Chapter 3
Numbers and chance
Numbers can mislead as easily as words. A coin has no memory, an average can hide what's typical, a risk that 'doubles' may still be tiny, and a survey of ten people proves very little. A few simple ideas from statistics help you read headlines, test results and sales pitches without being fooled. Nobody is immune, experts included, but knowing the names makes the traps much easier to notice, in yourself as well as in others.
10 cards
See the 10 cards
- Believing red is 'due' after a run of blacksGambler's fallacy
- Chance a fair coin lands heads after 5 tails in a row50%
- The middle value once numbers are in orderMedian
- Add them all up and divide by how manyMean
- The value that appears most oftenMode
- Risk goes from 1 in 10,000 to 2 in 10,000: the extra risk1 in 10,000
- Why a survey of 10 people tells you littleThe sample is too small
- Ignoring how common something is to begin withBase rate neglect
- Feeling better after a pill with nothing in itPlacebo effect
- Extreme results tend to be followed by ordinary onesRegression to the mean