Field of Science


Chimps Prefer the 2-Point Conversion

If non-human great apes were coaching more football games, you could expect to see fewer extra points being kicked. We risk-averse humans usually prefer kicking an easy extra point after a touchdown, rather than attempting a more difficult 2-point conversion. But chimps and other great apes, after considering their odds, usually opt for the greater risk and the bigger reward.

By "reward," I mean banana.

Researchers at the Max Planck Institute in Germany tested a group of chimpanzees, bonobos, gorillas and orangutans on their risk-taking strategies using chunks of banana. They wanted to know whether the apes' likelihood to go hunting for banana pieces hidden under cups, rather than taking a smaller banana piece already in front of them, depended on the "expected value" of their choices. Expected value is simply an item's worth, multiplied by your odds of getting it. If a 2-point conversion attempt is successful exactly half the time, then its expected value is 1 point.

The 22 apes each sat through a series of experiments involving banana bits in cups. On one side of a table, they saw a small piece of banana placed under a yellow cup. Next to that was a row of blue cups,  anywhere from one to four of them. Under one of the blue cups was a larger piece of banana.

The apes knew the larger piece of banana was hidden under one of the blue cups, but unless there was only one blue cup, they didn't know exactly where the banana was. (They understood the setup because there was also a series of trials in which the apes watched the banana being placed under one of the blue cups.) In each trial, an ape could point to just one cup and get the reward--if there was any--underneath.

The yellow cup was a guaranteed small reward. The blue cups were a gamble. And the size of the gamble (in other words, its expected value) depended on how many blue cups were on the table. It also depended on the difference in size between the two banana chunks. The "safe" piece of banana in the yellow cup ranged from one-sixth to two-thirds the size of the large piece.

The researchers found that the apes' decisions did correlate to the expected value of their options. Overall, as the expected value of picking a blue cup increased--there were fewer blue cups on the table, or the safe piece of banana was small and untempting--apes opted more often to try a blue cup. When the expected value of the gamble was lower--because there were a lot of blue cups to choose between, or the safe banana piece was large to begin with--they were more likely to stick with the yellow cup.

Adjusting choices based on the expected value of each option is similar to how humans would decide. But the apes were less human-like in their general propensity for risk. Even at the lowest possible expected values, apes chose to gamble on a blue cup more than 50% of the time.

In other words, apes acted more like humans playing the lottery than humans kicking an extra point after a touchdown. These apes, of course, didn't have their coaching jobs on the line. They might have just enjoyed playing the cup game. And in a human football game, there are plenty of situations in which a kicked extra point is better than going for 2--even though its expected value, with a success rate of about 50%, is the same.

But even outside of football games, humans are known by psychologists for being risk averse, especially when it comes to potential gains. We'd rather take a small guaranteed reward than a larger and riskier one. (For losses, though, we tend to feel the opposite way.)

When the researchers broke down their results by species, they found that while all four species were risk prone, bonobos were a little more conservative in their choices than chimps were. With only a small number of ape subjects, it's hard to draw any serious conclusions. But it's interesting to speculate about the differences between us and our two closest living relatives. Have chimps evolved to take more risks, always gambling on finding something better, because in the wild they must search for fresh fruit year-round? Can bonobos afford to be more conservative because their diet in the wild is more flexible? What factor in our past put risk-averse humans at an evolutionary advantage?

Next time your favorite football team takes an overly conservative extra point, don't blame the coach for his evolutionary history. You could always call up the owners, though, and suggest they hire a chimpanzee instead.

This post was chosen as an Editor's Selection for

Photo: Flickr/Mat_the_W 

Haun, D., Nawroth, C., & Call, J. (2011). Great Apes' Risk-Taking Strategies in a Decision Making Task PLoS ONE, 6 (12) DOI: 10.1371/journal.pone.0028801


  1. Hi Elizabeth,

    I wanted to let you know that I nominated you for the Liebster Blog Award!

    Here’s the post:



  2. "Have chimps evolved to take more risks"

    This is over-interpreting and even worse not taking all data into account.

    Those chimps KNOW they get a stable food supply each day, they don't need to rely on the gambled food - unlike the football players and coaches their 'livelyhood' doesn't depend on the game - and they 'know' it (reasonably expect it from prior experience, they have been fed all their lives haven't they).

    For those chimps the game is just that a game - it is more akin to someone take some money he doesn't really need and waste it in Vegas - the risk assesment is different under those circumstances.

  3. Anonymous: I agree that the study is limited. We don't know if the chimps would have behaved the same way with a reward that was more valuable to them. But when psychologists study risk aversion in humans, it's not necessarily about major gains and losses--you can see the phenomenon in students gambling for $5 in a psychology lab. They don't need those dollars any more than the chimp needs the bite of banana.

  4. But then it would be more appropriate to compare the chimps behavior with humans under similar conditions instead of (highly commercialized) football.
    Also it might be interesting to take the emotional state of participants (both chimps and humans) into account.
    I (completely out of the guts, not scientifically proven) claim the chimps participate in those tests in 'good mood' or 'fun' state of emotion (it's playtime for them).

    Humans have different risk assesment based on their emotional state too - checking against results from different tests (under pressure, neutral, happy/fun) might allow to get better data to compare.


Markup Key:
- <b>bold</b> = bold
- <i>italic</i> = italic
- <a href="">FoS</a> = FoS

Note: Only a member of this blog may post a comment.