Ask HN: Books with a high signal to noise ratio?

260 points by thewarrior ↗ HN
I like reading non fiction but a lot of it is like pop psychology with not a lot of informational content.

I've picked up two books recently and I've learnt a lot :

The Selfish Gene By Richard Dawkins

The 10 Day MBA By Steven Silbiger

There's very little fluff in these books and you learn something useful on almost every other page. I'd like to read more such books , especially on technical subjects.

212 comments

[ 4.8 ms ] story [ 61.5 ms ] thread
The Open Organization, by Jim Whitehurse
(comment deleted)
1. Walter Lewin's For the Love of Physics: From the End of the Rainbow to the Edge Of Time - A Journey Through the Wonders of Physics

2. Subroto Bagchi’s MBA At 16: A Teenager’s Guide To Business

3. Jugaad Innovation: A frugal and flexible approach to innovation for the 21st century by Navi Radjou, Jaideep Prabhu, and Simone Ahuja

Reclaiming History by Vincent Bugliosi.
Thinking, Fast and Slow by Daniel Kahneman.
I've finally started reading the book after reading about it on every corner. I'm halfway through and it's really good. A lot of information without unneeded fluff.
I started reading this book and am half way through and, although I'm not finished, I would have to disagree with you. This book is mostly pop-psychology with overstated implications and reliance on low-stake experiments that can not always be replicated. Also, the book glosses over actual results by saying "the participants were more likely to x than y". How much more likely?

A typical example of psychological discovery discussed in the book: 1. Ask someone to give 3 or 10 examples of when they were assertive 2. Ask someone to grade themselves on how assertive they are

And the study shows that people who were asked for 3 examples of how assertive they are were more likely to grade themselves as more assertive that those who were asked for 10 examples due to recollection bias.

I mean, I could see how this is sort of interesting but it's such low stakes. How would anyone be able to grade themselves from 1 - 10 on how assertive they are? What does that even mean? What implications does this have on anything?

I don't know. Psych experiments that are so glorified in the book seem to me to be too convenient and a form of story telling, which is fine except for the troubling implications that our cognitive biases and inefficiencies somehow trump our free will and freedom. But then again, I have my own biases.

It's supposed to be an accessible version of "Heuristics and Biases". Saying Kahneman is pop psych... Well if you think so, then we should consider anything in psychology to be "pop".
Another issue I take with the work is the certainty in talking about the analysis of his experiments. Phrases like "we now know..." are used often.

For instance, one experiment shows that frowning can have effects on your disposition. As an experiment, a subject is told to hold a pencil in their mouth in a certain way as to unknowingly produce a frown or smile and then do some task. The difference between behaviors is then attributed to smiling or frowning, as though the only thing going on with a pencil awkwardly in your mouth is the smiling or frowning.

Maybe I do view all psychology in a negative way but I imagine I find it distasteful the same way (I imagine) most people find the study of IQ differences among races distasteful. Is this a valid field of study? I don't know and I don't care to know because the racist overtones are so strong.

Unfortunately psychology, especially 'pop' psychology has been used to deny people their free will and restrict freedoms.

It goes along with an overall trend of the sciences to rely more and more on statistical methods which I find troubling.

Would you find What Intelligence Tests Miss - Psychology of Rational Thought by Keith Stanovich more convincing?
Oh give me a break, it's probably one of the most well corroborated pieces of research in that tradition.
Along the same lines is Moral Tribes by Joshua Greene. I learned a huge amount about the human condition from that book.
Most useful book I've ever read, period. Changed the way I think about decisions and influenced how I think about building products.
The Idea Factory, by Pepper White
Literary Theory: An Introduction, Terry Eagleton
Just to be clear, the simplest gloss on Eagleton's landmark essay (it's really too short to be a book) is that it's a takedown of belles lettres-style criticism.

Loosely speaking, Eagleton is best known for Marxist literary theory (dialectical materialism).

[Edit: removed html-style tags]

Good idea. Too many these books these days read like a long-form magazine articles (or blog posts) that have been stuffed with fluff to get to a book-length.
God, yes. I still can't get over "the no asshole rule". The entire book makes literally two points: "Don't be an asshole and don't tolerate assholes". It's stuffed with so many factoids and fallacies to belabour those points that I'm still mad at the author for wasting 6 hours of my day.
So the author was... kind of an asshole?
No, not at all, just terribly verbose and opportunistic, I guess.
Quantum Computing Since Democritus, by Scott Aaronson. It covers an astounding amount of stuff -- including complexity theory, the nature of randomness and information, and quantum information theory -- into less than 400 pages.
It's a great book (I'm more than halfway into it) but definitely not for laymen. I followed one quantum computing class as well as having had many QM classes and I find myself having to read a lot of background information to understand Aaronson. He writes well, but I lacked the proper computability knowledge to really enjoy the book in one go.
Within computer science, I found "Introduction to the Theory of Computation" by Sipser to cover a lot of ground in a concise and comprehensible manner.
(comment deleted)
"Technical subjects" is too broad to get started. But if you are serious about wanting a high signal to noise ratio, you want textbooks and monographs. Landau and Lifshitz' series on theoretical physics; Knuth's Art of Computer Programming; that kind of thing.
I don't think a mega-compilation of the Western Canon is what OP is looking for...
Eh, they certainly tend to avoid fluff, and by definition, until you get to Marx, James, Freud you're not going to find "pop psychology"! A lot of them are very much worth reading, I personally would recommend:

Homer and a few Greek plays

Sample a bit of the great story teller Herodotus, then read the birth of historiography in Thucydides' History of the Peloponnesian War, which by itself is also very interesting and important (wonder why the Founder of the US didn't like direct democracy? There are very important object lessons in it).

Surely Plato and Aristotle deserve some attention! The contents of the latter's Rhetoric is essential for when you can't reach people with dialectic.

Euclid's Elements is still about as good as you can get for what it teaches.

Plutarch is great, but I really like that period of history. To it I would add reading some of the earlier bits of Livy.

Read, or better yet listen to audio of a few of Geoffrey Chaucer's Canterbury Tales, out loud you can follow their Middle English.

Machiavelli's The Prince is still damned good, and a landmark in talking about politics as it is, not as how people would like it to be.

Shakespeare surely needs some attention by English speakers. Swift's Gulliver's Travels were amusing when I read them in their original, and obviously very influential.

So, yeah, check out some of the classics.

Process and Reality: An Essay in Cosmology by Alfred North Whitehead
The Art of Eating, by MFK Fisher: http://www.amazon.com/The-Art-Eating-Anniversary-Edition/dp/...

I love food, and I love the anthropology of food. The writing here is engaging and entertaining, but the content is quite interesting if you are at all curious about mid-century French cuisine from the perspective of an American ex-pat chef.

I second this. I have read every page of MFK Fisher that I can find. I have never regretted a single page. Not just a food writer; she possesses a poet's eye for the human condition, and genius-level writing skills.
I think if you're really looking to get a sense of how the world works, the best stuff to read are journalistic accounts and histories. Just pick a subject your interested in, do a bit of googling on well regarded accounts of that subject, than go to town.

The reason I think these are more valuable than the pop psychology/business airport books is they don't operate under the pretense that the world's great truths can be boiled down to 240 pages. Rather, learning about people's experiences and stories on there own terms helps you develop a much more nuanced worldview.

For example, I'm reading The Battle Cry of Freedom, an overview of the Civil War, and it's astounding how much more insight a book about something 150 years ago offers into today's society than just about any of the Gladwell genre stuff.

I've also been reading The Battle Cry of Freedom and it's been amazing. The analysis and insights the author can pack into every single page is just incredible. And all those political, social, and historical insights have just as much applicability today. It really has been essential reading for understanding the major fault lines that have defined American life, and continue to do so in major ways.
I've been (slowly) working through the series (Oxford History of the United States) and want to put a mention in for "What Hath God Wrought", covering 1815-1848. Similar in scope and depth to Battle Cry, gave me similar feelings of "How can a book so long feel like it's barely abel to get all this information in."

Thats the period I was least interested in going in, but have ranked as the best History book I've read.

(comment deleted)
I also think histories and biographies can be great for high S/N nonfiction. Good examples that I really enjoyed include "Ignition" by John D. Clark and "Excuse me sir, would you like to buy a kilo of isopropyl bromide?" by Max Gergel.
The best book I've read in the last 5 years is David Simon's, "Homicide: A Year on the Killing Streets"...if you've watched The Wire, half of its material is derived from this real life account of being embedded in Baltimore's homicide unit: https://books.google.com/books/about/Homicide.html?id=N8LS0b...

All of Atul Gawande's books, notably:

The Checklist Manifesto [1]: http://atulgawande.com/book/the-checklist-manifesto/

Better: A Surgeon's Notes on Performance: http://atulgawande.com/book/better/

[1] You can read the essay that this book is based on in the New Yorker: http://www.newyorker.com/magazine/2007/12/10/the-checklist

Homicide is one of those books I wish I could read again for the first time. I can think of no higher praise.
Homicide is one of those books I wish I could read again for the first time. I can think of no higher praise.
'Checklist Manifesto' is great, but I don't think it's high in SVN. It repeats itself a lot to drive home the point that using checklists is good, even (or especially) for experts in their field.
Reading that article is outrageous. Save thousands of lives a year in every single hospital? And what is the cost? About 2 to 3 million dollars? It would be an absolute bargain at 10 Billion dollars.

But we are not doing it. Why? Because we misallocate the founds.

I just did the calculations, if we ignore the money saved and assume the projects total expenses are going to be 4 million, then it only needs to save at least 1198 people to beat the top rated charity on GiveWell (Against Malaria foundation).

Looking at the data (http://www.nejm.org/doi/full/10.1056/NEJMoa1306801) there were 721,800 what they call health care associated infections in 2011 (I couldn't find newer data). If the rate goes down at what the article suggest, (75%) then we will save 541350 people a year using this method whereas it would cost more than 1.8 Billion dollars to save the same amount of lives using the Against Malaria foundation.

This is almost 500 times as effective!

How do we get this knowledge to people who have 2 million dollars to spare?

It should also be pointed out that this book was the basis for the show Homicide: Life on the Street, which Simon consulted on and later produced. That's what got Simon into television and eventually writing The Wire. (That said, Simon felt the show did not closely enough reflect the book, which is probably what fed into the creation of The Wire. Still, for a network TV drama, Homicide was pretty outstanding.)
Here are 3 that I particularly liked in the last few years:

The Information, James Gleick

A Universe from Nothing, Lawrence Krauss

Abundance, Peter Diamandis

I was surprised to see a recommdation for The Information. Gleick seems the epitome of a rambling author who says almost nothing. I made it through Chaos, but could not get through Feynman or The Information. This kind of InfoJunk is what the OP is trying to avoid, I think.
Maybe you are right, it is not in the same category as The Selfish Gene. However, I thought the parts about the early history of computing were written very well, as someone who did not know much about it before.

I was also thinking that a non-fiction book with no fluff and no injection of personality/flair from the author is a textbook. There are many outstanding textbooks (like Molecular Biology of the Cell) which offer pure information, but I think commercial non-fiction books are aimed at a more general audience. The author needs to fluff a book up a bit to make the material approachable. I also wonder how some non-fiction books would read if untouched by editors.

I disagree. The Information has a broad scope, so there is a lot of content but none of which I found to be InfoJunk. I thought it was very interesting and just kept on reading.
If history is of any interest, try Asimov's Chronology of the World, by Isaac Asimov (duh). It's ordered chronologically, and describes in concise and fact-rich passages what was happening during a given span of time in each region of the world.

Asimov was also a master at nonfiction writing for laypeople. Most of his books were nonfiction (they were easier for him to write).

Deep Simplicity: Bringing Order to Chaos and Complexity
Functional programming in Scala by Paul Chiusano and Rúnar Bjarnason. Programming Erlang: Software for a Concurrent World by Joe Armstrong. Even if you have no desire to learn Scala, Erlang, or functional programming, I still think most people would enjoy these books. Every page of these books contain a golden nugget.
Gödel, Escher, Bach, by Douglas Hofstadter.
If you consider extravagance as noise, this book will have a terrible SNR. (I think it's nonetheless worth reading. It's basically an introduction to the theory of computation with fables and plenty of illustrating examples by Escher and Bach (both of which have unknowingly used important concepts from computation in their works). It is definitely much easier to read after having attended to a theory of computation I lecture.)
We have access to so much great work on the topic of computation today thanks to the internet, and especially because of work in computer languages and research over the last 20 years, but GEB was published in 1979. At the time it was ground-breaking, making work that really only a handful of people knew about and were interested in accessible to many.
Completely disagree. GEB is the opposite of what OP was asking for. It takes pages to make points that can be done far more economically and concisely.
That was my experience as well. Another poster mentioned Quantum Computing Since Democritus by Aaronson. The latter seems more technically difficult but I enjoye it more than GED. I haven't been able to "get into" GED in the last 3 times I have started reading it. It's just long and rambling (to me). I don't "get it", but maybe I will one day, the book's still on my bookshelf...
For me it was

  Guns, Germs, and Steel: The Fates of Human Societies by Jared Diamond.
It blew my mind on almost every section! I loved it I suppose coz it answered a lot of questions I'd growing up and a bit more. I've recommended it to everyone in my family.

One critique I received was that there is repetition of ideas but I think its because primarily humans carry past successes anf failures so most of the changes are evolutionary to what worked previously and second this book is akin to a thesis and every chapter is essential to building up the case.

By far the best non-fiction I've read.

(comment deleted)
Yes the repetition was noticeable, I would typically start reading a paragraph, realize that it was a point made previously and end up skipping a whole page.
While no doubt Guns, Germs, and Steel is a good work, I will be hesitant to subscribe to ideas it presents, carte blanche. Please see this reddit discussion[1]

[1] https://www.reddit.com/r/AskAnthropology/comments/1rzm07/wha...

I'm sure I find books to read in this "Ask HN" but your comment is exactly what's the issue with it.

So the OP asks for high signal to noise ratio. There comes a recommendation, and then I read some good discussion on Reddit on how the book is actually kind of wrong.

This begs the question: What's "high signal to noise" anyways? Someone who has read 10 books on a subject will find barely anything new. One who just starts out might find "gems" on every other page.

And also: If a recommendation here in this "Ask HN" is given and then debunked as mediocre, how sure can one be to actually find a book with high signal to noise without reading the book?

In that sense: Thanks for your comment. In-depth discussions like the one on Reddit is really necessary, because I, as a starter, have no clue how to evaluate a book. Every recommendation should probably come with a lengthy discussion about its accuracy by people who know the subject :)

I kind of agree with you "high signal to noise" take. It is highly subjective. Most of the times, going with populistic choice seems to be way forward. Unfortunately we have very short time. Personally very liberal estimation of my reading prowess, I don't think I will read more than another 500 books in my life time.

My emphasis with my comment is to read books, ideas with healthy dose of skepticism, rather than worshipping it as gospel.

The first link in that reddit comment points to [1] which is a critique of another Diamond's book. But if you let me give two comments on it:

(1) In the first sentences, the author (a PhD student in history) cannot really hide his irritation that someone with a background in animal physiology has written a popular book about history. I have seen this ad hominem several times about Jared Diamond. Somehow those historians cannot stand that someone with no degree in history has written such a popular book.

(2) "Guns, Germs, and Steel attacked the notion that racial superiority explained Western global pre-eminence, a view taken seriously by almost no one who’s taken seriously"

Actually, what Guns, Germs, and Steel did, was to provide a first comprehensive theory (that I have heard of, anyway) that explains why Europe, and not someone else, rose to dominance. This is the merit. The main merit is not that it, as a side product, discredited the racially based theories.

Failing to understand this difference does not speak highly of the writer of the review.

[1] http://www.columbia.edu/~saw2156/HunterBlatherer.pdf