Nick Silver, author of a great new book “The Signal and the Noise: Why So Many Predictions Fail – But Some Don’t,” credits Thomas Bayes, the 18th-century English churchman and pioneer of probability theory, as his “guiding light.” Justly so – Bayes laid the groundwork for ways to address the uncertainties that confound us. After more than 250 years following his death his methods are as applicable today
In its 6 October issue, The Economist review cites Silver’s assertion that “humans are overconfident in their predictive abilities; that they struggle to think in probabilistic terms and build models that do not allow for uncertainty.” It is time, once again, to resurrect Reverend Bayes.
In addition to failing to grapple effectively with uncertainty, many people think the more data they have, the better their predictive capabilities, when says Silver, “the more data you have, the harder it is to distinguish the useful sort (‘the signal’ of the book’s title) from the misleading or confusing (‘the noise’). This is particularly noticeable during the dozens (hundreds?) of political polls that hit us daily. I really resonated with Silver’s reference to Philip Tetlock, a psychologist who declared predictions of political pundits to be “mostly bunk.”
Says Silver, people should not get hung up on uncertainty and subjectivity, which are inevitable. Rather they should think about the future as gamblers do: “as speckles of probability” (think poker). Dealing effectively with uncertainty enables one to continually update his/her assessments of probability as new knowledge is gained, basically improving the odds around the “accuracy” of their predictions.
Reference: The Economist, 6 October 2012, pages 97-98.
My white paper “Forecasting under Uncertainty Using Portfolio Navigator” sets forth a practical application of probabilistic evaluation to forecasting the economic value of investments in new business opportunities.