Daniel Kahneman worked with Amos Tversky to develop many of the concepts that today create the principle of Prospect Theory. Many people are familiar with the psychological and economic principle of Game Theory, and Prospect Theory is a similar psychological and economic theory of how people behave when faced with uncertainty. In his book Thinking Fast and Slow, Kahneman shares one of the early surprises of Prospect Theory that he and Tversky uncovered.
Prospect Theory gets its name from the way people behave when faced with different prospects, that is different potential outcomes with different potential likelihoods attached. This is similar to Game Theory, but instead of making decisions while another actor makes decisions that impact your final outcome, in prospect theory you generally are making a choice between a sure thing and an alternative minimal chance outcome. From the theory comes the fourfold pattern, which Kahneman uses to explain why large legal settlements are common, why people participate in lotteries, and why we buy insurance. What was surprising from Prospect Theory was the fourth block in the fourfold pattern, and it describes why some people are willing to take desperate gambles that have incredibly small likelihoods of paying off.
Kahneman writes, “when you consider a choice between a sure loss and a gamble with a high probability of a larger loss, diminishing sensitivity makes the sure loss more aversive, and the certainty effect reduces the aversiveness of the gamble.”
If you are suing a large corporation for damages, you are likely to accept a settlement far below what you are suing for. So, if you are suing the company for $1 million with a small chance of winning (say 5% or less), and the company offers you $95,000, you are likely to feel pressure to take the settlement to make sure you walk away with something. A guaranteed $95,000 is likely to be preferable to the tiny chance that you might actually win your lawsuit and walk away with $1 million. This is one square in the fourfold pattern that fit with Kahneman and Tversky’s prior expectations.
What surprised the pair was the tendencies of individuals when the tables are turned. Say you face the prospect of a large loss of $100,000 with a fringe legal possibility of getting off without facing any losses. If you are offered a settlement where your costs will be $10,000 instead of $100,000, you are likely to feel pressure to turn down the settlement if there still exists some possibility of getting away without any losses. When we look at the expected value we find that accepting the settlement is the risk averse option, but few of us will be content taking the settlement.
We see this with politicians who take an “I’ll risk burning it all down to stay in power” approach, with poker players who get in too deep and misread an opponent or a hand, and with hockey teams who pull the goalie knowing that a sure loss is coming if they don’t take a big risk and get another offensive player on the ice while leaving their net open. When the sure loss is severe enough, then even large gambles with a minimal chance of success are worth the risk.
Kahneman and Tversky were surprised by this because it seems to violate our normal pattern of acting based on expected value. We don’t consciously calculate the expected value of an event, but we usually do act in accordance to expected value. However, in these desperate situations, we actually choose the option with the worse expected value. We become less sensitive to the very likely large loss, and are unwilling to take the sure loss, violating expectations of risk aversion.
Novel Learning 