We’re often told we should learn from our mistakes. However, it is not always easy to identify the most persistent and deep-rooted sources of error. Late last year, an important bit of work was published that attempts to shed more light on this issue.
The paper, titled 'Bias, Information, Noise: The BIN Model of Forecasting', is the work of Philip Tetlock, co-author of influential book Superforecasters, along with his University of Pennsylvania colleague Barbara Mellers, plus Marat Salikhov of Yale and Ville Satopaa of INSEAD. What the researchers found about the sources of forecasting error is surprising and also encouraging for anyone interested in learning how to make better investment decisions (the subject of this week’s cover feature).
There’s no surprise that we have trouble predicting the future. As the BIN paper states at its outset: “Forecasters must often work under less-than-optimal conditions: too little or too much data as well as data of uncertain or varying reliability… In messy real-world situations, forecasters are bound to make mistakes – and misinterpret signals from the environment about what is likely to happen next.”