We all like to think we’re consistent. If you look at a file today and decide it’s a "yes," you’d like to believe you’d say "yes" tomorrow, too. But you probably won’t. That’s the messy, invisible reality of noise: a flaw in human judgment.
It’s different from bias. We talk about bias constantly—it's that systematic leaning in one direction, like a scale that always adds five pounds. Noise is more like a shaky hand. It’s the random variability that makes two doctors give different diagnoses for the same mole, or two judges give wildly different sentences for the exact same crime.
Daniel Kahneman, the Nobel laureate who basically invented the way we think about thinking, teamed up with Olivier Sibony and Cass Sunstein to put a name to this ghost in the machine. They realized that while bias is the star of the show in psychology, noise is the silent killer in business and law. It’s everywhere.
The Difference Between Being Wrong and Being Noisy
Imagine a target at a shooting range.
If all the shots hit the bottom-left corner, that’s bias. The sight is off. You can fix that easily because the error is predictable. But if the shots are scattered all over the paper like a handful of thrown gravel? That’s noise. You can’t just "adjust" for that because there’s no pattern.
Most of us assume that professional judgment is more stable than it actually is. We think experts are like calculators. In reality, they're more like weather vanes in a windstorm. In their book, Noise: A Flaw in Human Judgment, the authors highlight a study of federal judges. These were seasoned professionals. When asked to set sentences for identical hypothetical cases, the "noise" was staggering. In one case, the mean difference between two judges’ sentences was over three years.
Three years of someone’s life, decided by which judge happened to be on the bench that morning.
📖 Related: Destiny Tech 100 Stock Price: What Most People Get Wrong
Why your mood matters more than you think
It gets weirder. Occasion noise is a specific flavor of this problem where the same person changes their mind based on external factors that should have zero impact on the decision.
Are you hungry? Is your favorite football team losing? Did it rain this morning?
Researchers have found that physicians are significantly more likely to prescribe opioids at the end of a long day than at the start. They’re tired. Their "judgment" is being eroded by simple decision fatigue. This isn't a conscious choice to be a bad doctor. It's just the biological reality of being a human being. We are not the same person from hour to hour.
The High Cost of Unreliable Decisions
In the corporate world, this variability is a massive, uncounted expense.
Think about insurance underwriters. Their job is to look at a risk and put a price on it. If the price is too high, the company loses the contract. If it’s too low, they lose money on claims. You’d expect underwriters at the same firm to be within 5% or 10% of each other.
The reality? When Kahneman and his team conducted a "noise audit" at a major insurance company, the median difference was 55%.
The executives were floored. They had no idea.
🔗 Read more: Jamie Dimon Twin Brother: What Most People Get Wrong
This happens because companies often focus on "the right process" but rarely check for "the same result." If you have two recruiters interviewing the same candidate, and one sees a future CEO while the other sees a middle-manager, your hiring process is noisy. You aren't actually measuring the candidate; you're measuring the internal state of the recruiters.
Algorithms vs. Humans
This is where things get controversial.
Algorithms are almost always less noisy than humans. Even a very simple "linear model"—basically a basic formula—tends to outperform experts in predictable environments. Why? Because a formula doesn't get a headache. It doesn't care if it's Monday morning or Friday afternoon.
People hate this. We feel that human intuition has a "soul" or a "spark" that a spreadsheet lacks. We remember the one time our "gut feeling" saved us and forget the thousand times it led us into a ditch. But if you want accuracy, you have to kill the noise, and humans are just naturally loud.
How to Audit Your Own Noise
You can’t fix what you can't see. Most organizations suffer from "the illusion of agreement." We think we agree with our colleagues because we use the same vocabulary, but we haven't actually checked the numbers.
- Conduct a Noise Audit. Take a set of standardized cases—real ones from the past or hypothetical ones—and have several experts evaluate them independently. Do not let them talk to each other first.
- Compare the Spread. Look at the standard deviation. Is it 5%? Or is it 50%?
- Acknowledge the Gap. Most people find the results of a noise audit deeply insulting. It challenges their professional identity. You have to frame it as a systemic issue, not a personal failure.
Decision Hygiene: The Boring Path to Better Choices
Handwashing doesn't cure a specific disease; it prevents a thousand different ones from starting. Decision hygiene is the same. It’s a set of preventative measures to stop noise: a flaw in human judgment from creeping into your work.
One of the best techniques is "structuring" your decisions.
Instead of looking at a project and saying "I like this," break it down into five independent categories. Rate each category separately. Only at the very end should you look at the total. This prevents "halo effects," where one good trait makes you ignore four bad ones.
Another trick? The "Pre-Mortem." Proposed by psychologist Gary Klein, this involves imagining that a year has passed and your decision was a total disaster. Now, work backward. Why did it fail? This forces your brain to look for flaws it would otherwise ignore because of "groupthink" or simple optimism.
💡 You might also like: Chase Equity Loan Calculator: Why Your Results Might Be Wrong
The Power of Independent Thinking
If you’re in a meeting and the boss speaks first, the meeting is over.
Everyone else will subconsciously (or consciously) align their judgment with the leader. That’s noise amplification. To get a "clean" signal, you need people to write down their thoughts in private before a single word is spoken aloud.
Actionable Steps for Reducing Noise
Stop trusting your "gut." Your gut is mostly just a collection of biases and reaction to your last meal.
- Use "Sequencing": When hiring or evaluating, don't look at the person's name or background until you've looked at their work samples. Information should be meted out in a specific order to prevent early impressions from coloring later data.
- Aggregate Judgments: If you have the resources, get three people to look at a problem and take the average. The "wisdom of the crowd" works because individual noise tends to cancel itself out.
- Standardize the Scale: Don't ask people to rate things on a "1 to 10" scale without defining what a 7 actually means. One person's 7 is another person's 5. Give concrete anchors for every point on your scale.
- Delay Intuition: You don't have to banish your "feelings" entirely. Just wait. Do the hard, structured analysis first. If your gut still disagrees with the data at the end, then you can have the conversation. But don't let the gut drive the car from the start.
The goal isn't to become a robot. The goal is to realize that our brains are incredibly susceptible to the environment. By admitting that we are "noisy" by nature, we can build better systems—at home and at work—that lead to fairer, more profitable, and more consistent outcomes.