The Near Miss

Mo over at Neurophilosophy has an excellent summary of a new paper on near misses and addictive gambling:

Henry Chase and Luke Clark of the Behavioural and Clinical Neuroscience Institute in Cambridge have previously found that the brain responds to near miss gambling outcomes in much the same way it does to as winning. In moderate gamblers, both types of outcome activate the reward circuitry, and although near miss events are experienced to be somewhat less rewarding than wins, they nevertheless increase the desire and motivation to gamble. For games involving skill, near misses indicate an improvement in performance and spur the player to try again. But gambling is a game of chance, which distorts gamblers' thought processes - near misses cause them gambler to overestimate both the level of skill involved and their chances of winning. This spurs them to continue gambling.

The new study extends these earlier observations to regular gamblers, with the aim of establishing whether or not the response to near misses is related to gambling severity. Chase and Clark recruited 24 regular gamblers, The participants were asked to perform a computerized gambling task while their brains were scanned by functional magnetic resonance imaging (fMRI). Two slot machine reels, each with the same six playing icons, were presented to the participants on the screen inside the scanner. In one condition, they were required to select an icon on the left reel and then spin the right reel. In another, the icon was randomly selected for them by a computer. If the icons matched after the reels stopped spinning, they were rewarded with a small amount of money.

After collecting the fMRI data, the researchers focused on the midbrain, which contains neurons that signal reward by releasing the neurotransmitter dopamine. They again found that near misses activate the reward circuits, confirming the results of their previous study. Significantly, they also found that gambling severity, as measured prior to scanning by the South Oaks Gambling Screen, could predict the midbrain's response to a near miss. The more severe a participant's gambling habit, the stronger was the midbrain response to a near miss. In other words, near misses were most rewarding for the pathological gamblers, who experienced them as being almost as rewarding as a win. The participants who gambled less severely also found near misses rewarding, but to a lesser extent.

This data won't surprise Vegas. The details of slot machines are clearly rigged to trick gamblers into thinking they scored a near miss (say, two cherries and a seven) when, in fact, their near miss was just another random event.

The larger question is why this excruciating activity - we're losing money over and over again - is so addictive. The answer, I think, has a lot to do with how our brains process rewards. The popular myth of dopamine is that the neurotransmitter equals pleasure, that it's the hedonist chemical responsible for sex, drugs and rock n' roll. The dopaminergic reality, not surprisingly, is actually much more complicated. Consider the landmark work of Wolfram Schultz. His experiments followed a simple protocol: He played a loud tone, waited for a few seconds, and then squirted a few drops of apple juice into the mouth of a monkey. While the experiment was unfolding, Schultz was probing the dopamine-rich areas of the monkey brain with a needle that monitored the electrical activity inside individual cells. At first the dopamine neurons didn't fire until the juice was delivered; they were responding to the actual reward. However, once the animal learned that the tone preceded the arrival of juice -- this requires only a few trials -- the same neurons began firing at the sound of the tone instead of the sweet reward. And then eventually, if the tone kept on predicting the juice, the cells went silent. They stopped firing altogether. Schultz calls these cells "prediction neurons," since they are more concerned with predicting rewards than actually receiving them.

What's interesting about this system is that it's all about expectation. Our dopamine neurons constantly generate patterns based upon experience: if this, then that. While this neural software normally works great - it's an incredibly efficient form of learning - it gets reliably confused by random systems, from slot machines to Wall Street.

To understand why, it helps to think about the slot machine from the perspective of your dopamine neurons. While you are losing money, your neurons are struggling to decipher the patterns inside the machine. They want to understand the game, to decode the logic of luck, to find the events that predict a payout.

But here's the catch: slot machines can't be solved. They use random number generators to determine their payout. There are no patterns or algorithms to uncover; studying our near-misses won't tell us how to win. There is only a stupid little microchip, churning out arbitrary digits. At this point, our dopamine neurons should just surrender: the slot machine is a waste of mental energy. But this isn't what happens. Instead of getting bored by the haphazard payouts, our dopamine neurons become obsessed. When we pull the lever and get a lucky reward, we experience a rush of pleasurable dopamine precisely because the reward was so unexpected. (According to Wolfram Schultz, such unpredictable rewards are typically three to four times as exciting, at least for our dopamine neurons, than rewards that can be predicted. The clanging coins and flashing lights are like a surprising squirt of juice.) The end result is that we are transfixed by the slot machine, riveted by the fickle nature of its payouts.

A near miss, in other words, isn't just a near miss. It's also a reminder that we almost got a surprising reward, and thus keeps our dopamine neurons from getting bored. Nothing, after all, is quite as nice as that reward we didn't expect.

More like this

This reminds me of what the locus ceruleus does in the brain, which is a center (also in the midbrain) that releases norepinephrine (NE) rather than dopamine. An experiment was done in which monkeys faced a projector screen that flashed either a horizontal or a vertical strip of light. The monkeys had their hands pressing down on a type of 'space bar' on a keyboard, and they were trained to release the bar to a light flash that was a horizontal strip but to keep the bar compressed to a vertical strip. Each time the monkeys performed correctly they got a squirt of apple juice; each time they performed incorrectly they got shocked. The locus ceruleus released NE when they correctly released the bar and got juice, but what is interesting is this: the locus ceruleus released NE EVEN WHEN THE ACCIDENTALLY RELEASED THE BAR to a vertical strip of light and got shocked! NE is released more to the RESPONSE (the releasing of the bar) than to the stimulus (the strip of light, horizontal [correct] or vertical [incorerect]).

This is synonymous with the example of dopamine, which is that dopamine is released to correct responses (three pairs of cherries) AS WELL AS to incorrect responses (two paris of cherries and a seven).

Yes, but the interesting question is whether commercial slot machines use such teasing tactics or whether each spin and each wheel are truly independent (as one might assume). Is there any way to find out, or is that a trade secret? I assume there are laws governing mean payout, but what about independence? As long as you have a processor chip in there, it would be easy enough to program various patterns in, e.g. more than the expected number of "near misses", or even "if the machine has been idle for a while, and someone starts using it, increase the payout probability for a while to get the player hooked, then decrease it." Would this be illegal for any reason?
David

And what are the brain chemicals that create the sinking feeling of disappointment and denial?

Would the same mental processes work for prayer. Say a person prays ten times and, by chance, one prayer comes to reality. Would that hit encourage more prayer? And would the nine misses have the effect of "near misses"? Or am I reading too much into this?

By Nebularry (not verified) on 19 May 2010 #permalink

I'd be surprised if they didn't rig the machines to generate more near misses. Casinos are very clever about getting you to bet more.

I'm not a neuroscientists, but I believe the same chemical dopamine is also responsible for the disappointment of incorrect predictions. The trained monkey gets a dopamine reward when the tone is played, but if the expected apple juice is withheld, the dopamine levels dip drastically.

I think it's important not to put all gambling into the same group as games of pure chance like slot machines and roulette. Sports betting and poker, to name two examples that jump out at me, involve a considerable amount of skill and "near misses" in those endeavors can often be very meaningful.

To be honest...

The near misses were the men who should have kissed the ground and married me, but to whom I'd let go.

Talk about an ego.

Near misses in life are simple. You'd thought you nearly got to paradise, but one step less of the mark, you were derailed.

Yet, perhaps, it wasn't paradise you'd missed, but hell you'd averted.

Think about that. *wink*

Enjoy enjoy

Interesting article, thanks.

@Paul & David:

I work as a programmer developing slot machines, and I can say that, in the UK at least, it would be against current regulations for random running machines to pull such stunts as faking near misses. The source code has to go to external verification houses to ensure compliance, and they are *very* strict (I once had a game fail because a missing equals sign turned what should have been a 100% gamble into a 99.98% gamble.)

From what I know of the US casino market, the rules are, if anything, even stricter.

tsg's post of May 20 makes a valid distinction between games of pure chance such as slot machines and skill-dependent games like sports betting and poker. Likewise the securities markets, which many believe to be random processes, actually show, under statistical analysis, periods of strong serial dependency. While it's clear that many fund managers accomplish little more than gambling with their clients' money, there are also immensely successful traders whose methods are rigorously systematic.

By William H Phillips (not verified) on 24 May 2010 #permalink

In regards to the 'Prediction Neurons' theory/idea, which relates to some case I heard of where athletes would taste (but not swallow) Gatorade and reap the same performance increase (but following the full theory, this effect would rub-off as the brain finds error in the pattern), I'm wondering if (and if not, why not) my following idea would work:

Let's say that once per day I chew and enjoy sugar-rich junk foods Without actually swallowing, and spit the food out. Wouldn't I gradually lose my natural craving for sugar? And wouldn't this be an awsome dieting technique?

By Nick Green (not verified) on 27 May 2010 #permalink