modic.
Start exploring
Menu
A PATH THROUGH THE IDEAS

Build intuition.
One experiment at a time.

Start with a signal. Learn to question an average. End with the difference between doing well and being able to continue.

Begin the first experiment Jump to the glossary ↓

6 experiments · about 27 minutes · no prerequisites

PICK YOUR NEXT QUESTION

Find your next route into the world.

YOUR ROUTE

From evidence to consequences.

Follow in order, or take a detour.
  1. 01
    Base Rate Neglect · 4 MIN

    Start with the odds

    What did you believe before seeing the signal?

    Raise the spam base rate to 50%. Explain why the same flag becomes more informative.

  2. 02
    Regression to the Mean · 4 MIN

    Separate signal from luck

    Will an extraordinary result happen again?

    Remove measurement noise. Notice what happens to the selected group’s second score.

  3. 03
    Lindy Effect · 5 MIN

    Think in distributions

    What can the past tell us about possible futures?

    Compare low and high uncertainty at the same age. Read the interval as well as the median.

  4. 04
    Power Laws · 4 MIN

    Look beyond the average

    How much can a small minority account for?

    Compare exponent 0 with exponent 2. Watch the share going to the top tenth.

  5. 05
    Network Effects · 4 MIN

    Connect the pieces

    Does a bigger network always mean more useful connections?

    Grow the network with connection chance at 0%, then at 100%. Separate potential from use.

  6. 06
    Gambler’s Ruin · 6 MIN

    Protect the ability to continue

    Can a favorable game still end your journey?

    Compare small and large stakes across many lives. Read the ruin rate, not just mean wealth.

A SMALL WORKING VOCABULARY

Keep the ideas close.

86 terms
Allais Paradox
Two pairs of lotteries can expose tension in how we value certainty.See the idea in context ↗︎
Allee Effect
A very small population may struggle to recover even when resources are available.See the idea in context ↗︎
Amdahl’s Law
A part that cannot be sped up limits the gain from adding more workers.See the idea in context ↗︎
Anchoring Bias
An initial number can keep pulling an estimate toward itself.See the idea in context ↗︎
Antifragility
The shape of a response determines whether variation helps or hurts.See the idea in context ↗︎
Barbell Strategy
Separate a protected reserve from a small, explicitly risky allocation.See the idea in context ↗︎
Base rate
How common something is before you observe new evidence.See the idea in context ↗︎
Beauty Contest Game
Try to predict what others will predict, not a number's intrinsic value.See the idea in context ↗︎
Benford’s Law
Leading digits in scale-spanning measurements are often far from uniform.See the idea in context ↗︎
Birthday Paradox
A small room contains far more possible pairs than people.See the idea in context ↗︎
Braess’s Paradox
An extra shortcut can make selfish routing slower for everyone.See the idea in context ↗︎
Bullwhip Effect
Small customer changes can become larger swings in upstream orders.See the idea in context ↗︎
Butterfly Effect
Tiny differences in starting conditions can grow into very different trajectories.See the idea in context ↗︎
Centipede Game
Passing grows the pot, but each player can take it first.See the idea in context ↗︎
Chicken Game
Each side wants the other to yield, but mutual escalation is costly.See the idea in context ↗︎
Compound Growth
Small repeated changes build on everything that came before.See the idea in context ↗︎
Compounding
Repeated changes applied to a quantity that has already changed.See the idea in context ↗︎
Concentration
How much of a total is accounted for by a small subset.See the idea in context ↗︎
Conditional probability
The probability of an event after restricting attention to cases where another event happened.See the idea in context ↗︎
Confirmation Bias
Tests that agree with a theory may leave its strongest rivals untouched.See the idea in context ↗︎
Coordination & Focal Points
A shared convention can help people choose the same place without talking.See the idea in context ↗︎
Coupon Collector
The final missing item can take much longer than the first few.See the idea in context ↗︎
Decoy Effect
An inferior third option can change how two original options are compared.See the idea in context ↗︎
Diffusion of Innovations
Early adopters and imitation can create an S-shaped adoption curve.See the idea in context ↗︎
Diminishing Returns
Adding one more input can produce less extra output than the one before.See the idea in context ↗︎
Distribution
A description of possible values and how probability is spread among them.See the idea in context ↗︎
El Farol Bar Game
If everyone expects a quiet venue, the venue may become crowded.See the idea in context ↗︎
Ellsberg’s Urns
Known odds and unknown odds can feel different even when their midpoint matches.See the idea in context ↗︎
Endowment Effect
Owning an item can change the price at which someone is willing to give it up.See the idea in context ↗︎
Expected value
An average weighted by probabilities. It need not describe a typical individual outcome.See the idea in context ↗︎
Explore or Exploit
Trying an uncertain option can teach you something, even when another seems better now.See the idea in context ↗︎
False positive
A flag raised when the thing being sought is absent.See the idea in context ↗︎
Forgetting Curve
Memory fades, and the timing of review changes what remains available later.See the idea in context ↗︎
Framing Effect
Equivalent outcomes can feel different when described as gains or losses.See the idea in context ↗︎
Gambler’s Fallacy
A streak can feel like it makes the opposite result due, even when trials are independent.See the idea in context ↗︎
Hotelling’s Location Game
Two sellers compete for customers spread along one street.See the idea in context ↗︎
Information Cascade
People can follow earlier actions even when their own clue disagrees.See the idea in context ↗︎
Inspection Paradox
A random arrival is more likely to land in a long gap than a short one.See the idea in context ↗︎
Jevons Paradox
A more efficient tool can increase total resource use if demand expands enough.See the idea in context ↗︎
Law of Large Numbers
More observations can stabilize an average without making the next outcome predictable.See the idea in context ↗︎
Little’s Law
More items in a stable process usually means a longer time spent inside it.See the idea in context ↗︎
Logistic Growth
Growth slows as a population approaches a finite capacity.See the idea in context ↗︎
Market for Lemons
When buyers cannot observe quality, good products can leave the market.See the idea in context ↗︎
Matching Pennies
A predictable choice can be exploited even when neither pure choice is best.See the idea in context ↗︎
Median
A middle value: at least half of the observations lie on either side of it.See the idea in context ↗︎
Median Voter Game
Moving toward the middle can win more nearby voters in a simple one-dimensional election.See the idea in context ↗︎
Monty Hall Problem
A host who knows the answer changes what an open door tells you.See the idea in context ↗︎
Nash Bargaining
A negotiated split depends on what each side can get without a deal.See the idea in context ↗︎
Network effect
A change in the benefit of using something as other people use it. More connections are one possible mechanism.See the idea in context ↗︎
Noisy Channel
Repeating a bit can overcome some errors but uses extra capacity.See the idea in context ↗︎
Opportunity Cost
Choosing one useful thing means giving up another useful thing.See the idea in context ↗︎
Optionality
The right to act without an obligation changes the downside of uncertainty.See the idea in context ↗︎
Pareto Concentration
A small fraction can account for a large share when values have a heavy tail.See the idea in context ↗︎
Peak–End Rule
A remembered experience may emphasize its strongest moment and its ending.See the idea in context ↗︎
Percentile
A cutoff describing where a value sits in an ordered distribution; the 75th percentile has roughly three quarters of observations below it.See the idea in context ↗︎
Planning Fallacy
A project can take longer than the inside-view estimate suggests.See the idea in context ↗︎
Power law
A relationship where one quantity varies as a fixed power of another. In this experiment, attention falls as rank rises.See the idea in context ↗︎
Predator–Prey Cycles
Two interacting populations can rise and fall out of phase.See the idea in context ↗︎
Present Bias
An immediate reward can outweigh a larger future one—and reverse an earlier plan.See the idea in context ↗︎
Prisoner’s Dilemma
A tempting individual move can leave both players worse off.See the idea in context ↗︎
Public Goods Game
A contribution can benefit the whole group while costing the contributor.See the idea in context ↗︎
Regression to the mean
When selecting an extreme noisy measurement is followed, on average, by a less extreme one.See the idea in context ↗︎
Ruin threshold
A boundary at which this model stops a journey because its resources are too low.See the idea in context ↗︎
Schelling’s Segregation Model
Local preferences can produce a larger pattern that nobody explicitly chose.See the idea in context ↗︎
Secretary Problem
Stopping too early and searching forever can both lose the best candidate.See the idea in context ↗︎
Shannon Entropy
A fair binary event carries more uncertainty than one that almost always repeats.See the idea in context ↗︎
Signal and noise
The underlying quantity you care about, and the variation that obscures it.See the idea in context ↗︎
Signal Detection
A stricter threshold reduces false alarms but can miss real signals.See the idea in context ↗︎
Simpson’s Paradox
A trend inside every group can reverse when the groups are pooled.See the idea in context ↗︎
SIR Spread Model
Contact and recovery rates change the shape of an outbreak in a closed population.See the idea in context ↗︎
Skin in the Game
Who receives the upside—and who bears the loss—can change a decision.See the idea in context ↗︎
Social Tipping Point
A small change in visible adoption can trigger a larger collective shift.See the idea in context ↗︎
Stag Hunt
A valuable shared goal can require confidence that others will join.See the idea in context ↗︎
Sunk Cost Fallacy
Money already spent can pull a decision away from what happens next.See the idea in context ↗︎
Survivorship bias
Drawing conclusions from visible survivors while missing those that disappeared.See the idea in context ↗︎
Threshold Public Good
A shared project succeeds only after enough people contribute.See the idea in context ↗︎
Tragedy of the Commons
Shared resources depend on both individual demand and the rules people build together.See the idea in context ↗︎
Trust Game
The first person can grow a shared pie but cannot force a fair return.See the idea in context ↗︎
Tullock Contest
Spending more effort raises your chance of winning but everyone pays their effort cost.See the idea in context ↗︎
Turkey Problem
A long reassuring history can miss the mechanism that ends it.See the idea in context ↗︎
Ultimatum Game
An offer can be profitable and still be rejected.See the idea in context ↗︎
Uncertainty
The range of possibilities left open by what you know and by the assumptions you make.See the idea in context ↗︎
Vickrey Auction
Paying the second-highest bid changes the incentive to report your value.See the idea in context ↗︎
Volunteer’s Dilemma
Everyone benefits if one person acts, but each hopes someone else will.See the idea in context ↗︎
Winner’s Curse
The highest estimate may be the one with the largest upward error.See the idea in context ↗︎
Zipf’s Law
A ranked item can appear roughly in inverse proportion to its rank.See the idea in context ↗︎