The signature of the bluffing brain

The best poker players are masters of deception. They’re good at manipulating the actions of other players, while masking their own so that their lies become undetectable. But even the best deceivers have tells, and Meghana Bhatt from Baylor University has found some fascinating ones. By scanning the brains and studying the behaviour of volunteers playing a simple bargaining game, she has found different patterns of brain activity that correspond to different playing styles. These “neural signatures” separate the players who are adept at strategic deception from those who play more straightforwardly.

In the game, a buyer and a seller negotiate over the sale of an imaginary object. The buyer is told about the object’s value in private and suggests a price to the seller, who then sets the actual price. If the price is less than the value, the deal goes ahead, the seller gets the price and the buyer gets the difference between that figure and the object’s value. If the seller’s price is too high, the deal is called off and no one gets anything. This goes on for 60 rounds and at the end of each, the players aren’t told about the outcomes.

Because of this set-up, buyers do best if they set low prices, because they stand to gain the most profits if the sellers accept. Sellers, however, prefer high prices to make the most from their sale. To play successfully, buyers have to keep in mind the object’s real value, the price that they offer, how they think the seller will react to their move, how they could make the most money, and how they can manipulate the seller to accomplish that.

These are the same sort of mental gymnastics that poker players go through when they bet. Both games involve assessing “second-order beliefs”, namely “what people are thinking about what you’re thinking”. “The experimental task is very simple and extremely original,” says Giorgio Cirocelli from University of Southern California, Los Angeles, who also studies the neuroscience of strategic deception.

Bhatt found that by the second half of the games, the buyers had fallen into three different groups, each with a distinct playing style. The incrementalists are relatively honest players. They consistently suggest prices that are around half the actual value, so that both they and the sellers get an equal reward. The conservatives are perhaps the least subtle players; they took little note of the values of their objects and generally opted for low prices. Some even suggested the lowest possible prices on every turn.

The third group was the most interesting: the strategists. They were the best at manipulating the sellers and worked out that the best tactic was to offer higher prices for lower values and vice versa. If they’re given a cheap object, they stand to make very little in profit. They might as well suggest a high price; if the seller goes for it and the trade falls through, they wouldn’t lose much. Later, when a high-value object comes on the market, they can suggest a low price and get a high profit from an unsuspecting seller. In one of the strategists’ own words, “I tried to throw off [the] seller by saying the low things were high.”

The three groups didn’t differ in their IQs or their social backgrounds, so strategists aren’t necessarily more intelligent or educated than those who play with simpler styles. However, the strategists did stand out when Bhatt scanned their brains. As they played, she used a technique called functional magnetic resonance imaging (fMRI) to measure the blood flow in their brain and identify the most active areas.

As the players made their moves, the brains of the strategists were more active than those of the other groups in three areas. The first – Brodmann area 10 (BA10) – sits at the very front of the brain. It has been implicated in many complex behaviours including keeping our goals in mind and looking ahead to the future. Both are important to the strategist, who must bear in mind the long-term goal of making as much money as possible, while playing the short-term tactic of building up the seller’s trust.

The second – the right dorsolateral prefrontal cortex (DLPFC) – also sits to the front of the brain, but slightly off to the side. It’s active during tasks that involve memory, complex decision-making, mental control and social understanding. Again, all are important to strategists; they need to remember their previous suggestions, while holding back the impulse to play a simpler strategy. The more deceptively the players played, the stronger the blood flow in both BA10 and the DLPFC.

The third area – the temporoparietal junction (TPJ) – runs down the middle of the brain and has been implicated in understanding other people’s beliefs and switching attention between different goals. It was unique in that its activity also depended on the value of the different objects. The greater the value and the higher the prize at stake, the greater the activity in a strategists’ TPJ. It’s in these rounds when the ability to know what the seller is thinking really matters. By contrast, the different stakes had no influence on the TPJs of conservatives and incrementalists.

These results support the idea of three different classes of strategic behaviour that can be distinguished by looking at a person’s behaviour as well as their brains. As with many fMRI studies nowadays, some journalists will inevitably write about the possibility of mind-reading machines that can separate devious liars from honest boy scouts. But that would be an overstatement: Bhatt found that her three groups differed in their brain activity but the paper doesn’t say if the brain scans could  accurately classify people into the three groups.

Read Montague, who led the study, says that this should be possible but that it’s “very early days.” For now, he can’t put any sort of numbers on how accurate such a classification scheme would be.  Even so, he says, “One would be very surprised if recording from 50,000 sites in the brain during a bargaining interaction would not give some insights into behavioural strategies.”

Montague also points out that in the game, being a strategic deceiver wasn’t framed as something morally bad to do. It was simply a good money-making strategy. “It would be interesting to offer this style of play paired with some kind of clear moral imperative.  We have not done that experiment yet.” Similarly, Cicorelli says that this work is “of extreme importance for the understanding of the mechanisms of ‘deviant’ social behaviours, such as behaviour of sociopaths.”

This is just one of many future avenues for the work. In this study, the volunteers all ended up with a stable strategy, but it would be interesting to see if people would play as a strategist in one game but a conservative in another. Likewise, do people switch their styles depending on how their opponents play? How does training affect a person’s strategic ability and playing style?

Reference: PNAS http://dx.doi.org/10.1073/pnas.1009625107

Image: by Pokeravond

More related neuroscience:

• Will vs. Grace – are people honest because they resist temptation or because they don’t feel it?
• Undecided voters aren’t really undecided – the hidden side of decision-making
• Unconscious brain activity shapes our decisions
• Electrical stimulation produces feelings of free will
• When is attempted murder more acceptable than harming someone by accident?