Nate Silver, whose new book The Signal and the Noise aims to help people make better predictions, is celebrated in America for his ability to predict the outcome of the presidential election.
However, a former professional poker player, Silver explains that his advantage comes because his competitors are television pundits and they are widely inaccurate in their views.
The views of TV pundits are shaped by their own biases and they get lost in the narrative. By which he means the pundits believe in the candidates whose stories they most like.
The FT last week started a series on Big Data, where companies are looking for advantage through better customer understanding, searching the internet for hidden patterns and trends to give them an edge.
How much data is out there? 2.6 zettabytes are generated every year. How big is a zettabyte? It is 1 billion terrabytes. How big is a terabyte? It is 650,000 high quality photos.
Should an independent shopkeeper worry about this? Yes and no. Yes, because the supermarkets and big suppliers will be spending lots of money investing in big data solutions and coming up with big data answers.
No, because the evidence assembled by Silver suggests that most predictions and forecasts miss the mark. “The volume of this information is increasing exponentially. But relatively little of this information is useful,” he writes.
Instead, what you have to do is to think probabilistically. It is a big word and means something like “use your common sense”. Silver provides some brilliant examples of how to behave – and a mathematical formula (Bayes’s theorem).
For example, what happens when you come home and find a mysterious pair of knickers in your husband’s underwear drawer? Most wives immediately assume 100 per cent that he is cheating on them. However, using the formula, the chances that he is cheating on you are shown to be 29 per cent.
Hold that thought. Silver moves on to discuss how a successful sports gambler operates and how to make money playing poker. These chapters are riveting: especially when you apply them to your business of retailing.
What are the odds of selling that outer of 24 packs of crisps? What are the odds of realising its profit potential? What factors come into play?
Silver used to make money at poker because there were lots of big fish (bad players) at the tables. Their losses helped him win more than $100,000 a year. However, once the big fish disappeared, he started losing so he cashed out his chips.
At 450 pages, The Signal and the Noise is an important book. It teaches people to be more careful about their decision making. “Information becomes knowledge only when it is placed in context,” Silver writes.
To do this well, you need to make lots of forecasts. Don’t bet your business on these forecasts at first. But the more forecasts you do, the more successful you will be. I strongly recommend this book to regular readers. It will make you humbler and better at your job.