Past Experience is Good; Random Decision-Making is Better

We all know that making decisions from past experience is wise. We learn from the benefit of hindsight. We learn to make small improvements or radical shifts in our thinking and behaviors based on history and previous empirical evidence. Stock market gurus and investment mavens will tell you time after time that they have a proven method — based on empirical evidence and a lengthy, illustrious track record — for picking the next great stock or investing your hard-earned retirement funds.

Yet, empirical evidence shows that chimpanzees throwing darts at the WSJ stock pages are just as good at stock market tips as we humans (and the “masters of the universe”). So, it seems that random decision-making can be just as good, if not better, than wisdom and experience.

From the Guardian:

No matter how much time you spend reading the recent crop of books on How To Decide or How To Think Clearly, you’re unlikely to encounter glowing references to a decision-making system formerly used by the Azande of central Africa. Faced with a dilemma, tribespeople would force poison down the neck of a chicken while asking questions of the “poison oracle”; the chicken answered by surviving (“yes”) or expiring (“no”). Clearly, this was cruel to chickens. That aside, was it such a terrible way to choose among options? The anthropologist EE Evans-Pritchard, who lived with the Azande in the 1920s, didn’t think so. “I always kept a supply of poison [and] we regulated our affairs in accordance with the oracle’s decisions,” he wrote, adding drily: “I found this as satisfactory a way of running my home and affairs as any other I know of.” You could dismiss that as a joke. After all, chicken-poisoning is plainly superstition, delivering random results. But what if random results are sometimes exactly what you need?

The other day, US neuroscientists published details of experiments on rats, showing that in certain unpredictable situations, they stop trying to make decisions based on past experience. Instead, a circuit in their brains switches to “random mode”. The researchers’ hunch is that this serves a purpose: past experience is usually helpful, but when uncertainty levels are high, it can mislead, so randomness is in the rats’ best interests. When we’re faced with the unfamiliar, experience can mislead humans, too, partly because we filter it through various irrational biases. According to those books on thinking clearly, we should strive to overcome these biases, thus making more rational calculations. But there’s another way to bypass our biased brains: copy the rats, and choose randomly.

In certain walks of life, the usefulness of randomness is old news: the stock market, say, is so unpredictable that, to quote the economist Burton Malkiel, “a blindfolded monkey throwing darts at a newspaper’s financial pages could select a portfolio that would do as well as one carefully selected by experts”. (This has been tried, with simulated monkeys, andthey beat the market.) But, generally, as Michael Schulson put it recentlyin an Aeon magazine essay, “We take it for granted that the best decisions stem from empirical analysis and informed choice.” Yet consider, he suggests, the ancient Greek tradition of filling some government positions by lottery. Randomness disinfects a process that might be dirtied by corruption.

Randomness can be similarly useful in everyday life. For tiny choices, it’s a time-saver: pick randomly from a menu, and you can get back to chatting with friends. For bigger ones, it’s an acknowledgment of how little one can ever know about the complex implications of a decision. Let’s be realistic: for the biggest decisions, such as whom to marry, trusting to randomness feels absurd. But if you can up the randomness quotient for marginally less weighty choices, especially when uncertainty prevails, you may find it pays off. Though kindly refrain from poisoning any chickens.

Read the entire article here.