Bad mathematics about unsold books

Numbers fascinate, numbers charm, numbers click. But numbers can deceive and, in the hands of lazy journalists, end up expressing wrong things. This is what has happened these days with a headline I have seen repeated in quite a few media outlets: “Almost half of the books sell no copies and only 4.5% sell more than a hundred”. This one is from 3CatInfo, but I insist that it has spread like wildfire… even if it is wet gunpowder. The data comes from surveys of bookstores and, in reality, it says that, on average, in each of these establishments there are half the titles that do not generate a single purchase in a whole year. But this does not mean that they are the same references in all bookstores. Whoever produced this statistic wanted to indicate that, to start being a bookseller, one must invest in equipping oneself with a considerable stock and assume that half will only accumulate dust and add bulk to the store. On the other hand, it is not indicative that too many books are published, the litany that has already become a cliché in the publishing sector. I emphasize that I do not deny the excess of published books, but that this viral figure is not a measure of this phenomenon.

The error is the result of the secular low mathematical culture of many journalists, who do not detect the improbability of a figure, the laziness in picking up the phone to clarify what instinctively doesn't quite add up, and the rush of the digital age that makes you repeat the neighbor's news if you suspect it might go viral, without bothering to verify what is being said. Thirty years ago, John Allen Paulos wrote the highly recommended book A Mathematician Reads the Newspaper, in which he amusingly reviewed the errors of scribblers when dealing with numbers. I fear, in view of the facts, that it must be one of those books that, in the last year, has not sold a single copy in many bookstores.