When I was in middle school, I devoured the Nancy Drew and Hardy Boys SuperMystery books. These were the crossover books that put Nancy, Frank, and Joe together to solve mysteries, which was pretty much irresistible to me at that age. I was interested in the mysteries, of course, but I was also extremely invested in the relationships. Honestly, I always thought Nancy should dump Ned and get together with Frank. Apparently, I wasn’t inventing that possibility entirely in my own head because the series did play with an attraction between Nancy and Frank, although nothing ultimately came of it. I had opinions about this.

As I got older, though, I never really found my way back to mysteries. I tried them occasionally as a teenager and an adult, but they didn’t grab me the way those books had when I was younger. For a long time, I assumed that mysteries were simply something I had outgrown. Then I started discovering Agatha Christie read-alikes and realized that I had been wrong about my own reading preferences. I didn’t dislike mysteries. I was just looking for a particular kind of mystery.

It turns out that I love a closed-circle mystery, where there is a limited group of suspects and the answer is somewhere inside that circle. I particularly enjoy amateur sleuths who are figuring things out as they go rather than brilliant detectives who seem to be ten steps ahead of everyone else. Give me a group of suspicious people, clues that were sitting in front of me the entire time, and a solution that makes me mentally revisit everything I just read, and I am happy. Some of my recent favorites include Pomona Afton Can So Solve a Murder by Bellamy Rose, The Agathas by Kathleen Glasgow and Liz Lawson, How to Solve Your Own Murder by Kristen Perrin, and Vera Wong’s Unsolicited Advice for Murderers by Jesse Q. Sutanto. Apparently, the middle-schooler who wanted to solve crimes alongside Nancy Drew never completely disappeared. She just became considerably more particular about her fictional murders.

The more of these books I read, however, the more I have started thinking about what makes this particular type of mystery so satisfying. Part of the pleasure is that we aren’t simply watching someone else solve the puzzle. We are solving it alongside them. The author gives us information, introduces possible suspects, reveals motives, and drops clues into the story while we quietly construct our own theories about what happened. Somewhere along the way, though, something interesting begins to happen. Once we have decided who we think did it, we stop simply gathering clues and begin looking for evidence that proves us right.

That is usually when I get myself into trouble. I pick a suspect, construct what I consider an extremely convincing case, and then start interpreting everything that happens through the assumption that I am right. A suspicious glance becomes evidence. A lie becomes practically a confession. Something that should probably make me reconsider my theory instead gets creatively incorporated into it. By the time the actual solution is revealed, I sometimes have to admit that I wasn’t really solving the mystery anymore. I was building a case for the solution I had already chosen.

Psychologists have a name for this tendency: confirmation bias.

Step One: Pick a Suspect. Step Two: Prove Yourself Right.

Confirmation bias is our tendency to seek out, notice, interpret, and remember information in ways that support what we already believe. We don’t necessarily do this deliberately. Our brains are constantly trying to make sense of enormous amounts of information, and patterns and existing beliefs help us do that efficiently. The problem is that once we develop a theory, new information doesn’t always get evaluated on completely neutral ground. We are more likely to notice evidence that fits what we already think and find explanations for evidence that does not.

Mystery writers know exactly how to take advantage of this. In fact, some of the best mysteries depend on it. A satisfying mystery usually doesn’t work by hiding every important piece of information from the reader until the final chapter. If the detective suddenly reveals a murderer we have never met based on evidence we were never given, we don’t feel impressed; we feel cheated. The much more satisfying trick is for the writer to show us something important and allow us to interpret it incorrectly.

A character lies about where they were on the night of the murder, and suddenly we have our suspect. From that point forward, everything that person does can begin to look suspicious. Maybe they lied because they were committing murder, but perhaps they were hiding an affair, protecting someone else, concealing an embarrassing secret, or simply doing something they did not want anyone to know about. A character desperately needs money, and we see a motive. Someone becomes uncomfortable whenever the victim’s name comes up, and we decide that they must know something about the murder. They probably do know something, but the important question is whether they know the thing we have already decided they know.

Once I have selected my fictional murderer, I can become remarkably committed to the prosecution. If my suspect lies, that proves it. If they are nervous, that proves it. If they are unusually calm, that is obviously suspicious too. Even information that points somewhere else can sometimes be incorporated into the theory if I am sufficiently convinced that I have figured everything out. It is entertaining when the stakes involve beating a fictional sleuth to the solution. It becomes much more consequential when we realize that our brains can do exactly the same thing with actual people.

Our Brains Want the Story to Make Sense

Human beings are remarkably good at creating coherent stories from incomplete information. We have to be. We rarely have every possible piece of information before making decisions in everyday life, so our brains use previous experiences, patterns, expectations, and context to fill in the gaps. Most of the time, this is enormously useful. We could barely function if every situation required us to disregard everything we had previously learned and begin interpreting the world from scratch.

The difficulty is that a coherent explanation is not necessarily an accurate explanation. This is one of the things mystery novels exploit so effectively. The author gives us enough information to construct a plausible story about what happened, and once that story makes sense to us, we naturally become attached to it. New information then presents us with a choice. We can reconsider our theory, or we can reinterpret the new information so that it continues to fit what we already believe. Confirmation bias makes the second option remarkably tempting.

That is also why the reveal at the end of a well-constructed mystery can be so satisfying. The best reveals don’t simply tell us something we could never have known. Instead, they change the meaning of things we already knew. Suddenly, a conversation we dismissed becomes important. A behavior we interpreted one way has an entirely different explanation. The tiny detail we barely noticed turns out to matter enormously. The information itself has not changed, but our interpretation of it has, and suddenly we can reconstruct the entire story around a different explanation.

There is something useful in that experience beyond the pleasure of discovering who committed a fictional murder. It reminds us that there is often a layer of interpretation between what we observe and what we believe happened.

We Do This to People Who Aren’t Murder Suspects

Imagine that you have decided a coworker is lazy. Once that idea takes hold, you may start noticing evidence that supports it. They arrive late to a meeting, don’t volunteer for a project, and take longer than expected to answer an email. Each incident becomes another clue that seems to confirm what you already know about them.

What may be less noticeable are the pieces of information that do not fit the story. Perhaps they arrived early to every meeting for the previous month. Maybe they didn’t volunteer because they are already carrying a workload you cannot see. The unanswered email may have arrived while they were dealing with a problem you knew nothing about. None of those explanations necessarily means your original assessment was wrong, but confirmation bias can make us less likely to investigate them once we believe we already understand what is happening.

The same thing can happen with almost any label we give someone: difficult, negative, unreliable, resistant to change, not a team player, bad with technology, unmotivated. Once the label becomes part of our mental story about the person, their behavior can begin to look like a series of clues proving that our original judgment was correct. Eventually, we may stop asking what a particular behavior means because we believe we already know what kind of person would behave that way.

This does not mean that every negative impression is the result of bias or that we should ignore genuine patterns of problematic behavior. Sometimes a person repeatedly behaves in ways that support exactly the conclusion we have reached. The danger comes when we stop treating our conclusion as something that can be tested against new information. At that point, we become the mystery reader who decided on page fifty that the nephew did it and spends the next two hundred pages turning every raised eyebrow, missing teacup, and awkward conversation into further proof of his guilt.

A Good Sleuth Has to Be Willing to Be Wrong

Perhaps this is why I particularly enjoy amateur sleuths who are figuring things out as they go. They don’t necessarily walk into the story already knowing how all the pieces fit together. They develop theories, pursue leads, misunderstand clues, uncover new information, and sometimes realize that something they were certain about was completely wrong. Their ability to solve the mystery ultimately depends less on getting everything right immediately than on remaining curious enough to keep investigating.

That is a useful way to think about confirmation bias because the goal cannot realistically be to stop forming opinions. Human beings need hypotheses about what is happening around us. We need to recognize patterns, make judgments, and act without possessing perfect information. The challenge is remembering that a hypothesis is not the same thing as a verdict.

A good investigator has to remain interested in the evidence that does not fit. We can practice doing the same thing by asking ourselves what would have to be true for our theory to be wrong, whether we are giving some information more weight simply because it supports what we already believe, and whether there are other reasonable explanations for the behavior we are observing. We can even ask whether we would interpret the same evidence differently if we had started with a different suspect.

Those questions become particularly important when our “suspect” is another person. If we have already decided that someone is difficult, for example, are we interpreting a question from them as a challenge when we would interpret the exact same question from someone else as curiosity? If we believe an employee is resistant to change, do we hear their concerns as evidence of resistance rather than information about a problem we may have overlooked? If we think someone is unreliable, do we remember every missed deadline while barely registering all the times they delivered exactly when they said they would?

In a murder mystery, getting this wrong might mean accusing the wrong dinner guest for a few chapters. In real life, our conclusions can shape how we communicate with someone, which opportunities we offer them, how much we trust them, and even what behavior we elicit from them in return. Once that happens, confirmation bias can become more than a problem with how we interpret evidence. Our expectations can begin influencing the very relationship we believe we are objectively observing.

The Fun of Finding Out You Were Wrong

There is a particular pleasure in reaching the end of a good mystery and discovering that I was completely wrong. I may need a moment to maintain that my solution would have made an excellent alternate ending, but then I want to know how the author fooled me. I mentally retrace the story, looking for the clues I overlooked and the assumptions that sent me in the wrong direction. When the mystery is particularly good, I realize that the answer was available to me all along. I simply couldn’t see it because I had already decided what the clues meant.

Being wrong in real life rarely feels quite as entertaining. We become attached to our interpretations, particularly when they involve other people, and changing our minds can feel like admitting that our judgment failed. Yet changing a conclusion when new evidence warrants it is not a failure of judgment. It is exactly what good thinking requires. The problem is not that our first theory might be wrong. The problem is becoming so invested in proving ourselves right that we stop genuinely considering other possibilities.

Perhaps that is one reason I have fallen back in love with mysteries after all these years. I enjoy the closed circles, eccentric suspects, amateur detectives, suspicious alibis, hidden motives, and inevitable moment when the pieces finally come together. I still want the satisfaction of figuring everything out before the sleuth does, and I am delighted on those rare occasions when I actually manage it. At the same time, these stories give us a safe and entertaining place to practice something that is considerably harder outside the pages of a book: developing a theory, testing it against new information, and being willing to let it go when the evidence points somewhere else.

So I will continue accusing fictional neighbors, relatives, coworkers, dinner guests, and suspiciously charming strangers of murder, often with embarrassingly little evidence. I will continue changing suspects halfway through the book, overlooking clues that should have been obvious, and occasionally reaching the final reveal only to discover that I have been spectacularly wrong. There are worse ways to practice being willing to reconsider what we think we know.

As for my long-held belief that Nancy Drew should have dumped Ned Nickerson and chosen Frank Hardy instead? After an entire article about confirmation bias, intellectual consistency probably requires me to admit that I should at least be willing to reconsider that conclusion.

I am not. Nancy and Frank forever.