A Crowd Counted Jelly Beans Almost Perfectly Until It Started Talking
Private guesses averaged 1,771 against a true count of 1,776. Public guesses averaged about 900, a warning about listening to market chatter.
My First Million · There’s 98 days until 2027...Do this and you’ll thank yourself on Jan 1stThe jar sat at the front of the auditorium, full of jelly beans. Joel Greenblatt, an investor who beat the stock market for decades with his own picking strategy, invited the audience to come take a look.
Anyone could walk up and study it. They could not open it, but they could stare, count the beans in one row, estimate the height, and do the math in their heads. The one rule was to write the guess on a slip of paper and not show a neighbor.
When Greenblatt tallied the slips, the average was 1,771. The jar held 1,776. One hundred people, thinking alone, had missed the truth by five beans.
He kept that near-perfect result to himself. He tried the test a second way. This time, he said, just call out your guess.
The first person said 850. The next person had been thinking 2,000, and suddenly that felt way too high, so he said 1,300. The person who had been thinking 1,700 heard 850 and 1,300 and trimmed his own number, too. When the shouting stopped, the average had fallen to about 900. Nobody had left and the jar had not moved. The difference was that everyone could hear everyone else, and the crowd went from five off to about 50 percent off.
"This is the stock market," Greenblatt told them.
Left alone to study a business and estimate what it is worth, he said, smart people do pretty well. In the market, they stop thinking alone. They listen to the news, to television personalities like Jim Cramer, to analyst reports, and to the ticker itself. When the price falls, they assume the business must be bad. When it rises, they assume it must go higher and that they should get in.
His advice was physical. Look at the jar, write the number on paper, and trust that private judgment will land closer to true value than the noise of the room. He added that the advice is so hard to follow that, as Greenblatt put it, Warren Buffett says "you should buy low-cost index funds, even though he doesn't," and "I say you should buy low-cost index funds, even though I don't."
Questions people ask
Why does averaging independent guesses work so well?
High guesses and low guesses tend to cancel each other when people decide alone, leaving the middle close to the truth. The errors are random, so they wash out. Once people hear each other, the errors stop being random and lean one way.
Who is Joel Greenblatt?
He is an American value investor and professor known for a simple, numbers based stock picking formula he described in books for ordinary investors. His funds outperformed the market for decades. He often teaches with classroom demonstrations like the jelly bean jar.
Do index funds really beat most stock pickers?
Over long periods, yes. A low cost fund that owns the whole market keeps fees tiny and never makes big concentrated mistakes. Most professional funds charge more and trade more, and only a small minority beat the index over 10 or 20 years.