resources These slides at pulsipher.net/teaching1.htm IGDA.org (Game developers) Sloperama.com Gamespot.com, gamewire.com Gamesjournal.com (no longer published, but read the archives) Teachgamedesign.blogspot.com Boardgamegeek.com Boardgamedesign Yahoo Group rec.game.design (limited) Board Game Designers Forum (online) March 23, 2016 Questions? Comments?