The best specialist books are those that immediately impact the reader’s behaviour. I just finished Superforecasting: The Art and Science of Prediction by Philip E. Telock and Dan Gardner – and experienced this very effect.
As the title suggests, the book investigates and explains what qualities and skills are needed to make accurate predictions. The insights are based on a large-scale forecasting experiment conducted by one of the two authors and involving hundreds of forecasters.
Being able to excel in forecasting can be extremely valuable. In the field of digital technology, forecasting is a preferred activity by many. Pundits, analysts, entrepreneurs and everyone only slightly affiliated with the industry is constantly trying to predict the future. For financial gains through wise strategic decisions involving foresight, in order to build a professional reputation as visionary, or – unfortunately – in order to advance personal interests or the interests of “the forecaster’s tribe” (as the authors of the book put it).
Around each turn of a year, blogs and technology websites are loaded with predictions about possible events that will shape the digital world and economy within the upcoming 12 months. I try to resist the urge to participate, even though I am as weak as everyone to the temptations of predicting. In my mind I am constantly creating forecasts, which I find to be a great mental exercise.
As I mentioned, after reading the book, I instantly changed my forecasting behaviour: Instead of only using words to describe the likelihood of things, I now try to come up with percentage numbers. Words are very bad tools for precise forecasts, at least for those that are meant to avoid ambiguity. Philip E. Telock and Dan Gardner explain this by pointing to a widely mocked quote by Steve Ballmer. The former Microsoft CEO infamously stated in 2007 that “there’s no chance that the iPhone is going to get any significant market share”. He then predicted that the iPhone might achieve a market share of 2% or 3%.
He clearly underestimated the iPhone, no question there. The quarterly share of iOS-devices among all shipped smartphones globally usually hovers somewhere between 10 and 20 %, on average closer to 10 % than to 20 %. So that’s about 10 percent points more than what Ballmer publicly predicted. But does the iPhone have a “significant market share”? That’s a tougher question. It depends on how you look at it. From Ballmer’s point of view, a “significant market share” is something north of 60 % of shipments, as he suggested in the same interview. That’s the typical former Microsoft way to look at things. With that perspective in mind, Ballmer actually doesn’t deserved to be mocked that hard for his forecast.
When it comes to the financial performance, the situation is different, of course. Looking at the profits in the smartphone market, the iPhone completely rules, pocketing 94 % of the global smartphone profits. That quality, however, was never questioned by Ballmer. He even stated: “They (Apple) may make a lot of money (with the iPhone).”
Ballmer’s forecast was off by some measure. But it wasn’t as bad as it has been made sound afterwards. One cause for that was the public focus on the expression “significant market share”. An ambiguous statement which bloggers, journalists and analysts interpreted freely. As a serious, ambitious forecaster, you don’t want to give the audience the opportunity to misunderstand you. Which is why you should use percentage numbers instead of vague words.
Of course not every forecaster wants to remove the ambiguity, because that increases accountability. Everyone who has made public predictions knows the evil comfort that lies in hedging and in deliberately vague formulations. If the prediction turns out accurate, you reap the praise and credit. If it turns out wrong, you emphasize the ambiguity of your own prediction. It often works, but it is not sincere.
As predicting the future is so widespread (and partly necessary) in today’s fast changing world, this book is valuable for those who want to improve their forecasts, as well as for those who want to get better at evaluating the quality of other people’s forecasts.
The urge to predict the future – and how to do it right https://t.co/J0QgOi3eH4
— meshedsociety.com (@meshedsociety) January 19, 2016