excellent for anyone interested in statistics or economics ( )

http://pro-libertate.net/20100318/115-read-super-crunchers-why-thinking-numbers-... ( )

Ironically, I'm a bit of a super cruncher at work. For years, I had come at peace that my company can aggregate all the numbers they want, but I am impossible to market to. I am beyond a number. I can not be categorized. Or am I?

Can that spark of human inspiration make better decisions more than an evolving formula? I'm leaning toward the idea that numbers can better make at least as decision as humans when it comes to clear rule sets (like chess) and large amounts of data. Seriously, people, can you really understand what $100 billion is and how to appropriate it "fairly"?

Perhaps the more interesting battle would be humans vs. computer playing poker (which computers lost).

When it comes to meaningful questions about .. the meaning of life or does he/she love me? How should I raise my child? Or why do bad things happen to good people? The formula falls short.

Still, a very good read.







( )

Read Dolly's review, too: https://www.goodreads.com/review/show/409145842

I've noticed a theme in my reading lately. I'm apparently getting fed up with learning just how poorly our intuitions work ([b:Incognito: The Secret Lives of the Brain|9827912|Incognito The Secret Lives of the Brain|David Eagleman|https://d202m5krfqbpi5.cloudfront.net/books/1348669116s/9827912.jpg|14423132], [b:Predictably Irrational: The Hidden Forces That Shape Our Decisions|1713426|Predictably Irrational The Hidden Forces That Shape Our Decisions|Dan Ariely|https://d202m5krfqbpi5.cloudfront.net/books/1255573980s/1713426.jpg|3074803], etc.) and am looking for ideas about how to overcome those fallibilities. So I read a couple of books by Dr. [a:Nortin M. Hadler|490348|Nortin M. Hadler|https://www.goodreads.com/assets/nophoto/user/u_50x66-d9f6a4a5badfda0f69e70cc94d962125.png] and learned about how much medicine & health care rely on bad science. And then I read [b:Bad Science|3272165|Bad Science|Ben Goldacre|https://d202m5krfqbpi5.cloudfront.net/books/1327251503s/3272165.jpg|3308349], which turned out to focus mainly on health care.

And then I read Ayres. And I have hope. With more and more data available because of the digital revolution, and with more and more people becoming comfortable with using data & technology and with international collaborative efforts (think, for example, about the database of books and reviews here on goodreads), we will have more good science to help us make more rational decisions. For example, evidence-based medicine will become a given, not a controversy, as more studies are published online, and patients can check for themselves which treatments have been demonstrated to be effective.

Ayres' book is more inclusive than just healthcare, though. Gearheads are having their turn in the sun in baseball with sabermetrics, in wine with Orley Ashenfelter and his Journal of Wine Economics," and in travel with Farecast (which has apparently not done so well since being sold to Bing, but oh well).

The film industry has benefited from Epagogix. "There will always be legitimate and unresolvable tensions between artistic and commercial goals. However, there should be no disagreement that it's a tragedy to mistakenly interfere.... Epagogix is moving us toward evidence-based interference."

Businesses of all kinds could benefit from hiring gearheads. Jo-Ann, the sewing & craft chain, decided to test the sale "Buy two sewing machines and save 10%." Seems silly, no? But what happened is that people were effectively becoming sales agents and recruiting friends to buy the second machine, thus each saving money, with a win-win all around. How did Jo-Ann figure out what was happening? By counting click-throughs comparing this sale to another iin randomized testing. Ayres even tested the title of the book during pre-publication - the current title clearly beat out his first choice: "The End of Intuition." (You can try, if the link still works, Lulu.com/titlescorer.)

Of course, Ayres does take pains to point out that intuition has value. He does try really hard to find fields that would be better off letting humans continue to solve problems, instead of computers. But mostly he fails. What humans, and their intuition, are still needed for is to set up the tests. We are necessary because computers cannot figure out what questions to ask. You (I hope) remember from your science fairs at school that the problem must be defined, and the hypothesis stated clearly, before tests can be run - and this still holds true.

Too much to explain here, so read the book: Learn the 2SD rule when trying to come up with estimates. Understanding standard deviation (and it's really not that hard) will help you come up with a more accurate estimate of the mean. Also, learn the bit about Bayes' equation so you can understand better "how to update an initial probability given a new piece of evidence."

Other links to try. For healthcare: Infotriever, DynaMed, and FIRSTConsult. For Ayres' datasets: www.law.yale.edu/ayres/.

All this good information is presented in a book that's highly readable, even entertaining. I will look for more by the author.


" ( )

A must read for anyone wanting to understand better how important decision are made. ( )