Katy Milkman: Here’s a scenario. Two friends are watching the end of a Major League Baseball game on TV.
Announcer: Bottom of the ninth. Two out. Candelario is at the plate. He had an infield hit; that was in the first inning. The count is 3 and 2. Big pressure on Cardinals’ pitcher Jordan Hicks.
Speaker 3: Come on! Yes!
Announcer: And he lifts a fly ball to deep left.
Speaker 3: Oh come on, come on, come on, come on!
Announcer: Ozuna back at the wall. And it’s gone!
Speaker 3: Yes! Woo, yes!
Announcer: Tigers win over the Cardinals 5-3 with a walk-off here in the bottom of the ninth.
Speaker 3: Oh, I knew it. I knew he’d pull it off. Candelario is my guy. He has been hitting so well lately. I could just feel it.
Speaker 4: He’s the man!
Katy Milkman: OK, here’s the same scenario again but with a different outcome.
Announcer: Bottom of the ninth, two out. Candelario is at the plate. He had an infield hit; that was in the first inning. The count is three and two. Big pressure on Cardinal’s pitcher Jordan Hicks.
Speaker 3: Come on!
Announcer: Strike three and the Cardinal’s win it 3-2 over the Detroit Tigers.
Speaker 3: Ah, I had a feeling Candelario would choke. He’s a good hitter, but he cannot touch Hicks’s slider.
Speaker 4: Yeah, man, I hear you.
Katy Milkman: You’ve probably felt that way at some point after watching a game. When it was over, it’s almost as if you could see how it would play out all along. You had a feeling that it would turn out a certain way. Or did you? Today we’ll talk about a bias that toys with the way you perceive past events.
I’m Katy Milkman, and this is Choiceology, an original podcast from Charles Schwab. It’s a show about decisions and the impact those decisions have on our lives. It’s also a show about subtle but systematic mistakes in reasoning that can push people in one direction or another, often without us even realizing it. We try to give you some tools to fight back against those psychological forces and to help you avoid costly errors.
Speaker 5: France has made every preparation on the chance that the inhuman ambition of the ruler of Nazi Germany would rush blindly to catastrophe. And France is ready, too.
Katy Milkman: It’s September 1st, 1939. German forces have invaded Poland. And on September 3rd, Britain and France declare war on Germany. World War II has begun.
Speaker 6: Through Luxembourg, Belgium and Holland five German armies fanned out across France. Treachery and incompetence had doomed a nation that, only a decade ago, had been leader of Europe.
Julian Jackson: The basic story is that the French and British declare war on Germany in September 1939. And obviously, the Germans in that period overrun Denmark and they overrun Norway. And then, on the 10th of May, the Germans launched their offensive in the west against Belgium, Holland, Luxembourg, and of course most importantly, France.
Katy Milkman: This is Julian Jackson.
Julian Jackson: Professor of modern French history at Queen Mary University of London.
Katy Milkman: On May 10th, 1940, more than five months after the declaration of war, German armored units surprised the French by invading the country through the mountainous Ardennes region. This approach cut off and surrounded the French and Allied units that had been deployed to Belgium, where the French thought the Germans would focus their attack.
Julian Jackson: The Germans attacked across the river Meuse in France. And they crossed the river with their nine Panzer divisions. And although the fighting went on for another few more weeks, it was really all over then.
Katy Milkman: On June 5th, the remaining French and British divisions tried but failed to push back the Germans. By the time they realized that they’d misjudged the location and speed of the main German assault, it was too late.
Julian Jackson: The Germans were across. And then every attempt by the French to counterattack failed.
Katy Milkman: On June 14th, the Germans entered Paris unopposed. German commanders met with French officials on June 18th to negotiate an end to hostilities. The pro-German Vichy government was established. France had surrendered less than six weeks from the start of the invasion.
Julian Jackson: So the amazing thing about it is the extraordinary rapidity with which it happened. Roosevelt was very, very shocked by the speed of the French collapse. And the British, too, felt they’d been let down by the French. And so there was a quite strong feeling that the French were lacking in moral fiber, they were unwilling to fight, and so on.
Katy Milkman: So what explains the seemingly tepid resistance and the rapid fall of one of the great European powers?
Julian Jackson: The classic text, the one that’s influenced everybody ever since, was by a French historian, a famous French historian, Marc Bloch, who wrote this rather remarkable book, which is called Strange Defeat, in the summer of 1940. Pondering as a piece of instant history, because he was a historian, why this dramatic event had happened. That in six weeks what many people thought was the greatest army in Europe, that had held out against the Germans for four years in the first war, collapsed in six weeks. So there is, ever since that classic book, there’s a kind of idea that there was something inevitable about what happened in 1940. That France was in some kind of way rotten, that her political system was dysfunctional. And so the defeat was something about the whole moral defeat of a society. And another element of that, a pacifist society, a society of people who didn’t want to fight, who would prefer giving up to fighting. That it was written in the stars.
Katy Milkman: But was the fall of France actually inevitable? Historians have offered lots of takes suggesting that the answer is, “Yes, it was.” Some historians argued that the French couldn’t compete with a modernized German army. Some argue that the Germans were unified, politically and militarily, while the French had quarrelsome generals and undisciplined soldiers. But these were arguments made after the fact. If you follow the trajectory of the war before fighting broke out in France, the inevitability of defeat seems much less obvious. In fact, there was one event that occurred months before the invasion that had the potential to shift the advantage to the French before any shots were fired. It was called the Mechelen Incident.
Douglas Porch: Hi, my name is Douglas Porch. I am a military historian by profession.
Katy Milkman: We’ve asked Douglas Porch to walk us through the Mechelen Incident and its repercussions.
Douglas Porch: The Mechelen Incident occurred on the 10th of January, 1940. There was a Luftwaffe reconnaissance plane that was flying a Major Helmuth Reinberger, who was a logistics officer.
Katy Milkman: Reinberger was traveling to Cologne, Germany, carrying with him top-secret plans for a German invasion of France. During the flight, they ran into poor weather conditions.
Douglas Porch: The weather is getting worse and worse. They get in the clouds. It’s snowing. They get lost. And it’s debated whether they ran out of gas or whether the pilot inadvertently hit the switch that cut the gas off, but all of a sudden they start descending and they crash-land in a field. They have no idea where they are, so they get out, they find a local farm boy, and he tells them they’re in Belgium.
Katy Milkman: They’d crash-landed near the Belgian town of Mechelen-sur-Meuse.
Douglas Porch: Well, they panic. Because Reinberger has got the orders, the operations orders for his division, for this attack that’s supposed to be launched in a week. So he runs back to the plane and he takes his briefcase. And he goes behind a hedge and he pulls out his lighter and tries to burn the orders. But his lighter malfunctions. So he actually goes back to the farm boy and says, “Hey, do you have a light?” And so the farm boy gives him a match. And so he tries again, without much success. And then, all of a sudden, there are two customs officer that sort of pedal up on bicycles. And so they take him to the police station and then they go off to make a phone call and find out, “What do we do with these two German officers who’ve landed in Belgium?” So what happens is that they leave these two German officers momentarily. They come back, and there’s Reinberger stuffing the pages into a potbellied stove. So they grab the pages from him and, obviously, they’re going to keep them. But they’re charred, half-charred.
Katy Milkman: The plans had been damaged, but valuable information survived. The Belgian customs officers phoned their superiors in Brussels to ask what they should do with these partially readable plans. They attempted to mislead the German officers.
Douglas Porch: They basically tell the two German officers that the document’s unreadable and we couldn’t make heads or tails out of them. Which, in fact, is not true at all. They’re letting the Germans know that they don’t really have this plan. Whereas, in fact, they can recover enough of it that they take a two-page summary, and they give it to the French attaché in Brussels. So he has the basic outline of the plan.
Katy Milkman: The French High Command now had in their possession a secret German plan for the invasion of France. It’s hard to imagine more valuable intelligence. The plan was codenamed Fall Gelb, or Case Yellow.
Douglas Porch: Fall Gelb was the German plan that was developed in October 1939 for the invasion of France through Belgium. So what it showed was that the German attack wouldn’t be just through Belgium and Holland. There would also be part of it that would go through the Ardennes.
Katy Milkman: That’s the rugged mountainous region I mentioned earlier. The Ardennes, unlike other potential land routes from Germany to France wasn’t heavily fortified with obstacles and artillery. That said, it was still a treacherous route.
Douglas Porch: It was heavily forested. It had a very small road network of small, twisty, windy roads.
Katy Milkman: Maurice Gamelin was the commander in chief of the French army. He studied the German plan, and he had to decide how to respond. He had to guess whether the Germans would follow through with their leaked plan or change it.
In the end, Gamelin opted to prepare for a German invasion along the Dyle River in Belgium rather than move men and resources to the Ardennes. This would turn out to be a major strategic error.
Since Hitler had indeed suspected that his original plan had been compromised during the Mechelen Incident, he made a major change. The new German plan was to pretend they were going to attack in Belgium and draw the Allies there while now focusing their main attack through the less fortified Ardennes region.
This seems like a good plan in retrospect, but at the time, it too was a gamble.
Douglas Porch: It was a huge gamble because they had to put about, I think it’s 141,000 men, in all of their tanks, all of their motorized infantry, on four small roads through the heavily wooded Eifel Mountains of the Ardennes. They had to concentrate and move quickly before they were seen by French air reconnaissance.
They had to move at night in black-out conditions, driving for, in many cases, four nights, which meant that these soldiers had to stay alert. It required enormous road discipline to avoid traffic jams, to avoid accidents. It also required incredible logistics.
Katy Milkman: It was risky, but in the end, the German plan worked brilliantly, and the French were unable to recover. Historians debate the importance of the Mechelen Incident since it didn’t help the French thwart the German invasion. But as a thought experiment, imagine some of the other ways this could’ve played out.
What if the plans had been captured without the Germans ever knowing? Or what if the Germans had canceled their plans to invade through the Ardennes and stuck with the more obvious route through Belgium? Or what if Gamelin had opted to defend the Ardennes?
Things might have turned out very differently. It’s easy to hear about the pivotal mistakes made by the French military in the early days of the war and then to judge these mistakes harshly. But we have the benefit of hindsight.
If you were a German or French general at the time, none of these decisions would’ve been clear.
Douglas Porch: What was obvious about the German invasion is that none of it was obvious. Why would it be obvious that you would go through the Ardennes and take that sort of operational risk? And could pull it off?
Katy Milkman: In fact, the German high command was extremely nervous about the plan to attack France through the Ardennes. In several instances, they had tried to talk Hitler out of the plan. You can see how there were numerous opportunities for decisions that would’ve given France an advantage or at least a fighting chance. Again, only in retrospect do the events seem inevitable.
Douglas Porch: People have a tendency to see the fall of France as inevitable because they read history backwards. They go from the result and look for a cause, and that cause must be momentous because the result was so momentous, and they attribute the fall of France not to tactical and operational failures, but in fact to a weakness of a society, the political divisions of a society, the defeatism of society.
It served the purpose of certain people to see it that way. The fall of France wasn’t inevitable. It required a chain of events, beginning with the fact that Hitler chooses the weakest point to attack, the fact that the French didn’t bolster that point, the fact that the German military were able to sort of pull off a very difficult maneuver.
All of that went into the formula. The fog of war is that things simply aren’t clear. I mean, it’s one thing to look in retrospect and say, well, one should’ve done that. But you have to see it the way people did at the time.
Katy Milkman: Douglas Porch is a military historian and retired professor of national security affairs at the Naval Postgraduate School in Monterey, California. Julian Jackson is a professor of modern French history at Queen Mary University of London.
We’re sharing this story because it helps to demonstrate a common error in the way we tend to think about past events. Some call it the historian’s fallacy—where key moments in history, like the fall of France, come to be seen as inevitable, when in fact, they were anything but.
In psychology and behavioral economics, this tendency is called hindsight bias. That “I knew it all along” sense you get when an outcome arises that you’re sure you could’ve predicted. It was so obvious it was coming.
The thing is, when you don’t know the outcome, it’s often very hard to predict the precise things that seem obvious in hindsight. Let me give you another example.
We took a piece of recorded speech and electronically distorted it so it’d be difficult to understand. Have a listen.
Speaker 9: [garbled speech]
OK. Here’s the unaltered version.
Speaker 9: It is easier to recognize other people’s mistakes than our own.
Katy Milkman: As an experiment, we presented the unaltered version first to two people.
Speaker 10: I’ll get you to put on these headphones to start. What I’m going to do here is I’m going to play two clips for you and then just get you to comment on what you heard. Here is the first clip.
Speaker 9: It is easier to recognize other people’s mistakes than our own.
Speaker 11: OK. Got it. Makes sense.
Speaker 12: Yeah, that’s really clear.
Katy Milkman: Then we played the distorted version to see what they could understand.
Speaker 11: It sounds muffled, but I can hear it.
Speaker 12: It’s pretty garbled but yeah, I can totally understand it.
Speaker 10: OK, so now I want you to predict what percentage of people who hear only the garbled version will be able to understand it.
Speaker 12: I don’t know; I’d say like 30%.
Speaker 11: Maybe half? 50%.
Katy Milkman: Now we’re only going to play the distorted version for some other folks to see how many of them can decipher the words.
Speaker 9: [garbled speech]
Speaker 13: It sounds like audio is cutting out on a Skype call. It’s all cutting out. I can’t understand it.
Speaker 14: Something kinda blurry. It sounded like … [garbled speech]
Speaker 13: It was really unclear and kind of robotic sounding.
Speaker 10: Were you able to make out any of the sounds, any ideas what that was?
Speaker 14: A man talking? It sounded like maybe it was asking a question, but I could not really fully understand what he was saying.
Katy Milkman: Our original set of participants, the ones who heard the clear audio first, guessed that somewhere between a third and a half of listeners would be able to understand the audio. But in fact, almost no one we played the garbled clip for without a preview could decipher what was said.
This experiment demonstrates how hindsight bias affects our perception. When the participants could only hear the muffled audio, they had a really hard time figuring out what the content was in the distorted clip.
But the participants who were familiar with the contents of the audio clip ahead of time, once they knew what to expect, they could easily hear the words in the garbled version. In hindsight, it seemed the muffled clip wasn’t so hard to decipher.
That’s the hindsight bias. That something you would’ve found confusing or unknowable in foresight suddenly seems obvious in hindsight. It’s closely related to another bias called the curse of knowledge, where once we have information, we’re unable to accurately recall or simulate the experience of being naïve.
That is, we can’t accurately remember what it was like before the information was available to us. For example, as a professor explaining something to my students, I often suffer from the curse of knowledge.
The knowledge seems obvious to me now, and when I try to explain it, sometimes I fail because I fail at perspective taking. This inability to recall what it was like before you had certain information is part of what fuels hindsight bias.
Hindsight bias has presumably been around as long as people have been reflecting on the past. But the bias didn’t really make it into the academic literature until the 1970s. It was 1975 when Baruch Fischhoff highlighted the bias in a series of important experiments.
The experiments showed that telling someone an outcome leads to what Fischhoff called “an unjustified increase in its perceived predictability.” To get into the latest research on hindsight bias, I’ve invited Kathleen Vohs to join me.
Kathleen is a professor at the Carlson School of Management at the University of Minnesota. Hi, Kathleen. Thank you so much for being here.
Kathleen Vohs: Hi, Katy! I’m happy to be here.
Katy Milkman: Today, we’re going to talk about hindsight bias. What’s the easiest way for people to understand what hindsight bias is?
Kathleen Vohs: Hindsight bias, I think, is intuitive because I think a lot of us experience this a lot in our lives and we see other people do it. But briefly, it’s the phenomenon where once you know about an event, what happened, how it happened, then you say, “Oh, of course. I could’ve totally predicted that would’ve happened or it would’ve gone that way.”
It’s this idea that your evaluation and your predicting of the likelihood of an event is altered by knowing how the event played out, or whether it happened or not, when in actuality, you probably would not have been able to know that that was going to happen either at all or the way that it did.
What constitutes the bias is that there’s some error that creeps in once people know about an outcome that they think they would’ve been able to foresee, but they don’t.
Katy Milkman: So Kathleen, why do you think we have this hindsight bias? Is it somehow giving us some sort of evolutionary advantage, like it’s adaptive? Or is it just purely a bad thing?
Kathleen Vohs: The adaptive component of a hindsight bias is that it helps us to make sense of the world. It’s a very threatening idea that when we learn about events and we learn about information, that we would be caught off guard all the time, that we would never be able to make sense of something. So the adaptive part of the hindsight bias comes from what’s commonly called sense-making.
Katy Milkman: So the sense making idea that we want the world to feel orderly, and so when we get from point A to point B, we say, “I see. That was a straight path. It all makes sense now.” And we create this narrative that the path from A to B was inevitable and that there were no other twists and turns we could have taken along the way.
Kathleen Vohs: We build a narrative and we selectively choose different details to make sense of the event now that we know about it.
Katy Milkman: OK. Let’s talk a little bit about research. Could you describe some of your own research on hindsight bias?
Kathleen Vohs: My research on hindsight bias is really aimed at trying to take a broader look at that literature and trying to distill it into what causes it, what are its outcomes, what are the contexts in which we see it. So let me kind of tell you at a broad brush stroke how it’s normally done. There’s kind of two approaches to studying hindsight bias. One uses the same people throughout the study, where you ask them about an event prior to it having happened. And so for example, at the University of Minnesota, where I’m a professor, we might ask the likelihood that our football team, the Gophers, are going to win their football game this weekend. And then people rate the likelihood, and then a week goes by, and then we ask them later, “Looking back before you knew about the way the game played out, what do you think the odds were that it was going to go that way?”
So let’s say that the Gophers win. So we would say, “What do you think the odds were looking back that they were going to win?” And frequently what you’ll see is that people don’t recall necessarily their earlier estimate, but what they do now is make their estimate kind of become in line with the actual event as it played out. So if the Gophers won their game, but initially they’d said it was about 50/50 that they may win, then knowing about the outcome moves their probability judgements to be like 80%. I would’ve said there’s like an 80% likelihood that they would have won. But in matter of fact, we know that they said 50%, for example. And so that’s kind of one way to capture it.
And another way is to give two groups of participants information about an event, and one group gets partial information, kind of the general setup of the event, and then is asked to predict the likelihood of various different outcomes. And another group gets the entirety of the event, exactly how it was set up and then how it actually played out at the end. And then they too are asked, “If you didn’t know what you know now, what are the odds that it would have come out this way?” And what you see time and time again is that that group that knows about the outcome, they rate the odds of the outcome happening were just a lot higher than the group that has no information about what was actually going to happen.
Katy Milkman: Do you have any favorite studies on hindsight bias?
Kathleen Vohs: What I like about the hindsight bias literature is that it’s really steeped in all these different contexts. So there’s financial decisions, and there’s a lot about medical decision making. A lot of it is in a legal setting. A favorite paper of mine used professional judges, and they were trying to look at not only the hindsight bias but also this issue of negligence, which is essentially, like, the defendant—could or should they have known that the outcome was going to occur? And in this study they give half of the judges just the setup. That there is a patient at a psychiatric ward who after many months of good behavior gets released and takes a three-hour release with his new girlfriend. And then the other half of the judges get that same information, and then they learn that the man and his girlfriend have an argument. He goes off, goes on a big robbery and crime spree, and ends up killing an 80-year-old woman. And then both groups of judges now rate things like that outcome and how likely it was, also various other outcomes, how likely they are. And then knowing that the senior psychiatric staff is on trial now, how much should they be blamed for having let him out in the first place.
And judges show the hindsight bias. The group that knew about the outcome say it was foreseeable, and they place much more blame onto that psychiatric staff and say that they should have known and that they are criminally negligent. In this case, they should have seen the harm that was going to occur. Now, this study was based on a real legal case, so it’s a very … it’s just a rich and fascinating area, and it kind of goes to another broader point that you would think that experts in an arena would be less likely to make hindsight bias judgments, but that’s not the case. Generally they show them just as much as novices.
Katy Milkman: So interesting. And I always find the legal examples fascinating, in part because our legal system sometimes has these rules where you’ll instruct a jury, for instance, to ignore some information that was shared with them because it was actually inadmissible, and anyone who knows anything about the hindsight bias knows that that’s a completely preposterous set of instructions and that as soon as it’s presented to them, it’s going to weigh in on their judgments.
Kathleen Vohs: Exactly. The legal system and historical context are two situations in which everybody is interpreting something after the outcome is known. And so it’s just a very complex and also kind of very rich setting for us to observe hindsight bias.
Katy Milkman: Yeah. Do you have any advice to people on how they can avoid falling prey to hindsight bias or how we can reduce it, at least in our own judgments?
Kathleen Vohs: A lot of times hindsight bias not only involves thinking about the outcome but also thinking about how it occurred, some sort of causal factor. And so one way to kind of reduce hindsight bias is to get people to think about their favorite causal candidate: “It happened because of this reason,” but then also to start to consider lots of other potential causes for that same outcome. Because in reality, almost all complex social outcomes are due to a multitude of different factors, including randomness, including things that are mechanical, other people’s actions, etc. And the more that people center on one causal role, the more that they’ve just over-weighed its ability to have produced the outcome, and that can cause a lot of undue harm—as in the legal examples that we’ve already mentioned
And so when you get people to break down—there are lots of different ways that this same outcome could have occurred, through different pathways—that can help. And the other way it is to think about all the different ways that the outcome could have gone another way. So it would have been different in degree, or different in kind, or wouldn’t have happened at all, or it would have been delayed in time or pushed forward in time. And so when you get people to map out basically a host of alternate realities, that can help them to feel less confident that they knew it all along, which is the hallmark of the hindsight bias.
Katy Milkman: Kathleen, this was great. I really appreciate you being here. Thanks for taking the time.
Kathleen Vohs: Oh, this was a pleasure. It was so much fun. Thank you for having me.
Katy Milkman: Kathleen Vohs is a professor of marketing at the Carlson School of Management at the University of Minnesota.
Market events are notorious examples of hindsight bias. Lots of people will tell you that they saw the financial crisis and the Great Recession coming, or the bursting of the dot-com bubble. In reality, if forthcoming trouble were so easy to spot, it likely would have been avoided in the first place. And trying to time the markets is nearly impossible. That’s why Schwab also produces the podcast Financial Decoder. It’s designed for people who want to make better decisions with their money. Host Mark Riepe and his guests dissect the financial choices you might be facing, like how much risk you should take or whether your portfolio could be better diversified, and they offer tips to mitigate the impact of biases on your financial life. You can find it at Schwab.com/FinancialDecoder or wherever you listen to podcasts.
Reducing the effect of hindsight bias isn’t easy. As Kathleen Vohs mentioned, the bias makes it incredibly difficult to assess the likelihood of a given outcome once you know the actual outcome. By then it can seem inevitable. If you find yourself dealing with coworkers or managers who are regularly clever after the fact, you might find that waiting to announce the outcome of some decision until after they’ve given their own estimate or prediction will give you more balanced information.
Hindsight bias is revisionary, so it’s a great idea to keep notes, the way Kathleen kept track of how students judge the likelihood of a win for their football team. Having your judgments and predictions written down keeps you more honest once you know the results and will help you get a more accurate sense of your ability to make decent forecasts. Finally, like so many of the biases we discuss on Choiceology, it’s really useful to consider the opposite. Your favorite team might have lost the game, and the coach might have rationalized the loss after the fact, but how much of the outcome was simply luck? Or were there other factors like weather or location that might’ve had an impact? In short, while hindsight makes us feel our vision is 20/20, foresight is rarely that clear.
You’ve been listening to Choiceology, an original podcast from Charles Schwab. If you’ve enjoyed the show, leave us a review on Apple Podcasts, and while you’re there, you can subscribe for free. Same goes for other podcasting apps. Next time on Choiceology, we’ve got a story about a reality TV show model, and we’ll talk about a mental shortcut that can lead to errors in the way we perceive people, products and much, much more. I’ll also be participating at a special Schwab Live event on May 16th. You can tune in to watch and even join me for a live Q&A at Schwab.com/LivePodcast. I’m Katy Milkman. Talk to you next time.
Speaker 15: For important disclosures, see the show notes or visit schwab.com/podcast.