Betting Seed: How Birds Give in to Gambling
By Fink Densford
BU News Service
Pigeons are not stars of the animal kingdom. They aren’t featured alongside the patriotic bald eagle, the proud grizzly bear or the noble wolf. Mostly we see them on the streets. We spend money and time trying to keep the wrong ones away, and others we happily serve private banquets to – usually, in public parks.
At heart, however, pigeons are a lot like humans. They, like us, are animals that do well in urban environments and like to keep together. And in addition to the fact that the perky, rolling way they walk resembles certain elements of our own locomotion, they may have what was previously thought of as a uniquely human flaw – the urge to gamble.
In fact, new research is showing that pigeons, along with other members of the animal kingdom, may be as susceptible as humans to the demons of rolling the dice – even when it’s to their own detriment.
Research on this devious behavior began in 2009, when Dr. Thomas Zentall, a professor at the University of Kentucky, asked pigeons to make a simple choice. In his initial experiment, he gave 16 stark white Carneaux pigeons (more often seen in fine gourmet than laboratories) the chance to gamble. He and his colleagues shuffled the birds into metal boxes the size of large ice chests, and presented them with a set of lights that they could peck in hopes that they’d be rewarded with birdseed.
In his initial experiment, Zentall asked pigeons to choose between two options in the form of lit-up, round buttons – about the size you’d see in an elevator. On one side was a chance to gamble. If the bird pecked this button, there was a 50 percent chance they’d be shown a green light symbolizing a win and rewarding the birds with seed, and a 50 percent chance they’d see a red light, which symbolized a loss and ended with no seed.
The other button, if chosen, produced one of two random lights – yellow or blue – both of which symbolized a win, and rewarded the birds with seed 100 percent of the time. This choice was the steadiest option. It always rewarded the pigeons for their effort, and there was no chance of missing out on seed.
Zentall noted that the button colors in all experiments shifted during different trials with different birds, to make sure that there wasn’t simply a color preference at the heart of their decision-making.
Paradoxically, the birds made a decision to peck the gambling light 69 percent of the time in each trial. 13 of 16 birds showed a preference for this random choice that ended up giving the birds less.
The results were unexpected. Zentall said that while he had considered the possibility that the birds may choose poorly, he hadn’t expected to see such a preference for the gambling-like option. But while the results were interesting, he and his lab would need more data to show that the birds were more interested in the risky reward simply because it was a gamble.
In 2010, Zentall gave the pigeons a more human-like opportunity. Instead of a sure payout or none at all, they would get the chance to shoot for a jackpot.
In a small metal box – this time with a more sophisticated pellet delivery system – the pigeons would be asked to choose between two lit-up buttons. The button on the left was a gamble. When they pecked it, there was a two-in-ten chance they’d get a “green” light, resulting in a payout of ten food pellets. But eight out of ten times, the pigeons would get a “red” light that gave them no food, and symbolized a loss. The button on the right represented a non-gambling option. After pecking that button, the pigeons would be shown another colored button, and after another peck the birds would always end up with three pellets.
Again, the birds, like many people, seemed to be happy going for the long shot. In this experiment, 89% of the pigeons chose the worse option, betting on a one-in-five chance at 10 pellets over a sure three. And the reason may have been that the association they had with the light that symbolized that random chance at a jackpot, says Zentall.
At the heart of why pigeons might be so driven to choose the riskier option is a concept known as “contrast”. Kristina Pattison, a graduate researcher in the comparative cognition laboratory at the University of Kentucky explained that in humans, the chance of no payout in gambling leaves expectations low. When expectations and emotional state are low, rewards feel more rewarding. It’s like having a cup with water in it – a normal emotional state would be half full, low would be empty and high would be to the brim. To get from low to full takes more water, and feels more ‘rewarding’ to humans and animals. It’s often known as the ‘contrast effect.’
“You don’t see people crowding around soda machines, eager to feed it dollars to get sodas, or crowding around change machines waiting to get four quarters,” said Pattison. “But you do see them sitting for hours at slot machines.” Gamblers are known to often downplay losses, and slot machines do their best to help them forget, with no shows of light and sound to announce the loss. The event isn’t as impactful, and the player is urged to try again.
So why would pigeons be so convinced to make such a sour decision? Would this be the kind of thing they see in their environment naturally? Dr. Zentall doesn’t think so. “Nothing in nature is anything like the randomness of gambling,” said Zentall. The sort of skewed probabilities in gambling, both in his experiment and the type that hooks humans are not something animals have had to adapt to deal with in the wild.
The mystery of the gambling bird goes deeper than just contrast, though. Further investigation in pigeons has shown other factors also seem to play a part in their susceptibility to gambling-like behavior. In another experiment, published in the Psychonomic Bulletin & Review in 2013, Pattison and Zentall found that enriching pigeons by allowing them to socialize in large, communal cages significantly lowered how often the birds wanted to take the risky option.
The experiment mirrored the earlier model, offering the pigeons a chance at a large payout or a smaller, but more stable paycheck. The only difference laid in what some of the pigeons did with their free time. Half the group got four hours a day of socializing time with other pigeons. Normally, for health and safety reasons, the pigeons are housed independently and are not allowed to interact. But for this experiment, half of the group was cleared and allowed to roost together. The effect was dramatic. Within the first ten trials, socialized birds chose the risky option 50 percent less often than birds that had been left alone.
Pattison said that previous research has shown that enrichment and socialization has a dramatic effect on performing complex tasks in other animals, but this was different. It was effectively immunizing the animals from the dramatic preference – at least for a time. Even the socialized birds caved to the urge to gamble, showing preference against that slowly faded to a preference for the gambling option as the trials went on.
The results could say a lot about an unconscious basis for gambling behavior in humans, but these experiments can’t be called an exact replica of human gambling, said Zentall. He and his colleagues call it gambling-like behavior, mainly because they can’t recreate earning and spending money with animals, or any of the other cultural constructs that human gambling includes. Instead, they are asking the birds to make a statistical decision. “What we’re doing is, essentially, telling them they can have 3 pellets, much like earning money,” said Zentall. “Or offering them the chance at 10.”
The question, however, is close enough to humans to allow at least some theorizing of a simpler explanation for why people might gamble, says Zentall. “I think for all practical purposes, it’s the same.” He’s most interested in seeing if there’s a simpler reason why animals, and people, are so drawn to the jackpot.
Ben Hayden, an assistant professor at the University of Rochester studying Rhesus monkeys for similar gambling-like behavior, thinks the behavior might be genetic. He thinks there’s a possibility that there was, at some point in the past, an evolutionary advantage to be gained with going with the long shot. “Our economic biases are probably more reflective of something that shaped both human and animal evolution.”
Zentall isn’t sure of the cause of the effect, but believes that it can tell us something about humans. In nature, finding a large reward, in terms of food, usually means there’s a higher chance you’ll find more of the same around it. If the conditions are right for one plant to produce a lot of berries, then other plants in the same area are likely to do the same. “So I think the implication is that we think of gambling as an absence of appropriate morality,” said Zentall. “But I think there’s a general tendency, in animals and humans, to be attracted to large winnings, large rewards.”