ED vs. COGNITIVE BIASES, PART 2: SURVIVOR'S BIAS

Stop! This isn't a re-post from Wednesday. It's new. I swear. I'm doing them back-to-back because the names are so similar. But other than the root noun the two ideas have little in common.

Historical analysis often engages in lengthy searches for explanations of favorable outcomes. Take the Cold War as an example. There is no shortage of explanations for why World War III didn't happen and why we didn't all die in a nuclear holocaust a thousand times over. Take your pick – mutually assured destruction, skillful diplomacy, Ronald Reagan's mighty persona, the power of vast international alliances, credible first-strike, credible second-strike, and so on. Maybe, although we can't say with certainty, we just got really fucking lucky. Maybe we all should have been roasted to a crisp. Maybe there was a 99% chance of nuclear war breaking out under the given conditions and we were just very fortunate to have the other 1% bear out in reality. Other than Robert McNamara, I haven't heard too many thinkers come to this conclusion.

The reason, at least in part, is the Survivor's Bias – the tendency to overestimate the odds of success in light of a successful outcome. We managed to avoid global nuclear annihilation, so we reason that some policy ("deterrence!") was protecting us all along. The problem is that if we ever find ourselves in another similar situation, we will be wildly overconfident in our ability to achieve the same favorable outcome using the same policies. In reality, those policies may have been irrelevant. In a nuclear stare-down with China, deterrence might not work; we don't even know that it worked the first time.

Think of it this way: you're playing roulette, and you bet every penny you own on 14. You chose 14 because you developed an elaborate formula (your birthday divided by ambient humidity times the square root of David Ortiz's on-base percentage) and it gave you that number. The wheel spins and 14 is the winner. You're ecstatic. But instead of taking your winnings and going home, you significantly overestimate the odds of the outcome you just received. "My system," you tell the assembled throng of Social Security pensioners, hookers, and ex-felons, "made victory a near certainty." You play a second time and find out the truth about the odds, namely that they are 1:38 – about 2%. You just got lucky.

Sometimes, rather than finding some justification for our success, we change our view of the odds. The odds couldn't have been so bad after all – how else can the success be explained? For example, the odds of a borderline-insane stripper writing an Oscar-winning screenplay are astronomical. Yet in light of Sunday's awards (and thanks to dozens of rags-to-riches fluff stories in the news) I bet there are an awful lot of waiters, baristas, hookers, and overall louts in Hollywood telling themselves "See? It can be done. If this idiot could make it, it can't be that hard. I can make it!" Mathematically, you can't. Sorry. The odds are horrendous. This person is simply one in a million, and her "one" coming up big does not mean that the odds of it happening were anything but abysmal.

Being pessimistic creatures by nature, of course we do the exact opposite on occasion and imagine the odds to be much worse than they really are. We do have a tendency to let success go to our heads quickly, though, affecting our confidence going forward and our retrospective judgment.

Be Sociable, Share!

2 Responses to “ED vs. COGNITIVE BIASES, PART 2: SURVIVOR'S BIAS”

  1. JDryden Says:

    You mention Hollywood in your illustration, which points to a great example of this phenomenon–the "Whatever 'Worked' To Make This Picture A Hit Will Work If We Make The Exact Same Picture." The assumption being that producers can identify exactly *what* it was that 'worked.' Which, the complexities of a creative process involving the collaboration of hundreds if not thousands of people being what it is, they can't. At all. And yet we keep getting godawful "Keanu Reeves-saves-the-universe" movies because they all think *he's* the reason that THE MATRIX was a hit. Nobody will admit that movies are wildly improbable ventures, almost all of which are destined to recoup their costs but only just. Because (and this relates to your earlier post), all we ever focus on are the hits.

  2. Chris Says:

    Isn't this kind of like the gambler's fallacy? I definitely agree with the fact that 99.9999% of people are not going to be millionaires and should just accept it. Somebody does win the lottery, but it isn't going to be you. I wish more people realized this and understood that they are better off with having the necessities of life taken care of, like health care, housing, food, etc.

    Ed, you should come up with an anti-Horatio Alger character and write a book about it. If you haven't read Nathaniel West's book A Cool Million, you should. A kid goes out and tried to achieve the American dream, and West basically abuses and deforms him throughout the book before killing him. I believe he loses an eye and his teeth, looses his leg in a bear trap, and is scalped by an Indian.