Games
The 2-4-6 rule game
Find a hidden rule by testing triples of numbers.
AtlasReason
Biases, Bayes and better judgement
Reason: biases, Bayes and better judgement
Games and tests on cognitive biases and clear thinking: find a hidden rule, resist an anchor, calibrate your confidence, and work base rates with an icon array.
- 21games, tests and tools
- Newquestions every round
- Freeno sign-up
What these tests can and cannot tell you
A bias is a tendency measured across groups of people in experiments, so no quiz can tell you how biased you are. What these games and tests can do is put you inside the classic procedures, such as testing a hidden rule, estimating after an anchor, giving 90% ranges or reading a base rate, and show you from your own answers where the quick answer and the checked one came apart. The games and puzzle tests are built in your browser from a seed, every answer is explained, and nothing you type leaves the page.
Games
Play the classic procedures from the inside. Each game is generated from a seed, so the next round is new.
Games
The anchoring game
Estimate after a higher-or-lower question, and see if the number pulled you.
Games
Calibration trainer
Rounds of five 90% ranges, scored round by round.
Games
Spot the fallacy: speed round
Name the flaw in short arguments against the clock, or untimed.
Tests
Short sittings with a result you can read: what you got right, what the quick answer was, and why.
Tests
Reflection puzzles test
Seven word problems with a tempting wrong answer each.
Tests
Calibration test
Ten 90% ranges in one sitting, with a calibration curve.
Bayes and base rates
How often a flagged case is real, worked as counts of people rather than percentages.
Bayes and base rates
Bayes calculator with an icon array
Three rates in, counts and an icon array out.
Bayes and base rates
Bayesian intuition test
Six base-rate problems, counts against percentages.
Bayes and base rates
Bayes' theorem, worked as counts instead of formulas
Bayes' 1763 essay, a frequency tree, and the calculator.
Bayes and base rates
The conjunction fallacy, and why more detail feels likelier
The 1983 studies, the wording debate, and three scenarios.
Biases
Deep pages on single biases, and a map of fifty-one of them.
Biases
The Dunning–Kruger effect, and what the data can and cannot show
The 1999 finding, the curve that is not in it, and the artefact debate.
Biases
Confirmation bias, and the experiments it comes from
Wason's hidden-rule task, biased assimilation, and a self-check.
Biases
The sunk cost fallacy, and why the past keeps a vote it should not have
Arkes and Blumer, escalation of commitment, and three scenarios.
Biases
Survivorship bias, and the planes that never got counted
Wald's returning bombers, disappearing funds, and three scenarios.
Biases
The availability heuristic, and what comes to mind first
Letter frequency, feared causes of death, and three scenarios.
Biases
Anchoring, and how one number moves the next
The wheel-of-fortune study, a Many Labs result, and the game.
Biases
The halo effect, and how one trait colours the rest
Thorndike's officer ratings, an unaware audience, and a self-check.
Biases
The base-rate fallacy, and how a rare case gets lost
The lawyer-engineer study, the cab problem, and the Bayesian test.
Biases
The framing effect, and how a sentence moves a choice
The Asian disease study, a typology of framings, and three scenarios.
Biases
Hindsight bias, and why the past looks predictable
Fischhoff's outcome-knowledge studies and a write-it-down check.
Biases
The Bias Atlas
Fifty-one biases in six regions, on one map.
Myths
Popular claims about the mind, checked against the studies.
Myths
Psychology myths quiz
Fifteen popular claims, true or false, each with its study.
Also on this site
Named biases and fallacies, with examples.
The guide to fifteen cognitive biases groups the everyday ones by what they distort, the guide to twenty logical fallacies gives an example of each, and the spot the fallacy test is a fixed set of twenty arguments to practise on.
How to use these pages
Play a game before reading about it: the classic procedures teach more from the inside. Then read the method note under each one, which says what the result can and cannot mean. Nothing here diagnoses anything or ranks you against other people.