How Numbers Mislead Us

Numbers carry an air of authority that ordinary words rarely manage to achieve. A claim with a percentage attached always feels more solid than the same claim without one.

数字身上带着一种权威感，这是文字很难做到的。一个带百分比的说法，看起来就是比不带百分比的说法更可靠。

But a number is only ever as good as the question that produced it. The same set of data can support opposite conclusions depending on where the counting starts and stops.

但一个数字好不好，取决于产生它的那个问题问得好不好。同一批数据，从哪儿开始数、在哪儿停下，能得出完全相反的结论。

Averages are especially easy to misuse, and especially hard to notice. When a handful of very large values sit at one end of the range, the average describes almost nobody in the group.

「平均值」尤其容易被误用。当少数极大的数值都挤在一端时，这个平均值几乎不能代表群体里的任何人。

The problem is not that statistics actually lie; it is that they answer much narrower questions than we assume. Simply asking who was counted, and who was left out, exposes most of the difficulty.

问题不在于统计会说谎，而在于它回答的问题，比我们以为的要窄。问一句「数进去的是谁、漏掉的是谁」，大部分疑点就暴露出来了。

This does not mean we should reject numbers and return to impressions. It means we should read them the way we read any other claim: by looking for what is missing.

这并不是说我们该抛弃数字、退回凭印象判断。它的意思是：读数字要像读任何一个主张一样 —— 去找它没有说出来的部分。