January 28th 2013

the signal and the noise. why so many predictions fail – but some don’t. nate silver.

English: Icon representing Bayesian statistics

the signal and the noise. why so many predictions fail – but some don’t. nate silver.2012. ISBN 9781594204111. A hard book to classify. Better than Freakonomics and others of that ilk, this book has some pretty hard math and science behind it.  You will learn how gamblers, chess, baseball, poker, the stock market, climatology, weather, terrorism,  and earthquake forecasting schemas work and where they are weak.  As well you get some great advice on Bayesian mathematics, how to reduce  errors in your forecasts ( but never eliminate them, nor do you eliminate your bias).

Great lines (From many sources) :

  • The future seen as sprinkles of probability
  • With more evidence we can get closer and closer to the truth
  • The majority of scientific hypothesis deemed to be true are actually false
  • Big Data will make us more prone to failure
  • most data is noise
  • combining forecasts could reduce error 15-20%
  • the stock market can stay irrational longer than you can stay solvent
  • unknown unknowns are gaps in our knowledge – but we do not know they exist

Valuable book for your library – this is required reading to reset your brain away from common fallacies. Very readable, good for that trip to the UK and back . My only gripe is that the paper quality is too low – for my eyes I need a brighter white to  see the text.

Similar Posts:

Leave a Reply