Success: what great people actually share

— Name any trait.

— Any one of the ones you counted up. One.

— I'll bring you a thousand people who kept at it for 20 years and stayed nobody. You've never heard of them for exactly the reason that persistence didn't help them. Next one.

— Same thing, just a bigger graveyard. Look, I'm not making fun of you. I want you to notice that this trick works with any trait you name — and that what it means isn't "traits don't exist," it's something else. Arguing about a trait is pointless until somebody asks who assembled the list of people we're looking for those traits in.

— The outcome. The outcome assembled the list, and the historians showed up afterward to describe what it handed them. Say you want to find out which shoes hold best on ice. You go into the office and survey everyone who made it in. It turns out 90% of the ones who made it were wearing flat soles. The conclusion is obvious: flat soles save you.

— And the ones who slipped and ended up in the emergency room, you didn't survey. And they were wearing exactly the same shoes. You weren't studying footwear. You were studying the office lobby, which only lets in the people who made it.

— The idea is Wald's. The scene isn't.

— Wald wrote 8 memoranda in 1943, at the Statistical Research Group at Columbia University. Solid mathematics: he derives the probability that a hit on a given part of the machine brings the plane down. Nowhere does he say where to put the armor. The plane covered in red dots that you've seen was drawn by the designer Cameron Moll around 2005, and the Wikipedia image was made from his diagram.

— Survived, because it's a good-looking picture. Yes. For a long time that struck me as a funny coincidence, and then it stopped being funny. Notice what got selected: not 8 memoranda full of formulas, but a scene in a room where a clever man says a brilliant line. Selection is at work inside the story about selection too.

— They were real. I'm not arguing with that. But let's test it on a case where everything was measured. In 1982 a book came out about excellent companies — 43 of them, taken apart bolt by bolt, interviews, numbers, all of it honest. 2 years later BusinessWeek ran a cover story titled "Oops! Who's Excellent Now?"

— About a third of the exemplars were already in serious financial trouble. Meanwhile the book went on selling in the millions. Notice what the market actually tested: not the conclusions, but how good the book was to read.

— Take a gentler one. In 2001 a book came out about 11 companies that had made the jump from good to great: 15 years of ordinary results, then 15 years of beating the market several times over. Circuit City was one of them. In November 2008 it filed for bankruptcy, in 2009 it announced liquidation, 34,000 people lost their jobs.

— That's the most interesting part right there. The traits they were picked for didn't go anywhere. The discipline, the right people on the bus, everything just as the book says. The world simply shifted, and the same behavior started producing a different result. There's nothing to blame them for — and there was never anything to praise them for either.

— They have, and that's the best part of the story. In 2009 someone took the companies from 13 of the major books about success and checked statistically whether their results differ from what chance alone would produce. About three quarters — don't.

— Then the same authors did the thing nobody does, because it doesn't sell. They took not the winners but everyone: more than 25,000 companies traded in the US between 1966 and 2010. And they measured who actually breaks out beyond chance. It came to 344 companies.

— Out of 25,000.

— No. And this is where you have to be careful, because this is where everyone falls. From the fact that we're looking wrong, it doesn't follow that there's nothing to look at. Those are two different claims, and I didn't make the second one.

— That almost everything we call the traits of great people is a property of our optics, not of their psychology. Optics can be fixed. What's left after the repair is the whole rest of this conversation, and I'll warn you honestly: less will be left than you were counting on, but it won't be zero.

— I didn't like it either. I spent 8 years getting to that thought, and I ended up putting it roughly like this: the question "what do great people have in common" already contains its own answer, because it contains a sample. After that you can argue about traits as long as you like. First look at who put the list together.