Just finished Malcolm Gladwell's Blink: The Power of Thinking Without Thinking- I was leery at first because it sounds like the sort of book that's going to sell because people want to believe they can trust their gut and they're just looking for a book to back them up. And, to be fair, that's how it's marketed – and someone even wrote a reactionary book called Think dissing on it for that.
Also, I wasn't very fond of The Tipping Point – I thought his central thesis is just wrong – there is no point – in epidemics, isn't there just a steadily growing curve? You can't put your finger on a single point and say, “that's where it tipped.” Although, The Tipping Point was filled with interesting factoids which made it entertaining anyhow.
But, I'm happy to report, that Blink isn't really about that at all. Although it is filled with anecdotes about people trusting their guts and coming to correct decisions in small amounts of time, these anecdotes are all about experts who have been working in their fields for many years. So, yes, you can trust your gut, after you've got years of experience under your belt. This should surprise nobody. Kasparov can find the correct move on a chessboard in a second that would take me minutes or hours to analyze.
And I was glad to hear it, because I am a very not-trusting-my-gut sort of guy, even though I have fifteen years of experience making videogames. I'm always second guessing myself. That's going to change. Now I'll just second guess myself a little.
Also, a lot of Blink is about the times when your gut can fail you. In fact, it seemed like more of the book was devoted to how split-second decisions can go wrong.
One interesting kind of split-second decision that can go wrong is the Sip Test. Coke switched to New Coke in the late eighties because Pepsi supposedly was winning the flavor wars. They were winning on the basis of sip tests – you did a blind sip of beverage A, a blind sip of beverage B, and which did you like better? Pepsi, because it was sweeter and had a burst of citrus.
So Coke changed their formula – earned the hatred of millions – and then changed back and gained more market share than before, managing to blunder into a happy ending.
Al Ries claims that Pepsi tastes better and it's all about marketing, but one of the marketers at Coke says the problem is sip tests suck and to really rate a product you need to give them a home use test: you give them two cases of the two beverages and ask which they like better. (I suppose you better make sure the cases are just generically labeled “Product A” and “Product B”...I wonder if they do that?) Because the sweetness of Pepsi gets cloying once you've had more than a few sips, and that hint of citrus you detect with your first sip is lost on the second.
How does that apply to videogames?
It's conventional wisdom that a videogame has to grab you in the first 30 seconds or it's doomed. I'd like to challenge that conventional wisdom – look at Tony Hawk and Spider-Man 2. If you can fail a kleenex test—the videogame industry equivalent of the 'sip test'--these games did. It takes 15-30 minutes of playing before you start to feel even comfortable with these games, and it takes longer before you get good. And yet they're both multi-million sellers, outselling a couple of games that no doubt aced their kleenex tests – God of War and Ninja Gaiden – games where you feel good in the first thirty seconds, mashing buttons and explosively killing guys.
Apples and oranges, I know – multiplatform vs. one platform, franchise title, different marketing, etcetra. But, as NASA research shows, apples and oranges are really quite similar.
Also, this conventional wisdom seems to assume that players are playing demos when considering whether to buy, and the demo that catches their interest is the one they pick up. Yes, that happens sometimes, but I think far more often they pick games because of the box, or because of word of mouth, or because they played it a friend's house. (It's not reviews, unfortunately...) In which case they get more than a sip before they make their buying decision. Or they rent it, in which case they probably play at least a level before returning it, hopefully getting in that fifteen minutes.
What I'd like to see is videogame gameplay testing move closer to the world of home-use testing. Since we can't send devkits home with our testers, maybe the solution is to create a comfortable, living room like atmosphere for the testing – sofas and large-screen TVs – minifridges with soda and food. The testers are encouraged to play as long as they want – say, up to 10 hours if they want to – but they can also take breaks, whenever they want, and quit and go home whenever they want. The researchers stay out of the way – I'm wont to hover over a tester's shoulder when watching them play, I'd need to force myself to step back behind the mirror. Although sometimes you're doing gameplay testing – where does it get frustrating? -- in which case they're only given the one game, other times you're doing market testing, and then you'd give them a choice of games to play, only one of which is the game you're really interested in. (Although a publisher with multiple titles could put them all in the hopper to see which one the players like best and therefore is most deserving of marketing dollars.) At the end you'd ask them how they liked it, and which parts they liked best, and it wouldn't be until they'd rated it that you'd ask them to get more analytical. Because, as Blink mentioned, getting analytical about your opinions changes them.