A cognitive bias is a deviation in thought that moves in a predictable direction, rather than at random. A random error is just carelessness; a bias is an error everyone makes in the exact same direction. This distinction matters because a predictable error is something you can work on.
Many of these biases are actually useful shortcuts. Instead of calculating every decision from scratch, your mind uses quick rules of thumb. Most of the time, these rules produce the right result; under specific conditions, however, they fail systematically.
Here are twelve of the most well-documented cognitive biases, along with everyday examples and ways to counter them.
1. Confirmation bias
What it is: The tendency to search for, notice, and remember information that supports your existing belief, while ignoring or scrutinizing conflicting information.
Example: Deciding an investment is good, then reading praise for it while dismissing critics as "clueless."
Countermeasure: Before making a decision, write down "how would I know if this is wrong?" Make finding the strongest counterargument a distinct task.
2. Anchoring effect
What it is: The tendency for the first number you hear to pull subsequent estimates toward itself. Even if the anchor is completely irrelevant, the effect holds.
Example: In a negotiation, the party that names the first price largely determines the range around which the final price settles.
Countermeasure: Write down your own estimate before hearing the other party's number. Once you hear it, deliberately generate a range in the opposite direction and evaluate the midpoint.
3. Availability heuristic
What it is: Estimating the likelihood of an event based on how easily examples come to mind. Things that are easily recalled are assumed to happen more often.
Example: Believing flying is more dangerous than driving after seeing news coverage of a plane crash. High-newsworthiness events take up disproportionate space in memory.
Countermeasure: Ask "what is the base rate?" instead of "how many examples can I remember?" Look at actual numbers whenever possible.
4. Sunk cost fallacy
What it is: Continuing down a bad path because of unrecoverable past investments. Rational decisions look only at future costs and benefits; past spending is already gone.
Example: Trying to finish a boring book because "you're already halfway through." Staying in a degree program where you're unhappy because you've already given it three years.
Countermeasure: Ask yourself, "If I were starting from scratch today, would I choose this?" If the answer is no, past investment is not a valid justification.
5. Survivorship bias
What it is: Looking only at those who survived a process while ignoring those who were eliminated. The invisible portion of the dataset often carries the real lesson.
Example: Looking at a list of "billionaires who dropped out of college" and concluding that dropping out is a good strategy. Tens of thousands of people did the same thing and failed, but no books were written about them.
Countermeasure: In every success story, ask "where are the people who did the exact same thing and failed?"
6. Hindsight bias
What it is: Viewing an outcome as predictable all along once you learn what happened. "I knew it was going to happen anyway."
Example: Thinking a game's outcome was inevitable after it ends. Everyone remembering their warnings after a crisis, even though nobody took a defensive position before it hit.
Countermeasure: Write down your key predictions with a timestamp. Your memory will quietly update your past predictions later; a written record stops this.
7. Fundamental attribution error
What it is: Attributing other people's behavior to their personality, while attributing your own behavior to situational circumstances.
Example: Someone cutting you off in traffic is "disrespectful," but when you do it, you were "in a rush."
Countermeasure: When confronted with annoying behavior, take a minute to consider what circumstance would make that action reasonable if you were in their shoes.
8. Overconfidence
What it is: Overestimating the accuracy of your predictions. When you feel "completely certain," the error rate is often surprisingly high.
Example: Estimating how long a project will take. The planning fallacy is the most expensive form of overconfidence; projects almost always take longer than estimated.
Countermeasure: Give a range instead of a single number, and deliberately widen that range. For timeline estimates, look at actual durations of similar past tasks rather than your best-case scenario.
9. Framing effect
What it is: The way information is presented changing the decision made from it.
Example: A "90% survival rate" and a "10% mortality rate" describe the exact same fact, yet people choose surgery much more often when presented with the first frame.
Countermeasure: On important decisions, deliberately rewrite the information in the opposite frame and see whether your decision changes.
10. Loss aversion
What it is: The pain of losing something feels significantly stronger than the pleasure of gaining something of equal value.
Example: Holding onto a stock that is losing money in hopes it "comes back," even if you wouldn't invest the same amount of cash in that stock today.
Countermeasure: Frame the choice independently of ownership: "If I didn't own this asset right now, would I buy it at today's price?"
11. Clustering illusion
What it is: Seeing patterns in random data. Your mind is so wired to find patterns that it creates them where none exist.
Example: Thinking "it has to be tails next" after getting multiple heads in a row. A coin flip has no memory.
Countermeasure: When you spot a pattern, consider what the same data would look like if generated completely at random. Random sequences look far more "clustered" than you think.
12. Dunning-Kruger effect
What it is: People with low competence in an area overestimating their own skill level. A common summary: the less you know about something, the less equipped you are to realize how little you know.
An honest note: This effect is the most debated entry on this list. Some statisticians argue that the original finding might largely be an artifact of measurement error and data plotting. The debate continues. Still, the core observation—that people are bad at calibrating their own competence—is supported by other methods.
Countermeasure: Seek external feedback and objective measurement rather than relying on self-assessment.
Can you get rid of these?
Short answer: not completely. These biases do not stem from carelessness; they are side effects of how the mind normally works. Experts fall for them, and so do people who are aware of them.
What you can do is more modest, but effective:
- Slow down decisions. Most biases kick in during fast, automatic thinking.
- Write things down. Logging predictions, reasoning, and timestamps is the most practical defense against hindsight bias and overconfidence.
- Change the system. Setting up decision processes that block bias (like blind evaluations, preset criteria, and checklists) works better than individual effort.
- Find a dissenting voice. Overcoming your own confirmation bias on your own is hard; involving someone who will challenge you is much easier.
These systematic mental errors are not inversely related to intelligence—high achievers fall into the exact same traps. If you want to explore different aspects of reasoning, check out our guide on how to solve logic puzzles or take our free IQ test directly. Our article on multitasking and intelligence also explores how attention gets derailed during decision-making.