« Rules and Borders Part 3 | Main | Once Upon a Tank and a Time »

Feb 11, 2005



Do you just want to figure out "fun" or do you want to figure out "fun" in the context of "play"? The concept of flow and Raph's book (haven't read it yet -- sorry Raph!) seem targeted at fun as a form of play -- and if you want to get into play at all, I think Brian Sutton-Smith's work is really informative.

But I'm as interested as you are in getting a good reading list...


I guess I am most interested in fun itself, and from a more scientific rather than philosophical perspective. it would be nice if there was a generic theory of fun that quickly explained why buying lotto tickets, playing WoW, flying a jet are all fun. and if you say 'flow', well, what is flow doing to me? give me an evolutionary theory why flow makes me happy.


No matter how much money the average person saves, they're never going to have enough to live out their dreams. For $1 a day they can imagine it's possible.

An economist might say it's irrational, but compared to paying $100 for a psychologist to tell you to be happy with what you've got, it's very logical.


The attraction to play, in pretending to be something that you aren't, used to be an evolutionary advantage to prepare for adulthood.

You know the games: girls played nurturer with dolls and such, and the boys ran around hiding and play-fighting each other. Perhaps we haven't un-evolved the hardcoding in our brains yet.

But the attraction to the lottery is a mystery to me, too. Perhaps in instead of being linked to an evolutionary advantage, it is a personality disorder. However, I saw a show the other evening on the Discovery channel I believe, that explained how risk-takers played a role in evolution. The folks who would dare to climb the cliff to find the better cave, and so on. I guess evolution has a place for high-risk/big-reward behaviors.


Evolutionary explanations?

- flow is easy: productivity

- lottery is reasonably easy: Humans aren't good with probability. This is an advantage: if we were we would never have started inventing things or started up new businesses etc. Humans are pretty good at acting on inaccurate and incomplete information as well as following impossible dreams... Some good stuff comes out of this (progress and Art).


"Flow" is not "fun." You can get in flow from chopping wood, or doing breathing exercises. It is not the same, physically, as "fun," based on my reading. Flow is characterized by a certain alpha wave pattern, and fun is characterized by a particular endorphin surge.

My recommended reading list if you're going to jump on this subject is mostly here:


Sutton-Smith isn't on there, but should be.


"Fun" is time spent transporting your mind temporarily into a state of denial of the predictability of life. Your actions are temporarily guided by its own internal set of logical parameters, though usually the boundaries are still set by deeper moral values and knowledge of causality.

In the case of buying lottery tickets, you are letting the dream of fantasy life outweigh your ability to measure risk and probability rationally, and in the case of your average short term fun (movies, tv shows) you can experience a brief escape from the responsibilities of everyday life.

However, in the case of virtual worlds, you replace your inherent real world moral relativism and vision with an environment specific subsitute ones, with the freedom of a blank slate future. Although the ties with real life remain, you are allowing yourself a parallell destiny with its own set of value propositions and decision making criteria. And although this is usually nested within the value choice of paying for access to the world in the first place, the payoff is often so great that this fact is completely overshadowed. In some ways, the closest parallell is that of drugs, without the legal and medical side effects.


Bruce - I agree with you that fantasy is part of the appeal of the lottery, I would also think that part of the pleasure also finds its source in what Caillois calls "alea" (games of chance).

Ted: "When we buy a carrot, say, the value comes from the carrot alone, not the buying of it; the buying of a carrot is not in itself fun."

It would seem to me that the act of consumption in itself can be a significant source of entertainment/pleasure. I would imagine that economists/psychologists/marketers must have some theories on what makes shopping fun/addictive, etc.


Ted>What else?

Well, there's the theory of fun for virtual worlds that I outline in my book. Fun is whatever people are doing right now that will serve to advance their understanding of themselves. Er, I go into it in a little more detail than that in the book, obviously...



"Posted by Edward Castronova on February 12, 2005"

So where are you in the world, Ted, that it's tomorrow already?



One angle that I guess you need to take into account is the whole grind thing and why games can look very like work but are still described as ‘fun’ by the players. One thing here I think is to dissect different types of pleasure – as I suppose one way to understand fun is as a sub category of pleasure. Also I think you need to look at stuff like Espen Aarseth’s work, the key one I think is:

Aarseth, E. J. (1999). Aporia and Epiphany in Doom and The Speaking Clock: The Temporality of Ergodic Art. Cyberspace Textuality: Computer Technology and Literary Theory. M.-L. Ryan. Indiana, Indiana University Press: 31-41.

If you want to take a detour into psychology, Winnicott’s Playing and Reality is an interesting starting point, though I find the key parts on play to be quite frustrating and am looking for some secondary lit on this myself.

Oh and I'd add another vote for Sutton-Smith, anyone vaguely into these areas should take a look at that.


How do you consider buying lotto tickets to be fun? I can certainly see the purchase as a defiance of the known odds, in which case that's just a change in what utility is being maximized. If "having fun", whatever that might be, is considered to be of higher utility than alternative options (such as having more money), then it seems quite rational to me.

Nevertheless, I liked Raph's proposition quite a bit. It ties very well into the recommendation of "make learning fun" given to teachers of all varieties.

The challenge, if you will, of being presented with something new seems to be a key ingredient of fun-ness. Replaying something seems more nostalgia for the old fun than really fun proper. Bayesian probability judgment? And the demand of fun to be immediate, rather than long-term, smacks of hyperbolic discounting.

My two cents.


It's impressive how Ted managed to post this in the future so that it stays at the top :-)!


There's actually a very real reason to play lotteries and other games of chance -- the utility of the money.

For me, £1 has almost no real utility -- I am, in real terms, no better off for saving £1 each week. In fact, I could probably lose £10 each month without a significant impact.

By contrast, the utility of an £8million is quite enormous. It's trading off these two utilities that gives a lottery an actual point.

Of course, on top of that is the excitement, self-delusion, and indeed an element of fun. But there is an economic motivator -- sort of an inverse insurance.


Sorry, that should have been "£8million win".

Here are a couple of links, too:



Although it was not implied or questioned, I should probably clarify, given the mention of his name in this thread, that I am not Espen Aarseth, even though I probably do come from the same country and share the same first name. :) Any other similiarties are probably minimal.


The whole lottery ticket analysis has always bugged me for two reasons:

1) It is always assumed (sometimes condescendingly) that purchasers of lottery tickets are homogenous in their motivation for buying the ticket. (i.e. get rich quick with a total ignorance of the odds of that happening.)

2) I've never seen anyone discuss opportunity cost in the calculation of buying the ticket.

In terms of motivation I think it's an error to assume that lottery purchases are made as a result of a fundamental ignorance of statistics. It would be interesting to know if anyone had investigated lottery purchasers' awareness of their actual odds of winning. (Surely someone's done this.) My opinion is that most lottery purchasers understand it's a _long_ shot. So then the question is: "Why buy into such risk?"

While I think the ostensible reasons for buying a lottery ticket are going to be varied in the end I think that each purchaser has at least made an unconcious evaluation of the opportunity cost of the purchase versus the value (fun, hope, whatever) that they expect to receive. Is it irrational for someone to value a lottery ticket more than a bottle of Coke or a couple of candy bars?

In any case, I'm not sure that looking at the lottery as part of the quest for a Grand Unified Theory of Fun (GUTF ?:)) is really helpful unless you focus only on the lottery players that are actually playing for fun. (How you'd sort them out, I haven't a clue) Otherwise I think you're likely to get side tracked by the segments of the lottery population (and I think they exist) that play even though it's _not_ fun for them.


When considering the rationality of gambling, people often forget that these games really do payout much of the time. Consider roulette: according to Wikipedia, the American house average for single-number bets (and even/odd bets, according to my calculations) is 5.26%. That means that players keep, on average, 94.74% of each wager. If some player bets a total $1000 in a night of gambling, they need only receive $52.60 worth of entertainment for this to be a ‘rational’ act.


I've written a brief description of a psycho-social framework for describing fun and hypotheses related to it (IMO we're nowhere close to having a theory of fun) in an upcoming book ("Playing Video Games: Motives, Responses, and Consequences", Vorderer & Bryant, Eds., in press). The core of this idea is that there's not a single thing we can call "fun," but that fun is the result of a collection of psychological (or in some cases psycho-social) interactive responses. Since fun is inherently psychological (it's not something that exists independent of any of us), looking to our physiology and psychology for caustive factors seems reasonable.

I've separated this into several levels:
- Physical/perceptual/emotional: influenced by evolutionary responses, e.g., for motion, bright color, etc.; and also by our seemingly inherent (present in neonates) attraction to certain kinds of body motion (proprioreception). When graphics or fireworks dazzle and give an "oooh!" kind of experience, that's perceptual "fun."
- Short- and long-term cognitive: tactics and strategy in military terms; short-term puzzles and long-term goals. Myst was fun for many people because of the combination of being perceptually lush and its presentation of short-term cognitive puzzles. Other games (RTS games, e.g.) rely on stimulating interactive sequences of short- and long-term planning in rewarding ways. (And lotto tickets, FWIW, fit in here too - although they in some sense prey on the too-human inability to predict odds accurately, or to reason forward well about correrlation and causation).
- Social: we are (evolutionarily, culturally) incented to seek out others like us. When we can experience perceptual, emotional, or cognitive "fun" in the presence of others, those experiences become more meaningful to us.

Asking why this is -- or why we're attracted to bright colors, puzzles with appropriate affordances, etc., -- opens up a longer discussion of both evolutionary psychology and human performance measurement. This latter also brings us back to "flow," which seems more to be an indicator of a homeostatic balance of factors that often (but not always) engender a feeling of fun. Some fun is not flow at all -- the fun had on a thrill ride, for example -- and some flow-related activities are not typically considered fun (chopping wood was mentioned above).

But underlying the presence of flow is a balance between arousal and ability (going back to the classic Yerkes-Dodson curve from the 1920s): with too little arousal, performance suffers and boredom results; too much arousal (e.g. too many targets, too much input) and paralysis or panic results. As ability and familiarity increases, the ability to respond to greater amounts of arousal also increases -- this ascending arousal curve is perhaps the most common trope used in modern computer games. When the balance between arousal and ability for an individual is maintained over even a short period of time, and when the context lends to perceptual, emotional, cognitive, or social interaction, "fun" is often the experiential result.


What about fun that is related to "dissonance", where would you put it? (i.e. things that look out of place, comedy etc) Social?

One also have to consider situations that are both fun (you laugh) and unfun (you want it to stop, or maybe not) at the same time: teasing, tickling, heavy flirting... I wonder how that type of fun fits into the various "theories".

When something can be both fun and not fun, and we both want it to continue and to stop, at the same time then it clearly is a rather complex area to create theories for.


Ola, I'd venture that physical comedy, pratfalls, or other surprising or "out of place" comedy elements are probably a combination of perceptual and social: we react first to the unexpected visual or auditory sequence, and then apply a social overlay that informs us whether the events are comedic or tragic (the line between these being thin, blurred, and having been debated for millennia).

You're absolutely right though: the human sense of fun, like the sense of humor, is a subtle and complex thing. What I've outlined here is the merest bit of my current thinking on this; the chapter I wrote says more (and even that is at most a glimpse).

One interesting tidbit from our AI work: in many respects it's easier to identify conditions under which an AI agent could be said to be having fun (or close analogues of various forms of pleasure, delight, even joy), than it is to simulate any degree of humor (geek reference: "humor - it is a difficult concept" ;-) ).


Jeremy Neal Kelly>That means that players keep, on average, 94.74% of each wager

When I was at school, my weekend job was in an amusement arcade. We had slot machines in there, referred to as 70s, 80s and 90s. This was your return on investment, ie. if you put in 100 coins of one sort then at the end you'd expect to have 80 such coins in payout. Most were 80s, some were 90s and I think the 70s may have been illegal...

The normal arrangement was a bank of similar machines, one of which was a 90 and the rest of which were 80s. Someone would go on the 90, people watching would see it was a "payer", and they'd go on the other machines "of the same type". Those were 80s, so the arcade would make more money from them, with the sole 90 acting as a loss leader.

Although most of the machines in my day were honest, in that if you pulled the arm then there was no way the machine could decide where it was going to stop, there were some that weren't. They had a switch at the back, marked "liberal" and "conservative", so you could change the odds without changing the reels. Today's machines are far more cynical than that; I've no doubt that owners of large casinos can decide when they want the million dollar jackpot to be won...



I don't know much about humor, but that sounds spot-on. I suspect that most humor involves interpreting something as if it is depicting some aspect of the "human condition". The fun of some of the games I find enjoyable rely on turning society into a silly ant-like shape: GTA, Worms, Lemmings... From a design perspective: I wonder what qualities you need to get repeatable humor. Most humor is one-shot. What I love about Worms in two-player mode is that the humor is repeatable.

I also think it is interesting that designers write about how they conceptualise fun. I appreciate if they don't get too academic about it, which would make it less interesting even from a research POV (design researchers want to peek inside the brain of designers). I believe the optimal value from such efforts would come from designers honestly describing how they think about it and how they find that useful in relation to design. Maybe design oriented models will have to be different from cognitive models too.

One thing that has always bugged me is how difficult it is to judge a design without users. What is the most traumatising situation for one person can be the most exciting situation for another person: e.g. a plane that has to make an emergency landing. Arousal can go either way. I've noticed the same in game-play (e.g. XP-loss).


Richard Bartle wrote:

> The normal arrangement was a bank of similar machines, one of
> which was a 90 and the rest of which were 80s.

Interesting. This strategy seems to rely on the rationality of players, while simultaneously confounding it with asymmetric information. I’ll bet there is some interesting literature from the gambling industry on techniques like this. Suggestions, anyone?

> When I was at school, my weekend job was in an amusement arcade.
> We had slot machines in there...

That sounds like an excellent apprenticeship for a scholar of games! All I did was shuffle test tubes about.


Seymour Papert describes a concept of gratification that results from activities that can be frustrating, difficult and distinctly not fun (at least for short periods of time) as 'hard fun'. The term was coined by a kid trying to describe his experience of feeling enormously challenged and exhilarated all at once.

From Papert: "I have no doubt that this kid called the work fun because it was hard rather than in spite of being hard."

I find the term 'fun' limiting when it comes to games... the enjoyment and compulsion are so much more complex and meaningful than simple fun. I find it fascinating how quickly a situation (like a crazy MMOG battle that pushes everyone right to the limits of their abilities) that teeters on the edge of fun and frustration can so quickly fall either way, flow disrupted, and the depths of despair can be so low. Likewise, when everything works out, the highs can be so high. It's an interesting psychological condition, related to what Ola was describing: 'When something can be both fun and not fun, and we both want it to continue and to stop'.


Edward Castronova wrote:

When we buy a carrot, say, the value comes from the carrot alone, not the buying of it; the buying of a carrot is not in itself fun.

This seems to be the root of the error. Buying is, in and of itself, fun. Evolutionary psychology can explain part of why it's fun (or why your brain rewards you for collecting). There are many games based on buying and selling, trading, collecting, betting/speculating, and other similar behaviors. I think the actions themselves are fun, even outside the context of a game.

Raph said:

"Flow" is not "fun." You can get in flow from chopping wood, or doing breathing exercises. It is not the same, physically, as "fun," based on my reading. Flow is characterized by a certain alpha wave pattern, and fun is characterized by a particular endorphin surge.

I think flow is a kind of fun. Action ("twitch") games are fun at least partially because of flow. It is a particular type of mental state, and it's not a difficult one to reach. According to a primitive, self-made EEG, it's strongest around 8 or 9 Hz (low alpha/high theta). Someone's probably done serious (and more accurate) research on the subject, but google was not helpful.

I'd expect my brain to be fully engaged when playing Tetris, but it is actually running slower than usual. Usually my brain is strong in beta and delta, the highest and lowest frequencies, which makes my primitive LED readout look like an hourglass (unless it's noise or even a bad ground--I cannot stress enough the primitive nature of the device). When in flow, it is the opposite with all the activity in the middle ranges instead of the extremes.

Flow seems to occur (at least for me) when a task fully engages your consciousness, but it not especially taxing. Games can do this, and the feeling seems fun to me. But this is not the only kind of fun, and it is not usually the kind of fun game designers talk about. I don't think it's one game designers, particularly action game designers, can ignore.

There's alot of research on gambling (mainly from psychology rather than economics) that could be useful for game designers. Although I don't think it's directly related to fun, humans are poor judges of risk, time, trust, etc. And humans see patterns where there are none. Game designers can take advantage of this (and many have). Even a http://www.amazon.com/exec/obidos/tg/detail/-/0394722566/qid=1108330130/sr=8-1/ref=pd_ka_1/103-9309309-6591827?v=glance&s=books&n=507846>simplified communications theory book can have useful insights.


Its been incredibly interesting reading the responses on this thread but there's one thing that irks me throughout: the arbitrary signifier "fun" is being attributed to such a range of widely diverging situations and feelings that although they might be associated with a positive feeling are derived from very different sources.

Fencing can be fun. I bust my arse four times a week at it and sometimes, just sometimes flow comes into it and its incredibly "fun". Painting is a lot of fun, for a number of completely different reasons, as are well run RPGs to name a few random examples.

The problem with addressing the puzzle of puzzles is that what is being discussed is not an overall phenomenon, but a variety of related and non-related that are being bunged under the one term "fun".


Jeremy Neal Kelly>That sounds like an excellent apprenticeship for a scholar of games!

I spent most of my time calling Bingo, which certainly removed any fears I might have had about public speaking... It's quite a boring activity, even though there's a lot of interaction with the public. In the UK, Bingo is purely numbers (none of this "B 2" stuff), so after a couple of hours at the microphone I'd do things like add up the numbers and tell everyone what the total was (and sometimes, if the numbers were easy, what the average was).



AFFA, to be most specific, my research said that "flow" was specifically characterized by the release of dopamine, causing periods of "high attention." This was different from what we usually consider periods of pleasurable activity, e.g. fun, which were characterized by the release of endogenous morphines.

The two often go hand in hand, of course; Nicole Lazzaro in her "Why We Play Games" ended up calling fun characterized by flow as "hard fun," one of four types.


I don't know, you could probably classify meditation as "flow"... Flow is a rather high-level concept as I understand it.


At the risk of sounding all woo-woo and mystical, many people play the lottery for the chance to create a liminal state. They _might_ become a millionaire... and what the dollar pays towards is creating the state of "mightness", that millionaire-potentiality. Whether it pays off is a separate transaction, usually one which is treated with a great degree of stoicism even when money is won.


Most of what I'm thinking has already been said, though not together and in the way I'd like to put it here.

First the lottery question, since it seems easiest to me. I would say that the idea that people don't do well with interpreting odds is right-on. It isn't that people are typically incapable of understanding them either. It's just that we have a hard time interpreting the more large-ish numbers involved in a lottery into a real life setting.

Try this as an example. Can you mentally picture five hundred trillion dougnuts? I certainly doubt it. It's a very unrealistic idea. The odds of a lottery are the same. We can't really picture the improbability of winning and understand what it means in real life.

The same goes with the winnings, actually. The difference is that while we might not be able to understand eight million units of currency, we can definitely picture that car, house, jet, and private island we'll buy (exaggeration included here).

So the logic that comes out of this is a kind of "reality" logic, flawed by our limitations. The odds are very long in a lottery, but the winnings are very large, so why not? It balances out unless you can get a concept of the actual probability into your head. On risk of being hated for this, I have to say that I beleive intelligence plays a part in this grasping of the odds.

As for fun, I don't think the point has been stressed enough that fun isn't any one thing. I'm not qualified to talk any more scientifically about it, but maybe I can help get to the root of the question.

Fun is an emotional sensation, as I see it. It can be feeling accomplishment, intellectual stimulation, physical/sexual stimulation, release of mental or physical stress, and humor. There may be others of which I haven't thought.

What do these all have in common? They are emotional sensations that we have socially labeled "good". Heartbreak definitely isn't "fun". So, fun as I can see it is a positive emotional reaction, especially within the context of entertainment. Stimulating fun in people would then be as simple as provoking positive emotional sensations in people, preferably in a variety to keep from "exhausting" one particular sensation.


A couple of points (now that I'm back from some minor surgery):

1. What's "Fun" To You?

Considering "fun" in the larger context of "pleasurable sensations," I'm with Mike Sellers -- most of the more advanced mammals seem to have an innate capacity for fun, but individual humans don't all perceive the same activities as equally fun. Specifically, I think there are innate psychological preferences for perceving different kinds of sensations as "fun."

Consider roller coasters. Some people (I would say certain *types* of people) love the kick to the adrenal glands. Others can't see the attraction at all. Slasher movies ("boo!") fall into this category, too.

Now consider verbal wordplay. There are some people who can spend hours in punning contests or quoting Monty Python, and who would consider an evening spent this way to be the most fun they've had in months. But other people would probably rather have a root canal than be forced to endure such "fun." (Actually, they'd probably rather be riding a roller coaster.)

So I'd say that any GUTF must somehow take such differences into account. If the theory doesn't adequately capture real preferences for different kinds of fun, then its utility as a guide to creating enjoyable experiences will be limited.

2. Fun in a Game Context

There's an interesting suggestion by James P. Carse that there are ultimately two types of game (and two types of gamer). He calls these "finite and infinite games."

A finite game is played to be won or lost. A finite gamer is one who needs a game to conclude to feel satisfied.

An infinite game is played to continue playing, and an infinite gamer is thus one whose gameplay goal is to make moves intended to allow the game to continue indefinitely.

In his book _Finite and Infinite Games_, Carse (IMO) reads a bit too much into these two styles. To Carse, the finite gamer is always competitive rather than cooperative, and is resource-greedy due to a zero-sum mentality. Meanwhile the infinite gamer (whose gameplay is apparently illuminated by the light of the halo over his head) is primarily cooperative and strives for the expansion of resources rather than consumption.

I think this is simplistic. Competition isn't always a Bad Thing; it's an effective means for encouraging the efficient use of finite resources. That makes it an interesting process to simulate in a game -- i.e., some people find competition "fun" because it requires a certain kind of efficiency-oriented thinking.

Having said this, I think Carse may still be making a worthwhile point: Why make the assumption when designing games that everyone is only motivated by "beating" other players? In other words, would it be worthwhile to design a game whose structure flows from assuming that players would enjoy an infinite game?

Should there be a game whose meta-goal (as supported by the rules of the game) would be the infinite continuation of the game for all of its players?

Are some MMOGs already such a game? If so, is that part of the reason they're considered to be "fun" by some players?


The comments to this entry are closed.