Nate Silver is the wunderkind of American media. On his blog FiveThirtyEight he managed to predict the outcome of the last US presidential election with a precision never seen before. Silver bases his estimation on the works of Thomas Bayes and his „Essay towards solving a Problem in the Doctrine of Chances“. Silver’s success with this approach is astonishing and casts a shadow on the work of political pundits and experts. In “Signal and the Noise” he explains how he does that.

*„This book is emphatically against the nihilistic viewpoint that there is no objective truth. It asserts, that a belief in the objective truth – and a commitment to pursuing it – is the first prerequisite of making better predictions. The forecaster’s next commitment is to realize that she perceives it imperfectly.“*

Silver opens his book with the US subprime crisis and the over-confidence of financial forecasters. He calls it a „catastrophic failure of prediction“ and in fact, the custodian of forecasts – the rating agencies – are jobless today. They lost basically all their power. And financial analysis is a rather quantitative business. The scientists that had to deal with the 2009 Italian earthquakes in L’Aquila – another of Silver’s examples – had a much more difficult environment to drew predictions from.

What do these two examples have in common? They both live in „Extremistan“, how Nissam Nicholas Taleb called the field of social analysis, that exceed the bell-shaped curve invented by Carl Friedrich Gauß. It means, that there are some very rare events, that exceed everything we have seen so far – they follow the so called power laws – also mentioned by Philipp Ball and Paul Ormerod. One recent example is the floods in Eastern Germany that render obsolete the improved dikes build after the record flood in 2002. The main message is: whatever you have seen, the next crisis might even be bigger – statistically, nothing can be excluded.

Silver acknowledges that we are living in a world of randomness but also in a world of growing data and information. The question is, how to make sense out of that information given the limited capabilities of our brains and our analytical capacities – his answer is Bayes Theorem: „In its most basic form, it is just an algebraic expression with three known variables and one unknown one. But this simple formula can lead to vast predictive insights.“ In fact, it is not about the formula, but about it’s ability to offer a handrail for analytical thoughts. It prevents us from drifting into intuitive decision-making.

„Bayes’s theorem is concerned with conditional probability. That is, it tells us the probability that a theory or hypothesis is true if some event has happened.“ It takes new information into account and correlates it with the already existing probability. To apply the theorem you need a hypothesis and a percentage share estimation on how sure you are about your hypothesis. If new information comes into play you assign a probability to it how much this favors your hypothesis and how much the information contrasts this hypothesis. The ratio is the new „posterior probabilty“. If another bit of information gets available you run the same operation again – with the posterior probability being the prior probability now.

What is refreshing – apart from Silver’s success – is his position between intuition and Big Data: „With information and processing power increasing at exponential rates, it may be time to develop a healthier attitude toward computers and what they might accomplish for us. Technology is beneficial as labor-saving device, but we should not expect machines to do our thinking for us.“ To Silver computers manage to structure complex data, but reasoning is still stronger that algorithms.

It makes sense to watch the movie „Moneyball“ before reading Nate Silver – it portraits the 2001 team of the Oakland Athletics. The first team that valued numbers higher than the intuition of the scouts. Silver also made his first experiences in reasoning comparing baseball stats. From his descriptions it is pretty obvious what a geeky and time-consuming leisure Bayes Theorem can be – but Silver’s success proves him right. „The signal and the noise“ is a look behind the scenes of the most promising approach to predictive analysis.

*„The Signal and the Noise – Why so many predictions fail and some don’t“ from Nate Silver. Published in 2o12 by Penguin Books. 27,95 US-Dollar. *

http://www.us.penguingroup.com/static/pages/features/the_signal_and_the_noise.html