Bus Report: Stumbling on Happiness; A Theory of Fun

May 01, 2007 21:40

"Stumbling on Happiness" (by Daniel Gilbert) is a non-fiction book that is apparently very popular right now, as I was something like 139th in line for it in the queue at the library, and by the time I got it, I didn't remember what prompted me to put it on the list in the first place.

It's a very interesting book! It is largely about why we tend to be bad about doing things that will actually make us happy. We over-project our current mood onto our perception of our past and future moods, and we forget (both when remembering and projecting into the future) that our mood is made up of a variety of factors, only a few of which we may be thinking of.

An obvious example of projecting is shopping when you're hungry--you project your hunger into the future, and overbuy. A more serious example is that when you're depressed, you overestimate how depressed you were in the past, and overestimate how depressed you might be in the future.

An example of the undercalculating effect is people's predictions of how happy they'd be if they lived somewhere else. We take a few factors we think are important and estimate our mood based on only those things (weather, proximity to X), but we leave out a lot. Many people from all over the US believe that they would be happier if they lived in California or (if they already do), that they are happier because they live in California. But it turns out that people's happiness levels are actually not significantly affected by whether they live in California or not. It's one factor that seems like it should make a big difference, but turns out to be overwhelmed by other factors in people's lives. One of my friends, reading this book, realized that he had been sort of sad because he always thought Vancouver would be a great place to live, but that he would never be able to live there. The book helped him realize that the good times he associated with Vancouver were the times he had been there on vacation, and that if he lived there, there'd be good things and annoying things just like everywhere else. So he started to rearrange his life so that he had more vacations to spend in Vancouver.

Another similar example was illustrated by survey results before, just after, and well after the 2004 election. People were asked how they would feel if Bush won, then later how they did feel now that Bush had won, and much later how they had felt when they learned that Bush had won. Both the before and after predictions and recollections were much more extreme than how people actually felt when in the moment itself.

This was where I took issue with the book's conclusions, at least a little bit. My feeling is that our feelings about important events are actually better worked out in hindsight than in the moment. To some extent this is definitional: Daniel Gilbert believes that 'in the moment' feelings are the most important, while in this case, I believe that after-the-fact feelings are the most important. So the fact that people's predictions matched their recollections is satisfying for me, but the fact that people's predictions and recollections did not match the in-the-moment feelings is worrying to folks like Daniel. For me, though, it makes more sense to focus on the recollection, as that's the emotion that stays with you longer. And, since in the moment of the event, one's emotions about one thing are tangled in with one's emotions about everything else that's happening at the time, it makes a certain amount of sense to me that we would be able to 'untangle' those emotions in our memory by forgetting the extraneous emotions, and focusing on the emotion of the event on a more one-to-one scale.

Though the book is more focused on theory than advice, he does have a couple suggestions. One is that if you are wondering what it would be like in situation X, you should find people in situation X and ask them how they feel about it. Trying to project yourself into a hypothetical situation X is not something our brains do very well on their own, and asking people has much better odds.

He also recommends taking chances, at least as far as happiness goes. On the whole, people are happier about failed attempts than they are about missed opportunities. This is an area where projecting into the future and remembering the past are at odds: we predict we will be happier taking the safer option, but we remember being happier taking the risk.

Another tidbit that I picked up through implication, though he didn't outright state it: burn your bridges. Once you're committed to doing something one way and not another, your brain will naturally elide its negative aspects and highlight its positive aspects. This has obvious implications for relationships, too: if divorce is a viable option, you will be less happy in your marriage than you would be if it were not.

This is all conveyed in a breezy, lightweight style where he tends to make his points through repetition, and by making a lot of jokes along the way. But he references a lot of actual-factual experiments, and makes claims that you wouldn't necessarily have thought of, but make a lot of sense once you think about them, so that's good. At times I ended up skimming a bit, but overall, I enjoyed it.

One final caveat: if you are happy, I would not necessarily recommend this book, because (as he says therein), analyzing your happiness will make you less happy. Conversely, if you are unhappy, I would highly recommend this book because analyzing unhappiness makes you more happy. OTOH, I'm a pretty happy guy, overall, and managed to make it through the book unscathed, so you're probably OK either way.

"A Theory of Fun" (by Raph Koster) is, I felt, misnamed. It is A Hypothesis of Fun, with little to no presented evidence backing up any of his myriad of claims. Maybe I'm biased by the other nonfiction books I've been reading lately (like 'Happiness') or by my job in the sciences, but while Raph is no doubt an intelligent guy, and while he does indeed have a lot of experience in the field of game design, I was constantly disappointed by how many of his arguments were just him saying, "This is how it is." He said many reasonable things, but it was mostly just him rambling. I got the sense that if I tried following any of his tenets, I would be performing an experiment myself instead of applying any already-acquired knowledge.

I felt he also got himself into trouble whenever he would talk about what games 'should' do or 'should' be. The arguments he would surround these statements with were clearly intended to support these opinions, but that's all they were--opinions. There was never any exploration of the effect of doing things his way vs. the current way, or why one result might be better than the other. "You should stop playing a game when you've learned what it has to teach you." "Game designers should try to make art, and not just pander to what's popular." Well, you know, that's nice and all, but neither of them are backed up by what he imagines your goal to be, nor how doing what he recommends would move you closer to that goal. Very shallow.

If I was a social sciences guy instead of a physical sciences guy, I would jump into this field in a heartbeat, because there's so much here to be explored, tested, and discovered. Raph is pointing at interesting things people might want to explore, test, and discover, but sadly hasn't explored, tested, or discovered anything himself. Or if he has, he didn't tell us about it in this book, which is a shame.

Oh, and there is artwork. It's Raph's. It's not very good. I am not the world's most visually-oriented person, but even I found it distractingly bad at times. It is clearly intended to be endearing because he did it himself, and it's certainly better than what I could do, but I was not endeared.

book review

Previous post Next post
Up