Only 3% Escape This Sneaky Logic Trap
Most people read these puzzles too fast and walk straight into the trap. The answers feel obvious, which is exactly the problem — obvious is the bait.
Slow down for seven questions and your brain gets one job: resist the answer that jumps out first. Each correct pick earns points; each shortcut costs you. Score high enough and you join the tiny 3% who actually read the fine print of a problem. Ready to prove you're built different?
Read every possible outcome for this quiz and what each one means →
The bat, the ball, and the answer almost everyone gives
The bat-and-ball item in this set comes from the Cognitive Reflection Test, a three-question task published by Shane Frederick in 2005. Its charm is that the wrong answer, ten cents, arrives with total confidence before any arithmetic has happened. If the ball cost ten cents, the bat at a dollar more would cost $1.10, and the pair would come to $1.20. The ball costs five.
Frederick found that even at highly selective universities, large fractions of students handed back the intuitive error. That is the whole design: the task does not ask whether you can do the sums, it asks whether you notice that you ought to check them.
Fast thinking is not faulty thinking
Psychologists often describe two modes of reasoning, one rapid and automatic, one slow and effortful. The quick mode is not a defect. It is what lets you cross a road, read a face or hear a wrong note in a familiar song. Puzzles of this kind are engineered specifically to punish it, by dressing an unusual structure in familiar clothing, which makes them entertaining and makes them poor evidence about how you think for the other twenty-three hours of the day.
There is a practice effect too. Once the sheep question or the widget question has caught you, it never works again. A better score on a second run reflects exposure, not a sharper mind.
About that three per cent
Nobody counted. The number in the title is showmanship, the same way a pub quiz advertises itself as impossible. These items have not been given to a representative sample, no percentile has been computed, and the set is not a validated instrument for anything at all. A strong result is evidence that you read carefully, which is a genuinely useful habit, and it is not a measurement of intelligence, which this page cannot provide.
If the puzzles leave you curious, Frederick's original paper and the wider reasoning literature are open reading, and considerably more rewarding than any tier you land in.
