Fighting hindsight bias

I'm not sure if this is quite the same thing, but whenever a political issue is in the air, people make all kinds of predictions that essentially cover all of parameter space. E.g. if decision X happens, people will respond with outcome Y or Z or whatever. But it's only after the decision is actually taken that the response is really known. Suddenly it feels genuinely obvious that people respond as they do, even if their response was hardly predicted by anyone. Apparently it's either incredibly difficult to wargame responses even when the prospect of a decision is known, or maybe people are just fundamentally unpredictable.

I think it would be rather fun to keep track of predictions and it should probably be standard practise for any newspaper. To that end, while I was compiling this blog I labelled posts with a "predictions" category if I made (or the article therein contained) a specific, testable prediction. You can already examine these here, but at some point I'll post a review of them.

The saying goes that hindsight is 20:20, but Boyd says it’s worse than that. “It makes people think they can look back at past events and interpret something; it makes them think they have new ability to predict.” He says that in order to correct for hindsight bias, you have to realise you don’t possess a crystal ball.

Thaler says a simple fix could be to write things down; to make a record of how a decision was made at the time so that companies can learn lessons after the event. “Any company that can learn to distinguish between bad decisions and bad outcomes has a leg up,” he says. “Memorialise the fact that the CEO and the other people that have approved this decision all have the same assumptions, that no competitor has a similar product in the pipeline, that we don't expect a major financial crisis.” 

American businessman Warren Buffet, Boyd says, has a formula he calls ‘value investing’. “He follows it, but he knows the odds of succeeding are what they are. He doesn’t always hit magic; he loses occasionally, but he doesn’t say the odds were against him those times. Successful people understand risk for what it is and don’t make revisions in hindsight. Go with the odds, not the gods.” 

Kathleen Vohs says there is one de-biasing tactic she favours known as the ‘consider the opposite strategy’. Before rendering a decision you should flip it and think: how could this not have gone the way you anticipated? “Once you have your little narrative in your mind, think: how could the outcome go in a different direction or not occur at all?” she says. “If you flip the script like that in a number of ways you can reduce hindsight bias.”

EDIT : adding this nice essay by George Orwell on much the same thing.

How hindsight bias skews your judgement

On 8 November 2016 - US election night - the betting odds of Donald Trump winning the presidency had narrowed to five to one. By the following morning, he'd managed to pull off a feat many thought impossible, but that didn't stop the I-told-you-sos - friends, family members and even online soothsayers - who claimed to have known all along that Trump would triumph despite such bad odds.