How to Lie With Statistics (1954) [pdf]

Commenters praised How to Lie With Statistics as an enduring guide, saying the techniques it describes remain common and still persuade people. One singled out truncated-axis bar charts that make modest growth look dramatic, arguing that presentation and selective omission can matter more than the underlying data. Others called the book an excellent introduction to statistics; one said it helped inspire a statistics degree. The book’s availability prompted disagreement: one commenter asked why a PDF was being shared when the book remains in print, while another suggested its age and the author’s death made sharing reasonable.

Commenters also distinguished misleading statistics from the broader use of numbers without context. One cited the claim that a datacenter uses a million gallons of water per year as an example of a number that can mislead without explaining its scale or circumstances. Another objected that this is not necessarily a lie, but a use of implication: in the context of datacenter water concerns, the example could suggest that critics are engaging in deceptive rhetoric without directly saying so. The disagreement centers on whether misleading numerical claims should be understood as lies or as subtler effects of framing and audience inference.