Cognitive Biases

The systematic ways your brain misfires, and the small habits that pull it back into line.

Why Biases Exist

Your brain evolved for a different environment. Mental shortcuts that worked on the savanna cause systematic errors in modern contexts. These shortcuts are not random noise. They lean in predictable directions, which is the only reason this chapter is possible.

Biases are predictable. Knowing them lets you compensate.

The Most Dangerous Biases

Confirmation Bias

What it is: Seeking, interpreting, and remembering information that confirms what you already believe.

Example: You think a coworker is lazy. You notice every time they leave early, but not when they stay late.

Counter-measures:

• Actively seek disconfirming evidence
• Ask "What would change my mind?"
• Assign someone to argue the opposite position
• Read sources you disagree with

Sunk Cost Fallacy

What it is: Continuing a course of action because of past investment, regardless of future value.

Example: Staying in a bad relationship because you've already invested 5 years.

Counter-measures:

• Ask "If I were starting fresh, would I choose this?"
• Set kill criteria in advance
• Focus on future costs and benefits, not past
• The money/time is already gone. It's irrelevant

Availability Heuristic

What it is: Judging probability by how easily examples come to mind.

Example: After seeing news coverage, believing plane crashes are more common than car crashes.

Counter-measures:

• Look up actual statistics
• Ask "Is this memorable because it's common or because it's rare?"
• Recognize that vivid events distort perception

Anchoring

What it is: Over-relying on the first piece of information encountered.

Example: A high initial price makes a "discounted" price seem reasonable, even if it's still overpriced.

Counter-measures:

• Research independently before seeing their number
• Ignore arbitrary anchors
• Generate your own anchor from first principles
• Ask "What would I estimate without this information?"

Overconfidence Bias

What it is: Believing you're more right, more skilled, or more in control than you are.

Example: 93% of American drivers think they're above average.

Counter-measures:

• Track your predictions and review accuracy
• Assign probabilities, not certainties
• Ask "What's the base rate for people in my situation?"
• Seek feedback from others

Social Biases

Groupthink

What it is: Conforming to group consensus, suppressing dissent.

Counter-measures:

• Assign a devil's advocate role
• Anonymous voting on decisions
• Leader speaks last
• Encourage criticism explicitly

Authority Bias

What it is: Trusting someone because of their position, not their reasoning.

Counter-measures:

• Evaluate the argument, not the arguer
• Ask "Would this reasoning convince me from a random person?"
• Recognize that experts are often wrong outside their domain

Bandwagon Effect

What it is: Adopting beliefs or behaviors because others do.

Counter-measures:

• Form your view before seeing others'
• Ask "What would I think if I were the first to consider this?"
• Value independent thinking over consensus

In-Group Bias

What it is: Favoring people in your group, disfavoring outsiders.

Counter-measures:

• Actively consider opposing viewpoints
• Seek information from outside your group
• Evaluate ideas without knowing who proposed them

Memory and Perception Biases

Hindsight Bias

What it is: Believing, after the fact, that you "knew it all along."

Example: "I knew that startup would fail" (after it failed).

Counter-measures:

• Record predictions before outcomes
• Review decision journals
• Ask "What did I actually predict?"

Recency Bias

What it is: Weighting recent events more heavily than older ones.

Example: A manager rates an employee based on the last month, not the whole year.

Counter-measures:

• Keep logs and records
• Force yourself to consider full timeframes
• Create checklists for balanced evaluation

Peak-End Rule

What it is: Judging experiences by their peak intensity and end, not their average.

Counter-measures:

• Evaluate experiences systematically
• Don't let endings disproportionately color judgments
• Consider the whole experience

Decision-Making Biases

Status Quo Bias

What it is: Preferring the current state of affairs over change.

Example: Staying with a bad insurance policy because switching is effort.

Counter-measures:

• Ask "If I were starting fresh, would I choose this?"
• Quantify the cost of not changing
• Set calendar reminders to review recurring decisions

Loss Aversion

What it is: Losses hurt more than equivalent gains feel good (~2x).

Example: Refusing a bet with 50% chance to win $150 and 50% chance to lose $100.

Counter-measures:

• Frame decisions in terms of net expectation
• Consider portfolio of decisions, not single bets
• Ask "If I had neither, which would I choose?"

Endowment Effect

What it is: Valuing things more simply because you own them.

Example: Demanding more to sell something than you'd pay to buy it.

Counter-measures:

• Ask "What would I pay for this if I didn't own it?"
• Pretend you're advising a friend
• Focus on utility, not ownership

Choice-Supportive Bias

What it is: Remembering your choices as better than they were.

Counter-measures:

• Record decisions and reasoning at the time
• Review outcomes honestly
• Seek external feedback

Information Processing Biases

Dunning-Kruger Effect

What it is: Incompetent people overestimate their ability; experts underestimate theirs.

Counter-measures:

• For beginners: Assume you know less than you think
• For experts: Trust your expertise more
• Seek calibrated feedback

Survivorship Bias

What it is: Focusing on successes, ignoring failures that aren't visible.

Example: Studying successful entrepreneurs without considering the 90% who failed.

Counter-measures:

• Ask "What's the failure rate for this approach?"
• Seek out failure stories
• Consider the full population, not just visible examples

Narrative Fallacy

What it is: Creating stories to explain random events; seeing patterns in noise.

Counter-measures:

• Ask "Could this be random?"
• Look at large sample sizes
• Be skeptical of coherent explanations

Fundamental Attribution Error

What it is: Attributing others' behavior to character, your own to circumstances.

Example: "He's late because he's irresponsible" vs. "I'm late because of traffic."

Counter-measures:

• Consider situational factors for others
• Ask "What circumstances might explain this?"

Emotion-Related Biases

Affect Heuristic

What it is: Letting current emotional state influence judgment.

Example: Making important decisions when angry, excited, or tired.

Counter-measures:

• Delay big decisions when emotional
• Use 10/10/10 to get distance
• Create decision-making rituals that interrupt emotion

Optimism/Pessimism Bias

What it is: Systematically over- or under-estimating positive outcomes.

Counter-measures:

• Use base rates and outside view
• Track your predictions over time
• Adjust based on your historical tendency

Present Bias

What it is: Overvaluing immediate rewards vs. future rewards.

Example: Eating cake now vs. being healthy later.

Counter-measures:

• Pre-commit to future actions
• Make future rewards more salient
• Create environments that support good choices

Practical Application

Before Making a Decision

Ask yourself:

1. What biases am I most prone to?
2. What bias is most likely affecting this decision?
3. What would I think if I had opposite biases?
4. Who can give me an unbiased perspective?

Create Debiasing Habits

The Humility Check

You will still fall for biases. The goal isn't perfection, it's:

1. Recognizing when biases are likely
2. Taking steps to mitigate them
3. Making better decisions on average
4. Learning from mistakes

Next Steps

Continue to 03-risk-uncertainty.md to see how to think about probabilities, expected value, and the kinds of risk that biases love to disguise.