In a reading room at Yale, an Australian psychologist named Gina Perry put on headphones and pressed play on tapes that had been sitting in the university’s archives for forty years.
They were the original recordings of the most famous experiment in psychology. The one where a man in a lab coat tells an ordinary person to electrocute a stranger, and the ordinary person does it. The experiment every textbook cites. The one that supposedly proves authority can flip a switch in your brain.
What Perry heard on those tapes was not what the textbooks said. And although she could not have known it, what she heard would quietly knock a leg out from under a document written years earlier, three thousand miles away, by a man who had never run an experiment in his life. A document that investors treat the way medieval monks treated scripture.
We are going to tell you what was on the tapes. First you need to meet the scripture.
The lawyer who graded the psychologists
Cambridge, Massachusetts, 1995. A seventy-one-year-old securities lawyer from Omaha stands up at Harvard and tells the room he has identified twenty-five standard ways the human mind fails.◆
He has no psychology degree. He has never published a paper. He built the list the way a trial lawyer builds anything: fifty years of watching people lose money, lose lawsuits and lose their nerve, and taking notes. His own summary of his method runs twelve words: “I came to this after a life in the law and a life in business.”
The lawyer is Charlie Munger, Warren Buffett’s partner, the quiet half of Berkshire Hathaway. The talk becomes known as The Psychology of Human Misjudgment, and over the next three decades it turns into the closest thing investing has to a sacred text. Fund managers memorize the tendencies by number. The transcript circulates like contraband. When Munger died in November 2023, a month short of his hundredth birthday, half the eulogies quoted it.
Here is what none of the eulogies mentioned. We went looking, through three decades of tributes, summaries and reading guides, for a single one that had checked all twenty-five tendencies against the science. We found none.
So we did it ourselves. And two things surfaced that change how you should read every page of this book. The first is what Gina Perry found in those Yale archives. The second is stranger: something enormous is missing from Munger’s list. It hid in plain sight for thirty years. You know its name.
We will get to both. But you need the good news first, because the good news is almost unbelievable.
The miracle
Put yourself in 1995. Behavioral economics is not a phrase civilians use. Daniel Kahneman’s Nobel Prize is seven years in the future. Thinking, Fast and Slow, the book that taught the world to say “cognitive bias” at dinner, will not exist for another sixteen years. The research lives in journals that working investors have never opened.
And a lawyer from Omaha stands up and describes loss aversion.
He calls it “Deprival-Superreaction Tendency,” a name he invented, and defines it as the mind’s violent overreaction to losing things compared with gaining them. That is the finding at the center of the work that won Kahneman the Nobel. Munger described it accurately, from pure observation, under his own homemade label, before most economists had accepted it.◆
He does it again with availability: the mind overweights whatever is vivid, recent and easy to recall. Robust then, robust now, one of the few effects that sailed through every replication war psychology has fought since.◆ Again with contrast effects. Again with Pavlovian association. Again with incentive-caused bias, which is economics with the gloves off.◆ ◆
Run the full audit and roughly sixty percent of the list holds up against three decades of research the author never read. For a catalogue reverse-engineered from courtrooms and boardrooms, that number is quietly astonishing. Business books written last year, with full bibliographies and research assistants, routinely do worse.
If the story ended there, the worship would be justified. The story does not end there. Because about a quarter of Charlie’s list was not built from his own observations. It was built on borrowed evidence: the famous experiments of mid-century psychology.
And that brings us back to the tapes.
What was on the tapes
Press play with her.
The experiment on the tapes is Stanley Milgram’s obedience study, the spine of Munger’s “Authority-Misinfluence Tendency.” The tendency says humans follow authority figures even into madness. Milgram’s proof: ordinary people delivered what they believed were dangerous electric shocks, because a man in a gray lab coat told them to.◆
By the rules Milgram published, the man in the coat was allowed exactly four prompts, delivered flat, in order. “Please continue.” “The experiment requires that you continue.” “It is absolutely essential that you continue.” “You have no other choice, you must go on.” Four sentences. If the subject still refused, the session ended, and the refusal was recorded as disobedience.
That is the script every textbook quotes. It is not what Perry heard.
On tape after tape, the experimenter leaves the script and keeps going. He pushes. He improvises. He bargains with subjects the script required him to release, in some sessions coming back eight and nine times at a single wavering person. Perry sat with her headphones and counted what the legend never counted: a salesman closing, not a scientist observing. In the files around the tapes she found more. Participants who told the lab afterwards they had suspected the shocks were fake. And the number that conquered the textbooks, sixty-five percent obedience, traced back to one variation of the study, while other variations with very different results stayed in the drawer.
Authority influences people. Your own boss can confirm it. But the clean parable, the switch-flip in the brain that Munger built his tendency on, does not survive its own filing cabinet.
The same era of re-examination came for the Asch conformity studies, the engine inside Munger’s “Social-Proof Tendency.” The legend: people deny the evidence of their own eyes rather than disagree with the group. The data, re-read: most subjects, most of the time, did not conform. There is a tug toward the herd. It is real, and it is nothing like a puppet string.◆ ◆
Score it honestly and about a quarter of the list stands on evidence that has been downgraded since 1995. Not disproven. Downgraded. In our verdict language these tendencies are wounded and walking: real forces, weaker than advertised, resting on receipts that rusted.
Which raises an uncomfortable question. Munger’s whole method was to trust fifty years of direct observation over academic fashion. The one time he borrowed academic armor instead of forging his own, the armor failed. His system worked. His trust in their system didn’t.
And yet none of this is the strangest thing we found. The strangest thing is not on the list at all.
The dog that didn’t bark
Do something for us. Name the most famous cognitive bias in the world. The one everyone knows. The tendency to hunt for evidence that agrees with you and look straight past evidence that doesn’t.
Confirmation bias.
Now find it on Munger’s list of twenty-five ways the mind fails.
Take your time. We did. It is not there. Neither is anchoring, arguably the most replicated effect in the entire literature, the arbitrary first number that drags every later estimate toward it. Neither is the planning fallacy, the reason every project you have ever run came in late. The most celebrated bias catalogue in business omits the three most bulletproof biases science has ever measured. We counted.◆
How does that happen? The method explains it. Courtrooms show you incentive effects in neon. Boardrooms perform social proof daily. But confirmation bias is the one malfunction you cannot observe in others without committing it yourself, because every example you collect confirms the theory of human folly you already hold. A list built from observation was always going to miss the bias that corrupts observation. The instrument cannot photograph its own lens.
Now widen the frame, because this explains something bigger than the list. It explains the worship. For thirty years, thousands of intelligent, skeptical, professionally paranoid investors read this list, nodded, and never checked it. Checking scripture requires suspecting scripture. Every reader who loved the list went looking for cases where it explained the world, found them everywhere, and called that verification.
The Psychology of Human Misjudgment was protected for three decades by the exact bias it forgot to include. Charlie would have appreciated that. It may be the purest demonstration of misjudgment psychology ever run, and it was run on his own congregation.
The suit with three bodies inside
So what is this famous chapter, really? The audit says it is three documents wearing one suit.
Document one: verified psychology, sixty percent, described by a man who beat the academy to its own findings. Document two: the wounded quarter, tendencies leaning on Milgram and Asch and mid-century evidence that cracked under replication. Document three: the remainder, which was never science at all. Entries like “Twaddle Tendency” are things Charlie noticed and capitalized. Some of the noticings are excellent. “Persian Messenger Syndrome,” the observation that organizations shoot bearers of bad news until nobody brings bad news and the organization dies of surprise, is real wisdom; we have watched a version of it in every corporate collapse we have autopsied. But wisdom is not a finding, and the book never marks the difference.◆
Even the list’s crown jewel has this problem. Munger declares incentive-caused bias “the most powerful” of all the tendencies. Powerful it certainly is. But there is no experiment on Earth that fights twenty-five biases against each other and crowns a champion. That sentence is not a measurement. It is a closing argument delivered in a measurement’s voice, and after fifty years in courtrooms, nobody alive delivered closing arguments better.◆
That is the true flaw, and it is structural: the verified, the wounded and the folk wisdom all sit on one numbered list, in one confident voice, and the reader is never told which is which. The list has no immune system. Nobody ever gave it one.
The corrected playbook
Here is the list, sorted the way the book should have sorted it.
Trust with confidence. Loss aversion, availability, incentives, contrast, Pavlovian association. Replicated, armored, as useful in a deal memo as Charlie promised.
Hold loosely. Social proof and authority influence. Real tugs, not puppet strings, standing on downgraded evidence. Treat as tendencies, never as laws.
Enjoy as wisdom, cite as nothing. The named noticings, Persian Messenger included. First-rate aphorisms. Not findings.
Add what’s missing. Confirmation bias, anchoring, the planning fallacy. The best tribute to Munger’s method is fixing what the method could not see. Start with the bias that guarded his list for thirty years.
Chapter verdict: BS 4.2 out of 10. Mixed bag. We formally checked twelve claims from this chapter. Seven hold. Five are wounded. Not one failed outright, which tracks the roughly sixty percent survival rate across the full list of twenty-five. A 1995 field guide, built in a parking lot, that partially out-performed the academy, brought down only where it trusted the academy’s evidence instead of its own. Grade the man, not just the list: quietly magnificent, and overdue for the annotation you now hold.
One more thing before you turn the page. The eulogies all told you Munger was one of the greatest investors who ever lived. There is a number from 1973 and 1974 that every one of those eulogies left out. It is the size of the hole his own fund fell into, and it is larger than you are imagining.
Next chapter, we print it.