Caveman Logic: The Persistence of Primitive Thinking in a Modern World (31 page)

Here, once again, you can see that one-two punch: a faulty information processor (it evolved before probabilities were a major part of our information universe) and an overactive agency/pattern detector. So why does a hand of thirteen hearts seem so exceptional while that other hand seems so ordinary? Mundane things must have a higher probability of occurrence, right? Wrong. The simple fact is that the “random” unexceptional hand you were dealt a moment ago is just as unlikely as any
particular
hand in the deck. There are, in fact, over 635 billion of them just waiting for you. The confusion comes from the fact that you have mentally lumped all “undistinguished” hands together into a single category. There is even a name for this mental mistake. Psychologists call it the
representativeness bias
. It’s one of the easiest ways to distort calculations of probability.

There may only be a couple of dozen hands that would strike you as exceptional: for example, all hearts, spades, clubs, or diamonds, or perhaps six of one suit and seven of another. You might also find clusters of four jacks, queens, and kings an equally noteworthy hand, regardless of the thirteenth card. You have some intuitive sense of a “special hand” and you also know (from previous card-playing experience) that its probability is low. The trouble is, Caveman Logic has lumped all unexceptional hands together
as a group
that they don’t really belong to. Collectively, they do have a much greater likelihood of occurrence. But taken as individual events, which is what they are, they are every bit as exceptional as a straight flush from ace to king. They, too, are cause for celebration from a statistical point of view, even though they won’t win you anything in a card game.

This example offers an easy way to see—some people would say to
feel
—just how poorly our minds are designed to understand probabilities. The basic principle becomes easier to grasp if we reduce the numbers drastically and move from playing cards to a drawer full of socks. If you own a dozen pairs, only one of which is orange, what is the probability of reaching blindly into the full drawer and pulling out the orange pair? Most people understand the answer to be 1 in 12, or .0833333. Does that mean reaching in there twelve times will guarantee you success? Of course not. By continuing to sample from that drawer of twelve pairs, there’s a good chance you will repeat some of the earlier choices before you finally grab the orange pair. It may take fifteen or twenty tries, or you may hit the orange pair the very first time. That’s precisely what might happen with our deck of cards. And remember that “nonorange” is not a single category. It consists of eleven different outcomes (e.g., blue, red, black, brown), every one of which has the same probability as orange. Chances are, you’ll receive some of those “losing” hands or socks a second time before you finally get to the one you were looking for.

Gambling is a wonderful place to look for distortions in human perception and cognition. Although life and death are rarely at stake, the experience is usually symbolically strong enough to trigger some pretty fundamental circuitry. Researchers interested in perceptual distortions, cognitive biases, and errors in judgment and decision making have invested major effort studying those who gamble.

Just to be clear, we are not talking about
pathological
gambling. The kinds of errors and distortions that are studied are by no means the work of diseased or abnormal minds. That’s just the point. These mental deficiencies are likely to occur in normal people whenever a hand of cards is dealt or a pair of dice are rolled. There is an entire literature of such cognitive research including journals dedicated solely to studies of gambling behavior (
Journal of Gambling Studies
,
International Gambling Studies
) as well as those with overlapping interests (
Journal of Behavioral Decision Making
). The results have revealed a fairly consistent and not altogether glowing report about how our minds work when information is fragmentary and probabilities are involved.

Gambling situations differ in how much control the gambler actually has (e.g., betting on the lottery versus playing poker), but even in situations in which the gambler has no control over the outcome, most gamblers do a passable job of convincing themselves that some kind of control exists. It is a rare person who buys tickets at a local convenience store for some kind of state-supported numbers game and doesn’t try to bring a measure of personal power (e.g., betting a child’s birthdate) into a situation that is demonstrably out of his or her control.

People will tell you that they really
know
such strategies don’t work, but such knowledge doesn’t mesh with the results of studies in which gamblers provide ongoing verbal narratives during games of chance. Whether or not people truly understand what randomness or independence are, they certainly act as if these were alien concepts. For example, we are not only bad at recognizing randomness, but we cannot even do a passable job of creating it. Back in the precomputer days when psychology experiments needed a random number sequence, it was common knowledge that the worst strategy was to allow the experimenter or his assistant to make one up. They were almost invariably nonrandom in easily demonstrable ways. Embarrassingly, some research from my own lab
¹⁰
demonstrated that even rats could see right through our pathetic human attempts at randomness.

In 2003, Sevigny and Ladouceur
¹¹
monitored the thoughts of gamblers before, during, and after playing a slot machine. Prior to the experience, 98 percent of subjects believed that the outcome of such games was primarily determined by chance. That number dropped slightly to 91 percent following the experience, but remained a sizable majority. However, during the game, in the heat of battle, as it were, the majority of subjects’ thoughts were classified as erroneous and irrational. The authors suggested that two sets of beliefs exist and most subjects “double switch” effortlessly between them. Arguably, subjects know what is sane and reasonable and readily say it when interviewed by a psychologist. However, once thrust into the heat of the game, some other “self,” whose perceptions are distorted and whose judgments are irrational, takes control of the body.

There does not appear to be a single study of gambling that doesn’t underscore how irrationally humans behave. The extent of our mental malfunctioning is truly staggering. The research is not far-fetched; our mental deficiencies are simply that apparent.

The best known of these cognitive distortions is called the
gambler’s fallacy
. Imagine you are in a casino observing a roulette wheel being played. Incredibly, it has landed on red ten consecutive times. You may not have the binomial distribution with you, but you know that outcome seems pretty extreme. In fact, the odds of getting ten consecutive reds (in an unbiased red-black binary system) by chance alone are 1 in 1,024. Suddenly, someone hands you a thousand-dollar bill and tells you to bet it. What do you choose? This is where the gambler’s fallacy kicks in. Most people will choose to bet on black. Why, you ask them? They will tell you “because black is overdue.” In the great cosmic ledger book in the sky, in which black and red have equal probabilities of occurring, we’re going to have to play some serious catch-up to get black back up to that 50 percent total. This trial is as good a place as any to start compensating.

Interestingly, those who do not bet on black tell you they have to choose red. Why, you ask? Because, they explain, there’s obviously something amiss with this roulette wheel and it’s clear that whatever mechanical or ethical quirk is at fault here, it’s going to keep on happening.

The truth is, it’s hard to fault either account and harder yet to see why they represent some kind of breakdown in mental functioning. But they do. The reason is simply that in both cases the bettor has failed to realize that each spin of the wheel is an independent event, neither influenced by nor influencing any adjacent trial. In short, the real probability of rolling an eleventh red after ten consecutive reds is not 1 in 2,048, but 1 in 2. Most people will simply tell you that account feels wrong. And so we have a handful of psychologists telling the rest of the human race that they shouldn’t listen to mental circuitry that has taken hundreds of thousands of years to evolve.

This impasse is tied to the fact that gambling casinos are every bit as unnatural environments as psychology labs. In the real world events are
not
typically independent of each other. You can, indeed you
should
, use past cases as predictors of present ones. In that sense, gambling casinos, like psychology labs, are dangerous and misleading places where unnatural things can happen. Casinos are not registered charities. They are there to take your money and they depend on your making foolish choices. Your mental modules can get you into trouble, but that is not because they are faulty. It is because the environments in which they evolved are very different from where they are being tested. If you can anticipate that and take a moment to second-guess the messages delivered by those modules, you’ll be in much better shape.

The usually unstated corollary is that such mental errors are not confined to gambling. Gambling is simply a value-neutral way to trigger irrationality. There is no other domain in which we dare hold the human mind up to such ridicule. We
know
what is correct in a gambling experiment. If a person believes that a previous outcome on a roulette wheel influenced the present one, we can say, “No, it didn’t. Your perception is faulty. We are right and you are wrong.” We do not have to worry about being politically incorrect or offending a powerful group who can affect our advertising revenue or our national security. Who cares if we tell some nitwit at a roulette wheel that he is seriously deluded? Hardly anyone will come to his defense. But can you imagine telling someone that winning the lottery was not proof of the power of prayer? In social or political terms, not all Caveman Logic is created equal. The underlying mechanisms may be the same, but the manner in which these distortions are expressed runs the gamut from silly to sacrosanct.

And so, using gambling as a venue to study mental distortions is socially acceptable. In fact, it is even highly valued. It’s not that humans are in a hurry to discover the inadequacies of their minds. Rather, the major impetus for studying these cognitive distortions—at least as far as the public is concerned—is the alleviation of problem gambling. Problem gambling is bad, and we want to “cure” it like we do any other addiction. Such research is thus on the side of the angels; it will reduce human suffering. Of course, try to take our findings from gambling research and apply them to everyday, socially supported situations and one will find a quick end to government funding for this line of research.

Just ask any of my colleagues who happen to be in their forties, fifties, or sixties and they will be only so glad to tell you what is wrong with their memories. This litany of complaints runs the gamut from petty annoyances (“I’ve lost my car keys again,” “I can’t remember that student’s name”) to more serious concerns about incipient Alzheimer’s disease. Few people I know are happy with how their memories are working.

Part of the problem is that we are keenly aware of the occasional frustrating or embarrassing flaws and almost oblivious to the majority of cases when our memories serve us well. This alone makes us highly suspect as data analysts, since we are prone to what we have previously labeled as the confirmation bias (“I have a rotten memory and now I’m going to collect evidence to prove it.”) Worse yet, we are unlikely to view those occasional lapses as evidence of something good. However, rather than signaling that we are going straight to cognitive hell, such infuriating glitches might actually suggest that everything is working according to factory specifications. In
The Seven Sins of Memory
,
¹²
Daniel Schacter concludes that memory’s most common failings are “byproducts of otherwise adaptive features of memory, a price we pay for processes and functions that serve us well in many respects.”

The problem is simply that the architecture of our memories has been designed by evolution to serve us well in the
majority
of cases. And it does. But, by definition, there will be exceptions in which the same strategies fall flat, leaving us to wonder how human memory, especially our own, can be so flawed. Experimental psychologists and cognitive neuroscientists love to design situations to short-circuit normal memory. Such research is not necessarily geared to discrediting the facets of cognition that have served us and our ancestors well. Rather, it allows us to understand the mechanisms of memory more clearly by revealing its boundary conditions.

Here are some examples: We cannot possibly recall all the information to which we have been exposed. It is reasonable, even desirable, to set aside information (like unused phone numbers) that we no longer actively require. On those occasions when we are required to retrieve such dormant information, we are likely to come up empty. Is this really a design flaw? Perhaps it feels that way during the palpable struggle for retrieval. But consider the alternative: everything is slavishly encoded and stored. Nothing is lost. The trade-off seems obvious. Yes, that dormant phone number (or old license plate) can be retrieved, but the entire system is slowed down by the sheer weight of storing and retrieval of all-but-useless information.

Schacter goes on to detail a variety of cases in which forgetting is a virtue and how the inhibition of irrelevancies keeps us functional in a constantly demanding world. But there are downsides to occasional problems with human memory as well. The effects of suggestibility can have devastating effects on courtroom testimony. Confusion about the source of a memory (even when the content itself is intact) can lead to misattribution or unintentional plagiarism. There are numerous instances of this in the music business, for example, when a composer mistakenly believes that he has written a melody, which in fact stems from an earlier listening experience whose source has been forgotten. Examples of this include well-known composers and compositions: George Harrison’s “My Sweet Lord” turned out to be an unconsciously plagiarized version of “He’s So Fine” by the Chiffons. Not even jazz musicians are immune to the process. The jazz anthem “Misty” came to Erroll Garner during a dream, he often reported. He may have been sleeping, but the song he was dreaming of was called “Ebb Tide,” which was a major pop hit at the very time Garner spontaneously wrote “Misty.”