Some light geek reading.....

[woodgeek]

My geeky FIL gave me the Nate Silver book:

http://www.amazon.com/The-Signal-Noise-Many-Predictions/dp/159420411X

and I just finished it.

I love Nate in blog form, and I thought the book started a little slow and dry (with no surprises to anyone with a good working knowledge of statistics). After the first few chapters, though, I found myself really digging the material and his thesis. Bottom line, we live in an age of Big Data, and are running again and again into the perils and pitfalls of misinterpreting noisy data. There are some wins, like better Hurricane forecasts, but in general, the information age seems to be harder to deal with than we might have hoped naively 20 years ago.

Overall, I think he is going for a book in the style of Michael Lewis' work, and pulls it off. He is not as funny as ML, but introduces us to interesting characters as the chapters unfold to tell his story. He presents more mathematical material along the way....overall it reads as a geekier version of the 'big short'.

He covers a lot of topics....financial crisis, weather forecasting, internet poker, computer chess, day trading, national security, climate change and more (I skipped the moneyball chapter). Seems a refreshingly non-partisan take to me...let's see what the data show, and if the result is ambiguous, let's agree we don't know the right answer. Of course, he discusses the world as it is, not policy at all.

The climate change chapter was IMO esp well done, and changed how I thought about it (not easy).

I'd recommend it to anyone with an interest in the subject matter.

 

[begreen]

Thank for the review. I'm interested to hear how he changed your climate change view. I'm stuck in the middle of the book on Google. It kept putting me to sleep so I've read a couple other interesting books in the meantime, both fiction, but well written. Just finished Ordinary Wolves last night.

 

[woodgeek]

The book def starts slow/dull. Then he hits his stride a ways in.

NS underscored that many researchers building climate models themselves do not rate them well, due to systematic uncertainties within them. He also convinced me that the statistical 'significance' of the warming signal over the last century or two is greater than I had appreciated. Overall, a thoughtful and skeptical take (in the positive sense of skeptic, not 'denialist') that still leaves little room for doubt for the AGW phenomenon.

 

[fossil]

Glad to read that post, geek, thanks. I've had Silver's book on my shelf in the "to be read" section for some time, it just hasn't made to the top of the queue yet. Or even the on-deck circle. Maybe I'll give it a bump. Rick

 

[Doug MacIVER]

My geeky FIL gave me the Nate Silver book:

http://www.amazon.com/The-Signal-Noise-Many-Predictions/dp/159420411X

and I just finished it.

I love Nate in blog form, and I thought the book started a little slow and dry (with no surprises to anyone with a good working knowledge of statistics). After the first few chapters, though, I found myself really digging the material and his thesis. Bottom line, we live in an age of Big Data, and are running again and again into the perils and pitfalls of misinterpreting noisy data. There are some wins, like better Hurricane forecasts, but in general, the information age seems to be harder to deal with than we might have hoped naively 20 years ago.

Overall, I think he is going for a book in the style of Michael Lewis' work, and pulls it off. He is not as funny as ML, but introduces us to interesting characters as the chapters unfold to tell his story. He presents more mathematical material along the way....overall it reads as a geekier version of the 'big short'.

He covers a lot of topics....financial crisis, weather forecasting, internet poker, computer chess, day trading, national security, climate change and more (I skipped the moneyball chapter). Seems a refreshingly non-partisan take to me...let's see what the data show, and if the result is ambiguous, let's agree we don't know the right answer. Of course, he discusses the world as it is, not policy at all.

The climate change chapter was IMO esp well done, and changed how I thought about it (not easy).

I'd recommend it to anyone with an interest in the subject matter.

you might feel better after you read this,http://www.huffingtonpost.com/michael-e-mann/nate-silver-climate-change_b_1909482.html .beware the dark side geek, have a day.

 

[woodgeek]

I like Mann, and I like Nate, and I have made a conscious decision never to read huffpo (b/c of the ample, dangerous non-factual content). Following the link, it seems that Mann read a different book from me. In my reading, Nate was deferential and accepting of Mann, and Mann is now calling him out as a denier (!) b/c he also talked to some deniers. In my reading, NS seemed to be poking holes in the deniers logic left and right, and suggested that they were out of their depth applying mathematical models from econ to the climate problem. But apparently that was not explicit enough for Mann.

 

[woodgeek]

Glad to read that post, geek, thanks. I've had Silver's book on my shelf in the "to be read" section for some time, it just hasn't made to the top of the queue yet. Or even the on-deck circle. Maybe I'll give it a bump. Rick

I guess the rule with Tolkien is that you can skip the songs. If you get bogged down, I think you can skip the parts that are uninteresting.

 

[Doug MacIVER]

I like Mann, and I like Nate, and I have made a conscious decision never to read huffpo (b/c of the ample, dangerous non-factual content). Following the link, it seems that Mann read a different book from me. In my reading, Nate was deferential and accepting of Mann, and Mann is now calling him out as a denier (!) b/c he also talked to some deniers. In my reading, NS seemed to be poking holes in the deniers logic left and right, and suggested that they were out of their depth applying mathematical models from econ to the climate problem. But apparently that was not explicit enough for Mann.

every thing I read about mann is about as you put it. don't agree and follow him you are wrong, that's it. he recently said "proof is for mathematical theorems and alcoholic beverages. it's not for science." take what you will from that, it's from wuwt.

 

[woodgeek]

every thing I read about mann is about as you put it. don't agree and follow him you are wrong, that's it. he recently said "proof is for mathematical theorems and alcoholic beverages. it's not for science." take what you will from that, it's from wuwt.

Agreed. Before I judge, though, I would have to walk in his shoes...he has been hounded and excoriated by loonies for decades now....if I were in that situation, I would prob go hide under a rock and never come out.

As for the 'proof' comment, as a scientist myself I would have to agree with it. Very few things in science are 'proven', and in important areas (including climate, medicine, etc) getting something like proof is essentially impossible.

 

[Doug MacIVER]

Agreed. Before I judge, though, I would have to walk in his shoes...he has been hounded and excoriated by loonies for decades now....if I were in that situation, I would prob go hide under a rock and never come out.

As for the 'proof' comment, as a scientist myself I would have to agree with it. Very few things in science are 'proven', and in important areas (including climate, medicine, etc) getting something like proof is essentially impossible.

i'm obviously not a scientist ,just interested in the subject. he would agree with you as he speaks of science as credible theories and best explanations. thanks for discussion and lead on a read.

 

[woodgeek]

Yar. The thing that makes me laugh is when scientists are portrayed as protecting the existing 'law' against all assaults from new data, always reticent to try new things. Fact is that the best way for any young scientist to make a name for him/herself is to find data that don't fit and to find a new explanation. We are in fact rewarded and respected in proportion to how much prior thought we overturn!