“How to lie with statistics” is the title of a little book by Darrell Huff, first published in 1954. Here’s an example of the kind of question Huff writes about in the book: A new business makes a very respectable one hundred thousand dollars in profits in its first year. In its second year, the business enjoys a 100 percent increase in profits, relative to the previous year. But in its third year, the business suffers a 75 percent decrease in profits, relative to the previous year. Look on the bright side, someone says. Up 100 percent the second year, then down 75 percent the third year–overall, this person says, profits are still 25 percent above the first year’s level. What’s Wrong With This Argument? What’s wrong, of course, is that the argument glosses over the fact that each of those percentages expresses profits in terms of the previous year’s profits.