From Tuesday’s science pages: A new study confirms that humanity is designed to underestimate the luck of the draw
Where is the line dividing luck and skill? How much does chance determine success? It turns out that, from sport to the stock market, we are constantly deluded by an illusion of order, when in fact we are assailed by randomness. And a study released this month claims that, in one area where we ultimately expect skill to triumph – namely poker – we may be far more at the mercy of fortune than we believe.
Anyone who has ever played Texas hold ’em or the like thinks they know how much talent is involved. But, according to Professor Gerhard Meyer and his team at the University of Bremen’s Institute of Psychology and Cognition Research, they are wrong. He says that we are, as in so many parts of our life, fooled by randomness into believing that we are masters of our destiny.
Poker, he insists, is largely a game of chance. Some games are clearly luck – roulette, for instance – while others, like chess, are matters of pure skill. But, as Meyer explains, “most games are heterogeneous, because they involve both chance and skill, and poker falls into this category”. The question is, how much skill is involved? Meyer’s study found that the workings of randomness far outweigh the importance of talent or hard work.
This result does not surprise Michael Mauboussin, who is an investment strategist and the author of The Success Equation, which examines the boundaries between luck and skill. He says that it is very difficult for humans to tell luck and skill apart. We are simply not designed to notice randomness.
This disconnect is perhaps most visible, he says, in sport. “Imagine a continuum, with all-luck-no-skill on the left and all-skill-no-luck on the right,” he says. “In all-luck-no-skill, games might as well be the toss of a coin; in all-skill-no-luck, Team 1 always beats Team 2, Team 2 always beats Team 3, and so on.” Using a computer algorithm, Mauboussin has compared these scenarios to real life. Most team sports, he says, are about halfway along the spectrum.
Over the course of a long season, the best teams tend to win, and the worst to lose. But that is the product of a large sample: in the short run, randomness reigns. Several bad results in a row will be ascribed to poor form – but it could simply be the equivalent of a series of tails coming up in a coin-toss.
Indeed, the evidence that such a thing as form exists is, at best, equivocal. “There are some sports where it may be a factor,” says Mauboussin. “Bowling and archery are two of them. But for the most part it doesn’t exist. Yet as a player, or as a fan watching the game, you sense it; it’s a strange mirage.”
In Prof Meyer’s study, 300 poker players took part, playing 60 hands each on tables of six. They were divided into “expert” and “average” players, and their ability to make money from good, bad and average hands was assessed. It turned out that – as you might expect – “expert” players lost less money on bad hands; but surprisingly, they did no better than average players on mediocre hands and even made slightly less on good ones. The conclusion, says Meyer, is that “poker players overestimate the skill factor in their play”.
It is not just sport where this applies, but the stock market, too. Some years ago, Professor Daniel Kahneman, a Nobel Prize-winning psychologist and the author of Thinking, Fast and Slow, analysed the work of financial traders. He found that while the traders themselves had a subjective experience of making sensible predictions of complex scenarios with great skill, they were in fact playing a game based almost entirely on luck.
“The test of skill is whether you get the same individuals consistently doing well,” he explains. “With stock-picking, different people do well every year; someone who did well one year had no advantage the next.”
This tendency to imagine order where there is only randomness has been observed for decades. Kahneman tells a story about London during the Second World War. Towards the end of the conflict, the city was bombarded with V2 missiles. Newspapers noticed that the strikes were clustered in various areas. It was suggested that the reason was that they were targeted at working-class districts, or that German spies were living in the unaffected regions.
A statistician called William Feller eventually explained the real reason: the pattern of clusters and empty spaces was exactly what you’d expect if the distribution was indeed purely random. But to the human brain, hungry for meaning, it looked like the product of deliberate targeting.
The question is, why are humans so keen to see patterns where there are none? Part of the reason is a phenomenon called hindsight bias. “If I say that what’ll happen in the future will be a mixture of luck and skill, you’ll nod and agree,” says Mauboussin. “But once it’s happened, we forget that, and attribute it to skill.”
This is dictated in part by our need to explain our experiences in terms of a causal story. “There’s a part of our brain that’s called the interpreter,” he says. “It’s designed to make sense of what we’ve seen, to give it a narrative. And we always see causes; so if Person A succeeds where Person B fails, we assume that Person A had some skill that Person B didn’t.
“Even when we know it’s random, we can’t help but see the workings of skill.”
This hyperactive pattern-detector is likely to be an evolutionary adaptation, says Kahneman: a false positive will generally be less harmful than a false negative, an imagined lion less of a problem than an unnoticed one. “There will be evolutionary advantages to being hypersensitive to clues in the environment, to patterns,” he says.
Teasing out the role of chance is controversial. For his part, Mauboussin thinks that poker may involve a little more skill than Meyer believes: “There’s a simple test for whether something is a game of skill: can you lose on purpose? In roulette, no; in poker, yes.” Professor Meyer agrees that the difference between good and bad players may be visible in the long run, but for most casual players of internet poker, he says, the randomness plays the decisive role.
Either way, poker can certainly reveal to us the biases in our thinking, says Todd Simkin, a financial trader who now uses it to teach strategy to other traders. Humans are are obsessed with outcomes, he says. “You have to distinguish the outcome from the decision process.” So we tend to credit a player with doing well if a long shot comes off – or a trader if his high-risk stock pick comes good.
But, he says, that’s not how we should work. “Say you bet, and the next card [which you need] is the six of clubs,” he says. “That’s irrelevant. What matters is that there were only three sixes in the pile, and 40 other cards. You got lucky.” He uses the game to teach the risks of hindsight bias to traders.
Almost all areas of life are plagued by randomness, but our story-loving brains find it almost impossible to see. “It takes a lot of discipline to think abstractly,” says Mauboussin. “It’s amazing, in sport for instance, how few people – even coaches, or reporters – notice it… we’re fooled by randomness all the time. The important thing is what we do about it.”