the signal and the noise (sic)

[Please note: The word “Featured” on the left side above was NOT inserted by this blogger, but apparently was inserted by WordPress, and it cannot be removed. NO post in this blog is sponsored.]

The Book of the Week is “the signal and the noise (sic)” by nate silver (sic), published in 2012. In this volume, the author described in redundant and wordy terms, why human beings are so fallible in their predictions and forecasts (and explained the difference between the two). Basically, humans get distracted by noise, so they don’t zero in on the right signals in order to tell the future correctly.

Ironically, the author used less-than-ideal language in describing the epic failings of ratings-agencies in the 2008 financial crash. He should have pointed out that they could have mitigated, just a little, their false advertising by using better risk-assessment wording.

Silver wrote, “… trillions of dollars in investments that were rated as being almost completely safe instead turned out to be almost completely unsafe.” (Never mind the awkwardness of the word “being” in the middle of the sentence, or “it” in the middle of a sentence– so many recently published books have that kind of bad writing.) The ratings agencies should describe investments as “low-risk” or “high-risk” and use the adverbs “extremely” or “very” or “somewhat” or “slightly” as applicable, but never use the word safe.

Anyway, another irony was that the author appeared to be distracted by vast generalizations that were just noise– as cherry-picked data tend to be. He provided all sorts of line graphs and scads of data on housing bubbles. He cited a study on market prices of the “American home” completed by Robert Schiller and Karl Case that created an index based on a century’s worth of data– the years between 1896 and 1996, inclusive.

The research indicated that an inflation-adjusted home bought for $10,000 in 1896 would be worth $10,600 in 1996. Is that noise or what? Silver didn’t specify what “American home” meant. Anyhow, who would buy a home in 1896, and sell it in 1996?

Silver did admit that predictions and forecasts were less inaccurate when qualitative data supplemented statistical models. Worded facts are considerations that add real-world conditions because numbers never tell the full story in complex situations, which are dynamic.

Incidentally, at the book’s writing, he had had success in making predictions in professional baseball because: 1) an excessive amount of data on it had been collected, and 2) he claimed its rules didn’t change. The latter is not true anymore. And besides, performance-enhancing drugs, not to mention new stadiums– among other factors– have put new noise and signals in baseball statistics.

The author pointed out that more data actually made for worse accuracy in predictions in many areas of life. Technology in the form of software that can process scads and scads of data in record time has improved humans’ ability to specifically forecast severe weather, but not earthquakes. As an aside– in any area that involves linguistics, technology is overrated. A chatbot cannot comprehend complex concepts and nuanced language (like sarcasm, irony and idioms). American English is especially fraught with words that have multiple meanings, so it is highly contextual.

There are still financial crashes, gamblers who lose big-time, and “experts” who can’t modify conditions to improve the economy with certainty. Incidentally, as is well known, more and more, daily life in America has been infiltrated by politics.

Read the book to learn about futuristic pronouncements of: television pundits, professional-sports commentators and gamblers, seismologists, chess software, national-security advisers, poker players, and many others.