A Theory of Fun for Game Design

Despite a growing interest in the field, books on game design can be jargon-filled textbooks too intimidating for the average game player. Raph Koster's A Theory of Fun for Game Design takes an entertaining look at a subject that has, in some ways, been taken too seriously by other authors. The book is thoughtful as well, providing a groundwork for a discussion of games as learning tools, art, and societal shapers. Read on for my thoughts, and some commentary from the author, on this distillation of a designer's viewpoint.

A Theory of Fun for Game Design
author	Raph Koster
pages	244
publisher	Paraglyph Press
rating	9
reviewer	Zonk
ISBN	1932111972
summary	Game design as examined by a skilled craftsman, with a unique look at the larger context of games.

Raph Koster speaks often on the subjects of game design and interactive narratives. A Theory of Fun for Game Design is an approachable version of the larger body of writing and speaking Koster has produced in his years of design work. Its unusual accessibility is clear as soon as you open the book: while the left-hand page page contains text and observations, the right hand page makes (sometimes snide) commentary on design via comics drawn by the author.

Mr. Koster kindly agreed to answer questions when I was preparing this review. When asked about the audience of the book, he said "The book was intended in large part as something I could give to my parents, or to other relatives, or to non-industry friends, as a way to explain what it is that my profession is all about." As such, the comics and plain-spoken writing bring design concepts into focus for readers who may not want to spend the rest of their lives on these topics.

The chapters of Theory of Fun are not organized formally, but the book seems to fall into three sections. The first section sets the stage by discussing what exactly a game is. "Games are puzzles to solve, just like everything else we encounter in life." Koster's thesis is, essentially, that games are learning puzzles. In his experience, simple games are created by children to teach themselves useful skills. More formal games have similar goals, but modern games exist almost entirely to provide the elusive substance of fun to the player. This assertion resulted in a brisk discussion on the site Terra Nova. Exactly what people want when they pick up a joystick is very much in debate even by industry professionals.

The central portion of Koster's theory ruminates on the roles games play, why games are designed the way they are, and what matters in a game. The meat of the book is here, in discussions about why gamers cast aside the ethical quandaries brought up by games like Grand Theft Auto (they're playing the game mechanics, not the fiction surrounding the mechanics) and in the observation that the destiny of all games is to become boring. An amusingly astute statement about cheaters caps off a discussion of the tendencies players have to finding the optimal solution to a game: "When a player cheats in a game, they are choosing a battlefield that is broader in context than the game itself.&quot

At the end of the midsection, the eternal discussion of games as art makes an appearance. Instead of equivocating, Mr. Koster makes his opinion very clear. "Art, to me, is just taking craft seriously. It's about communication (as I have said many times, in the book and elsewhere). Taking what we do seriously, *even if for frivolous ends,* just leads to better work. Considering what you are doing to be art tends to emphasize high standards, experimentation, expression, thoughtfulness, and discipline -- even if your goal is to make a gag-a-day newspaper strip or macrame hangings for your window."

To close his discussion on games and to provide a larger context against which to examine them, Mr. Koster steps outside the bounds of game design and makes some fairly dramatic statements about what games should be. While other media portrays the human condition almost as a matter of course, he argues, games rarely connect with the most basic aspects of our lives. To his mind, in order to truly achieve respect alongside the novel or the musical composition, games should "illuminate aspects of ourselves that we did not fully understand."

In his epilogue, Koster goes even further, arguing that -- as authors of art -- game designers should take responsibility for their creations. "I have little patience for those who hide behind the statement that 'it's just entertainment.' To deny our influence while simultaneously crowing about our financial success is at best naïve, and at worst irresponsible."

The book itself is well laid out, with the thoughtfully edited and often humorous text set amid plenty of whitespace on the right and the usually well-drawn comics on the left. The comics set the tone for the whole book, which in format resembles more of a collection of Far Side strips than it does a technical guide. The back of the book contains an extensive commentary section where offhand references and asides are explained in depth.

If you're planning on entering the field of game design, A Theory of Fun won't help you to storyboard a plot, model a texture, or develop a code base: if you're looking for the technical aspects of game design or deep academic consideration of the field, other titles will hold more for you. The intended audience of this book is quite wide, and Koster does an excellent job of making everyone feel included in the conversation that occurs between the pages. While game players and professionals new to the field alike can get a lot from what he discusses, the reader who may benefit the most from Theory of Fun is the seasoned game industry worker.

With the endless rehashing of game and design concepts currently in circulation and parent groups growing ever more shrill at the release of morally ambiguous titles, Raph Koster's book is a refreshing read. The book is an unpretentious examination of what it is that makes a game a game. He steps beyond the dehumanizing aspects of game mechanics to look at games and their designers in a broader societal context. If for no other reason that that, Theory of Fun is worth a look to read the opinion of someone who gives a damn.

---

You can purchase A Theory of Fun for Game Design from bn.com. Slashdot welcomes readers' book reviews -- to see your own review here, read the book review guidelines, then visit the submission page.

Nice, but not necessary.

[NashCarey]

As a novice game creator I must say that I have yet to read a book and feel I am doing fine thus far. I created http://ruaware.org/ [ruaware.org] AWARE and even had an article published in the NY Times (nothing on /. I am afraid) as a review in theory. The game I created was successful enough to even warrant a sequel.

I specialize in Alternate Reality Gaming and the games are much more cerebral than most, so when Dave Szulborski wrote "This Is Not A Game" (seen at http://www.immersivegaming.com/ [immersivegaming.com]) not many had anything to complain about. Our players tend to like an intellectual challenge.

Yet, I can imagine that many 13-15 year old DOOM players may have their head spin when discussing game theory.

Re:Yeah, but it's Raph 'SWG' Koster

[john_anderson_ii]

I thought SWG was a lot of fun untill it gave way to marketing demand and became Jedi Wars Galaxies.

What makes a game fun

[Radres]

I've done much thinking on the subject, and I contend that there are 4 main elements that lead to a game being fun:

#1) Storyline. This is the most basic element; a computer game can be looked at as a form of interactive movie. However, storyline is not essential since games have elements that movies cannot provide. An example of a game the excels at storyline without the other elements is Star Wars: Knights of the Old Republic. It basically immerses the player in the Star Wars universe without requiring too much in the way of critical thought or reflexes.

#2) Hand-eye coordination. At it's most basic level, a game requires the player to learn how to interact with the environment via some input device, whether it's a mouse, keyboard, joystick, or what have you. An example of a game that does this without the other elements is the original Space Invaders. Not much thought is needed to perform in that game, but learning how to press the fire button and move quickly is important.

#3) Tactics. Forcing the player to make a decision that has both benefits and weaknesses. Forcing players to make real-time decisions in a fantasy world leads to a sense of immersion. It's hard to think of a game that is purely tactical-based, but for an example of what I'm talking about, let's look at Contra. The game takes the basic shooter hand-eye coordination premise that a game like Space Invaders has, and adds the requirement that the user be smart enough to figure out what weapon to use for a given scenario. There are of course better examples, but this particular example gives you the basics of how tactics can be used to enhance a game.

#4) Strategy. Forcing the user to come up with an overall plan for how to do things. An example of a game that excels in this area is Civilization. Provoking critical thought from the user in order to solve a detailed problem (albeit a fictional one) involves the user on a higher level that can be appreciated. I find that the games with the most longevity tend to feature a lot of strategy.

The most successful of games will combine all 4 of these elements. My favorite game is Starcraft, and it is clear to see how all of these elements are used. The storyline is okay, the hand-eye coordination required is immense, the tactics involved are complex, and the strategy level is great. Other games can be broken down similarly. For example, Counter-Strike has no storyline, but there's hand-eye coordination required for aiming the weapon, tactics for deciding what equipment to use, and strategy for deciding how to approach the level with your team.

Think about it, and I bet you'll be hard-pressed to find another way to evaluate gameplay. I only wish there was a game review magazine that took these factors into account!

Re:Raph Koster

[pHatidic]

Koster was also the first to realize the value of "elder games", i.e. the things that keep players into a game after they have already hit the max levels. These include things like collecting rare items (stamp collecting), player housing, guild warfare, becoming a counselor or seer (in UO a counselor is like a minor GM and a seer facilitated role playing, basically player GMs with some limited powers). Anyway I have heard claims from others that he ruined UO in the later years, but in the first couple years at least he was doing a great job. He also communicated very well with players and started the trend of fortnightly player chats with game devs in IRC which no other game to my knowledge has done before

Comment removed

[account_deleted]

Comment removed based on user account deletion

Re:Yeah, but it's Raph 'SWG' Koster

[servognome]

No it wasn't fun from the beginning. Aside from the game's bugginess, there were several core design problems. I would call SWG more of an experiment than a game.
- HAM system - an experimental alternative to the typical HP/Mana systems of most RPGs. Both the penalties of specials (using specials injured you) and the arbitrary nature of damage (rifle damage injured "Mind" not health) just made it overly complicated and unintuitive.
- Player run economy - interesting system, which I think worked well in some respects (gave the "feel" of a real economy). Unfortunately the breakdown occurred because risk/reward system was not in place for adventuring types. If the best stuff was made by players what was the use of taking risks adventuring.
- Housing/building system was nice, though not completely new, it was I think one of the best implementations, though the downside was extreme lag in certain locations
- Skill Structure - bland, and not particularly valuable. Getting higher skills in some respects would give you access to technology that you wouldn't use because there were better lower level alternatives
- Mentorship - interesting, but not particularly valuable, and later became more of an annoyance.
- Entertainers - once again interesting, but not engaging in terms of gameplay.
I think I could have lived with the bugs, in the end I did not like the game due to intentional failures of design decisions. Overall it is something that could be learned from for future game designs. (ie. Discovering that many people wanted to be entertainers, so now how can you make an entertainer class engaging)

Re:Cheating == No Context

[Jace of Fuse!]

One of the best things about medium to difficult games is the satisfaction of defeating them in the end.

In most "Eventuwin" games that are out now days, the average (read, Unskilled) gamer will beat them with sufficient devoted time.

Granted, there are different TYPES of player skills. Logical reasoning, navigation, resource management, memory, hand eye coordination, reflex speed, attention to detail, the ability to multitask, and any combination of the above all are different skills that might be important in different types of games.

On the most basic external levels, all games will have to have some kind of initial sensory appeal, either artistically or contextual.

But after that they must start challenging a person's skills. Some games aren't fun because while they may (or may not) be well presented they just don't play worth a shit.

Having played literally thousands of games in my life (starting with the Atari 2600, and now owning all of the current systems (and only missing a few obscure mid-90's systems), I have played everything from the insanely difficult to the boringly easy.

I can honestly say there is a great satisfaction in accomplishing something exceptionally difficult in a game knowing fwe other people would be able to.

Also, I want to say, while I admit cheat codes built into games can add a new layer of amusement after the game has been defeated, I think it's a horrible shame that some of the greatest games of today are so horribly tainted with built in cheat codes that are widespread before the game even hits store shelves. Very, very, sad.

Re:Cheating == No Context

[syukton]

I guess it depends upon the fundamental outlook of the player.

The best example I can think of right off the top of my head is Trickjumping in any FPS that uses the quake3 engine. Wolfenstein: Enemy Territory is my preferred Q3-using game, so I'll use it for my example. Here's a little background so I don't lose anyone: W:ET is an axis-versus-allies team-based online multiplayer first-person shooter. It is largely objective-based and there are 5 character classes available: soldier (heavy weapons), field ops (hands out ammo / calls in airstrikes), medic (can give out health packs / revive fallen teammates), engineer (can build/destroy things), and covert ops (can take enemy uniforms and throw smoke grenades). The game also has one-way doors that only members of a specific team (or a covert ops in that team's uniform) can open.

There's this map entitled "Siwa Oasis" or just "oasis" for short. On this map, you need to repair the water pump at the oasis and/or run through the tunnels to the old city, capture the old city, and then by either blowing up the old city wall or repairing the old city water pump and draining the tunnel, make your way to the two anti-tank guns and destroy them.

Seems pretty straightforward, right? Well, it's actually possible to JUMP over the wall (without blowing it up and drawing attention to yourself), by exploiting what I guess you could call "nuances" in the physics engine. You see, it's possible to convert downward momentum into forward momentum by "bouncing" off of a curved or slanted surface, much the same way that a ray of light changes direction when hitting a mirror at an angle. Utilizing this method, it is possible to generate enough forward momentum from a position at about the same height as the wall, that you can propel yourself over the wall and be well on your way to blowing up the anti-tank guns before anybody knows what hit them.

It requires a great deal of timing and skill to pull off this jump. I have personally spent hours getting it just right. Is totally circumventing the old city wall cheating? I could have gone through one of those one-way doors I mentioned earlier if I had the assistance of a covert ops and an enemy uniform, and that is certainly within the realm of gameplay mechanics. So, the circumvention of the wall is something that seems to be acceptable (given the door options presented to me), so does the method of wall-circumvention matter? Is jumping over the wall actually cheating?

I totally agree with the statement made, "When a player cheats in a game, they are choosing a battlefield that is broader in context than the game itself." I like this statement because it speaks to my nature. I am the sort of person that likes to approach things from outside the box, and sometimes that means choosing a battlefield that is broader in context than the box itself.

Idiot Testing

[cirby]

There's a bad habit among some game designers. They use friends and "nice" people to playtest their games.

You have to include idiots and assholes in your test sequence. You need to have That Guy - the rules lawyer, the "I didn't mean to do that" fellow, the "I don't understand this" twit. And you need to build your system to shut them out when it's done. For MMORPGs, you need the sort who will get a medium-powered character and hunt down the newbies. You need a complete lunatic for driving games ("why can't I drive across the river here?"). You need a tactical asshole, who will camp on a resurrection point in a shootemup.

(The idea of "idiot testing" was laid out quite nicely by Steve Jackson about 25 years ago, in "Game Design: Theory and Practice"). It was about board games, but the concept holds even more for online games.

Re:Yeah, but it's Raph 'SWG' Koster

[zeno_2]

I bought UO a few weeks after it came out. First online game I played really, and the first few weeks just amazed me. The first few years, I remember them doing a lot of quests involving seers. They were people who worked for them, and role-played a person who would have a big quest ready for players to embark on, for those who haven't played it before.

I remember a big quest that was over the course of a few months, and at the end, I was picked to stand atop this tower at the maze, and hold one of the staves of the elements. When we brought them all together (1 other player and 2 seers if I remember correctly) it spawned a *ton* of elementals all around the maze. Later on they took the staff I held, and put it in the moonglow museum with a tooltip that said, "Staff of Air, Held by Cloud" or something to that effect.

I played on the sonoma server, which was also the home of Oasis, and Fight Night. Fight night was this "event" i guess you could call it, where a bunch of people had duels in the desert (Oasis) on friday nights. This got so popular, that they were able to get the people who ran the game to build arena's and special buildings and such. It was also big fun when the pks would teleport in and try to distrupt things. We'd have the biggest battles.

Those first few years were a blast. I'm not sure if it's because I was new to the whole mmorpg thing, or because it was just an excellent game. When I think about it, it's probably the latter. I quit a bit after they split the world into 2 realms. I did come back a few times though with new accounts, but it just isn't like it was. Since then i've played EQ, DAoC, SWG, AC, and currently WoW, and none of them are as fun as UO was when it first came out, IMO.

Some ideas and whining

[PromANJ]

Sadly, most games today have such weak and inconsistant rulesets that they have too resort to invulnerable characters, impregnable forcefields and linear scripted setups.
Bending the rules is one of the things I enjoy the most in games, but nowdays everything is too controlled in favour of storylines.
It also seems to me that back in the day people/geeks made games because they wanted to and they had an idea, and nowdays it has to be 'safe' and they want to make money, so they make another WW2 shooter with some sneaker elements cuz that seem to be popular too.

Here's some games I'm still wating for:

• Exile was a little confusing and hard but it's just blew my mind. It's strange how a game that runs on a BBC micro with 16/32kb of mem can beat the crap out of ...something new and recent. When will I again see a game where I can just blow the door with a big pile of 'nades up instead of finding the key?
• Paradroid? Great concept, it had the sneaker and LoS paranoia element, lots of replay value with the free roaming and different droids you could take over.
• Lemmings. Is anyone else wondering why they aren't making a lemmings-like game with like... structural integrity and awesome physics?
• Wrecking crew would be pretty neat with physics too wouldn't it?
• SuperCars II? I want to buy missiles and armor again. I'm tired of running over symbols or driving in normal traffic.
  
• Utopia K240... it was like Sim City and a RTS in one. Beats HomeWorld by a parsec if you ask me.
• SimLife... imagine this with some new life algoritms and 3D morphing.
• Star Control II! Imagine that with a dynamic realtime political universe with vast fleets of Umgah and Spathi or whatever going at it.
• Metroid and Zelda never really had any sequels if you just look at how unlinear they were. There's no more walking into Dungeon 8 right from the start anymore. Sigh.
• Blaster Master + Excite bike = my wet dream. It's so awesome to jump out with Jason and just swim or crawl around. Nowdays they don't accept characters smaller than 48 pixels even on GBA...
• Stunt Car Racer felt so real! You were actually in that rollcage when you played that game.
• Dogs of War was really unlinear, but it was balanced so it was just better to pick the easier missions from the start. I really like that kind of responsibility.
• Elite. No elite... no, I don't want to play as "Slater - the ungrommed mercenary looking for his auntie's killer"... I want to play Elite... no, I don't want to play 'missions' where I shoot generic plasma balls on hoardes of pointless enemies... Just give me another Elite Dagnabbit!
• Scorched tanks/earth, again, imagine what could be done with modern processors.

Oh well, maybe I'm just old and nostalgic, but so are many others my age, so it should be a market, no?