Let’s play a game. I am going to flip a coin. If you guess correctly which side it will land, I will give you $1. If you get it wrong, you will give me $1. We will play this game 100 times.
Since we are using my coin, it may be weighted to preferentially land on one side. I am not going to tell you which way the bias works, but after 20 - 30 flips, you will notice that the coin lands with ‘heads’ facing up three times out of four.
What are you going to do?
This was the question posed by Professor Andrew Lo from the Massachusetts Institute of Technology, during his talk last week for the Origins Institute Public Lecture Series. Lo’s background is in economics. It is a research field with a problem.
In common with subjects such as anthropology, archaeology, biology, history and psychology, economics is about understanding human behaviour. While the findings in these areas may not always be relevant to one another, they should not be contradictory. For example, studying mating behaviours in anthropology should be consistent with the biological mechanics of human reproduction. The problem is that economics is full of contradictions.
In the example of the weighted coin, the financially sound solution made by a hypothetical ’Homo-economicus’ would be to always choose ‘heads’. This would correctly predict the coin toss 75% of the time and net a tidy profit. However, when this choice is given to real Homo-sapiens, people randomise their choices. In fact, humans match the probability of the coin weighting, selecting heads 75% of the time and tails 25% of the time. Moreover, if the coin is surreptitiously switched during the game to one with a weighting for 60% heads, 40% tails, then the contestants change their guesses to also select heads 60% of the time.
Economics tells us this is not the choice we should be making. So why do we do it?
This selection choice holds in even much more serious situations. During World War II, the U.S. Army Airforce organised bombing missions over Germany. These were incredibly dangerous flights with two main risks: the plane could receive a direct hit and be shot down, or it could be pelted with shrapnel from anti-aircraft artillery. In the former case, the best chance of survival would be a parachute. In the latter, it would be safest to wear a metal plated flak jacket. Due to the weight of the flak jacket, it was not possible for pilots to select both pieces of equipment. The army therefore gave them a choice of items.
The chances of being hit by shrapnel were roughly three times higher than being shot down. Despite this, pilots selected the flak jackets only 75% of the time. When the army realised this was delivering a high death toll to their pilots, they tried to mandate the flak jacket selection. However, the pilots then refused to volunteer for such missions, claiming that if they were taking their lives in their own hands, they ought to be able to pick the kit they wanted.
What is perhaps even stranger is that humans are not the only species that does this probability matching. Similar behaviour is seen in ants, fish, pidgeons, mice and any other species that can select between option A and B. This can be tested on fish surprisingly simply. If you feed a tank of fish from the left-hand corner 75% of the time and from the right-hand corner 25% of the time, then 25% of the fish will swim hopefully to the right side of the tank when it is time to be fed.
Lo’s feeling was that this common probability matching trait must have stemmed from a survival advantage. He therefore turned from economics to evolution.
Lo admitted that crossing from one field of research to another can be daunt-ing. He humourously recalled an attempt by his university to bring together researchers in different areas during an organised dinner.
“We were speaking different languages! I never went back!” he told us.
Despite this disastrous first attempt, Lo has found himself drawn into projects that span fields.
“Interdisciplinary work comes about naturally when you’re trying to solve a problem,” he explained. “Problems don’t care about traditional field boundaries.”
To explore the possible survival advantages of a probability matching instinct, Lo considered a simple choice for a creature: where to build a nest?
In Lo’s system there are two choices for nest location: a valley floor by a river or high up on a plateau. While it is sunny, the valley floor provides the best choice. It is shaded with a water source, allowing the creature to successfully breed. Meanwhile, the plateau is too hot and any off-spring will die. However, when it rains, the situation is reversed: the valley floods and kills all the creatures, while the plateau provides a safe nest.
If it is sunny 75% of the time, which is the best choice?
Lo ran this experiment as a computer model for five sets of imaginary creatures. One group (the ‘Homo-economist’ creatures), always picked the valley. The second group selected the valley 90% of the time, the third 75% of the time and the forth and fifth groups 50% and 20% of the time.
At the beginning of the simulation, the ‘Homo-economists’ had a population boom, successfully producing more children then any other group. However, then it finally rained in the valley. All of this group were killed, leaving only the groups which had some creatures on the plateau. After 25 generations of creatures (corresponding to 25 chances to rain or shine), the group which selected the valley 75% of the time were in the lead. In short, while the valley was the better bet for any one individual, the species did better if it probability matched.
Lo’s model is simple, but it does offer an explanation to why humans may not make the most logical decision when confronted with two choices: it is an instinct that helped the human race survive. Perhaps even more interestingly, it may even explain the origins of intelligence.
Lo next posed the question: What is intelligence?
He admitted this is a difficult quantity to assess, and suggested the reverse is much easier to recognise: What is stupidity?
To demonstrate this, Lo presented three pictures. The first showed two men listening to the radio in a swimming pool. Said radio was attached to the electricity supply via a cable held out of the water on a floating flip-flop shoe. In the second image, a man was holding open a crocodile’s jaws so he could take a photograph within its gaping maw. The final photograph depicted a man working underneath his truck, which was raised at a precarious 120 degrees using two planks of wood.
This caused laughter from the audience. We recognise this as stupid, Lo suggested, because we know these people are about to take themselves out of the gene pool. If intelligence is the opposite of stupidity, it must mean intelligent behaviour is linked with our chances of survival. By evolving to maximise the probability of our species surviving, we are selecting for our own impressive IQ.
--
The Origins Institute is based at McMaster University in Canada. Their public lectures are free to attend and live recordings can be found on their website.