Say you’re making a significant purchase – such as buying a new house or a new car. At what point in your search should you stop looking and pick the best alternative?
Well, the answer depends on what type of person you are. Some people are maximisers, who will not settle for anything lesser than the best deal. Maximisers search and compare until they are convinced that they have picked out the best option, and bargain until they find the best terms. Other people are satisficers. They have a certain standard they are looking for, and once a choice meets those standards, they make a purchase and stop looking further. When maximisers looks for clothes, they are likely to survey multiple brands, compare deals and after all this research, make the best possible purchase. With satisficers, they walk into a store and pick the first item that meets all their basic criteria. All of us lie somewhere on the spectrum between satisficer and maximiser.
Now coming back to our question about the purchase, Tom Griffiths, a cognitive and computer scientist, offers a rule of the thumb for decision making. Based on optimizing algorithms used in computers, Griffiths says that the answer is 37% of the sample set. Once we’ve looked at 37% of the available houses in a city, he suggests that we pick the next house that is better than anything that we have looked at so far. Framed another way, if you’ve set aside 1 month for picking a house to buy, you must set the benchmark as the best house you have seen in 11 days and in the subsequent days, pick anything that exceeds that benchmark.
But later in his TED talk, Griffiths pulls a fast one. He points out that even with this strategy, the probability of finding the best option would be 37% – logically the odds of finding the best option bears a linear correlation to the ones we have already looked at. So how, then, are we benefited by stopping our search at 37%?
Perhaps the answer lies in the difference between satisficing and maximizing. Consider a basket with 1 red ball, 4 green balls and 5 blue ones. Here, the red ball represents the best option, while the green balls represent the good-enough options for a satisficer. With 3 picks (without replacement), the probability that the red ball is picked is merely 30%, whereas the probability that either one green ball or one red ball is picked is a whopping 94%. With maximisers, since their standard is set to the absolute best option in the market, looking at each additional option bears only a linear correlation with the total number of options examined. With satisficers, each additional option that they consider in the beginning shoots up the probability that it would meet their standard, until the curve tapers off at some point. Needless to say, the choices we have in real life are more complicated than picking balls out of a basket. For the average real world choice, my guess is that we hit the point of diminishing returns at 37%.
With all of our decisions, from ordering food at a restaurant to choosing life-partners, we fall on the satisficer-maximiser spectrum. Satisficing offers the benefit of climbing the green curve that tapers off at 37% rather than walk the linear, red tightrope of maximization.
Inspiration: Tom Griffith’s TED talk on computer science and decision making