Card 01

🏥 🔬 📊

If a medical test is 99% accurate and you test positive, what's the chance you actually have the disease?

💭 How to Think About This

Most people say "99%!" But that's wrong—often dramatically wrong. The answer depends on something crucial: how common is the disease in the first place? This is called the BASE RATE, and ignoring it is one of the most common mistakes in probability thinking.

What's your intuitive answer about your chance of having the disease?

🔒 Write your thinking above to unlock hints

99% accurate means: "IF you have the disease, 99% chance of testing positive." But that's different from: "IF you test positive, 99% chance of having the disease." These sound similar but are NOT the same question! Which direction is the test accuracy measuring?

Think about a rare disease affecting 1 in 10,000 people. Even with a 99% accurate test, 1% of healthy people test positive (false positives). In 10,000 people, that's about 100 false positives! But how many truly sick people? Just 1. So most positive results are wrong.

Our brains focus on vivid, specific information (the test result) and ignore boring background statistics (how rare the disease is). This is called BASE RATE NEGLECT. The "99% accurate" feels so strong that we forget to ask: "But how common is this disease anyway?"

This trap appears everywhere: security screening ("suspicious" doesn't mean "guilty"), investment ("promising startup" still usually fails), news (rare dramatic events get covered, common boring ones don't). Always ask: "How common is this in the first place?"

The answer isn't 99%—it could be as low as 1%!

Your intuition is the MOST common response—and it's the exact trap this problem reveals. Here's what went wrong:

The key insight: "Test is 99% accurate" tells you: IF sick → 99% positive. But you need: IF positive → how likely sick? These are reversed questions with very different answers!

The math with a rare disease (1 in 10,000):

• 10,000 people tested

• 1 truly sick person → tests positive (correctly)

• 9,999 healthy people → 100 test positive (1% false positive rate)

• Total positives: 101. Only 1 is real. Your chance: ~1%!

The lesson: Before reacting to specific information (test result), always ask: "How common is this thing in the first place?" The BASE RATE changes everything.

You're right to sense something is missing! The crucial factor is the BASE RATE—how common the disease is in the population BEFORE testing. If 50% of people have it vs. 0.01% of people have it, the same test accuracy means very different things for your actual risk.

99% accuracy with COMMON disease (50% base rate): Most positives are real. 99% accuracy with RARE disease (0.01% base rate): Most positives are false! The test doesn't change—but what it means for YOU changes dramatically based on prevalence.

Imagine testing 10,000 people for a disease affecting 1 person. The 99% accurate test finds that 1 sick person (great!) but ALSO falsely flags 100 healthy people (1% of 9,999). So 101 positive results, but only 1 is real. Your chance if positive? About 1%!

You spotted what most people miss! This "it depends" thinking applies to: medical screening (why doctors do follow-up tests), security alerts (most flagged people are innocent), startup predictions (most "promising" ventures fail). Always ask: "How common is this to begin with?"

Exactly right—it depends on the BASE RATE!

You showed excellent probabilistic intuition by sensing that test accuracy alone isn't enough information.

The key insight: "99% accurate" tells you P(positive | sick)—probability of testing positive IF you're sick. But you want P(sick | positive)—probability of being sick IF you test positive. These are reversed, and VERY different when base rates are low!

The math for a rare disease (1 in 10,000):

• 10,000 people tested

• 1 truly sick → tests positive

• ~100 healthy → test positive (false positives)

• Total: 101 positives, only 1 real = ~1% chance

For a common disease (1 in 2):

• 5,000 sick → 4,950 test positive

• 5,000 healthy → 50 test positive (false)

• Total: 5,000 positives, 4,950 real = 99% chance

The habit: Always ask "How common is this in the first place?" before interpreting any test, claim, or prediction.

Good instinct! You sensed that "99% accurate" might be misleading. The key is understanding what 99% accurate actually MEANS—it's about the test's performance, not your actual risk. These are different questions entirely.

Your actual risk depends heavily on the BASE RATE—how common the disease is. For a rare disease (1 in 10,000), even a 99% accurate test produces mostly false positives! The rarer the condition, the worse this "false positive paradox" becomes.

10,000 people, 1 has disease. Test finds that 1 sick person (99% sensitive). But 1% of 9,999 healthy people = ~100 false positives. Total positives: 101. Only 1 is real! If you test positive, you have about a 1% chance of actually being sick. Counterintuitive but mathematically certain.

This base rate logic applies everywhere: airport security (most "suspicious" passengers are innocent), startup investing (most "promising" companies fail), criminal profiling (fitting a profile doesn't mean guilty). You've found a powerful pattern for probabilistic thinking!

You're right—it can be surprisingly low (around 1% for rare diseases)!

You avoided the trap that catches most people. Your suspicion about the "obvious" answer shows strong probabilistic thinking.

The key insight: "99% accurate" means P(positive | sick) = 99%. But you want P(sick | positive)—these are reversed! For rare diseases, false positives vastly outnumber true positives.

The math (disease rate 1 in 10,000):

• 10,000 tested: 1 sick, 9,999 healthy

• True positive: 1 (the sick person)

• False positives: ~100 (1% of healthy)

• Your odds if positive: 1/101 ≈ 1%

Why most people miss this: Our brains focus on specific, vivid info (test result) and ignore background statistics (base rate). This BASE RATE NEGLECT is one of the most common probability errors.

Your advantage: Asking "how common is this?" before interpreting specific information is a superpower for decision-making.

🤔 Which thinking lens(es) did you use?

Select all the lenses you used:

Learn more about the 7 Thinking Lenses →

👨‍👩‍👧 For Parents & Teachers

🌱 A Small Everyday Story

A school screens 1,000 students for a rare condition.
The test is 95% accurate.
10 students test positive.
Parents panic—but only 1 actually has the condition.
The other 9? False positives.
The base rate made all the difference.

See more guidance →

🧠 Thinking habits this builds:

Always asking "how common is this?" before interpreting new information
Understanding why accurate tests can still produce mostly false positives
Resisting the urge to jump to conclusions from vivid, specific information
Thinking in populations, not just individual cases

🌿 Behaviors you may notice (and reinforce):

"But how common is that in the first place?" questions
Skepticism about dramatic statistics without context
Understanding why rare events dominate news coverage
Asking for base rates before making judgments

How to reinforce: When they encounter a statistic or claim, ask: "What's the base rate here? How does that change your interpretation?"

🔄 When ideas are still forming:

Some learners may think all tests are useless if they produce false positives. Help them see that base rates determine interpretation—the same test is more useful when the condition is more common.

Helpful response: "When would this test be really useful? When would it be misleading?" Guide them to see context matters.

🔬 If you want to go deeper:

Look up Bayes' Theorem for the formal mathematics
Explore why COVID testing strategies changed as prevalence changed
Discuss how lawyers use (and misuse) base rates in criminal trials

Key concepts (for adults): Base rate neglect, prior probability, Bayes' theorem, false positive paradox, conditional probability, population thinking, prosecutor's fallacy.

If a medical test is 99% accurate and you test positive, what's the chance you actually have the disease?

🔄 Compare All Perspectives

📈 "Very high (99%)"

🤔 "It depends"

📉 "Surprisingly low"

🤔 Which thinking lens(es) did you use?