Author Topic: Week 1: Intro and Chapter 1 discussion  (Read 2926 times)  

Offline Eltanin Publishing

  • Status: Scheherazade
  • *****
  • Posts: 1262
  • Charlotte, VT
  • Cathy Ryan - Editor-in-Chief
    • View Profile
    • Eltanin Publishing
Week 1: Intro and Chapter 1 discussion
« on: November 12, 2012, 06:39:04 AM »
Hi everyone! As I mentioned before, I've never led a book club before, so feel free to steer this in any direction you want.

Some thoughts on the intro:

  • I thought the history of the printing press was fascinating. Silver wrote,"exposure to so many new ideas was producing mass confusion. The amount of information was increasing much more rapidly than our understanding of what to do with it, or our ability to differentiate the useful information from the mistruths." (~ location 88) I think this is the overarching theme of the book - how to take the huge, huge amount of data (noise) that is available to us and discerning what the relevant, useful data (the signal) is.
  • "The story the data tells us is often the one we'd like to hear, and and we usually make sure that it has a happy ending." (~ loc 119) Data is data, just numbers, but we can present it, manipulate it, and use it (consciously or subconsciously) to support the conclusions we want to be true. It is nearly impossible to keep our own biases out of our analysis of the data. "The numbers have no way of speaking for themselves. We speak for them. We imbue them with meaning." (~loc 176) "We can never make perfectly objective predictions. They will always be tainted by our subjective point of view." (~loc 267).I believe the book focuses on the question of "How can we apply our judgment to the data - without succumbing to our biases?" (~loc 297)

Chapter 1:
  • First off, I am fairly confident that there is a typo in Figure 1-2 (~loc 473). The general idea is that if you're betting that 5 mortgages won't default, there is a greater risk if they are connected or correlated in some way - such as if all 5 houses are located in the same town (which could get hit by a natural disaster and wipe out all 5, or a factory in the town could close and many people in that town are out of work). So the risk, the probability of losing the bet, is higher with correlation than without. I believe the column headings "if defaults are perfectly correlated" and "if defaults are perfectly uncorrelated" should be switched.
  • There are many ideas in this chapter (feel free to bring up any of them). But the one that sticks in my head is the idea that risks we have considered, but think are no big deal, can be more dangerous that risks we haven't even thought of: "In this book I'll discuss 'unknown unknowns' - the risks that we are not even aware of. Perhaps the only greater threat is the risks we think we have a handle on, but don't." (~loc 441) "The ratings agencies quite explicitly considered the possibility that there was a housing bubble. They concluded, remarkably, that it would be no big deal." (~loc 429) And part of the problem is that "false confidence can be contagious." (~loc 441). Surely, some people, before buying into these deals, asked, "What if housing prices drop?" and were told, "Yes, we've thought of that, but don't worry - it won't be a problem." Can you think of other examples, in your life or society, where the downfall was a risk that was considered, but it wasn't thought to be a problem, or it was believed that the problem could be handled? I can think of one - the Titanic. The risk of hitting an iceberg had been considered, but engineers thought the hull was so strong that an iceberg couldn't cause a hole. Partially because of this, they didn't bother putting an adequate number of lifeboats on the ship.
  • Another interesting idea was the one of Risk vs. Uncertainty (~loc 490) and Confidence vs. Accuracy (~loc 505). There were some politicians a week ago who were very confident they would win...

Again, if you feel there is a more helpful way of posing ideas/questions and sparking discussion, feel free to let me know and/or jump in and guide the discussion, pose your own questions, etc.

Offline Ann in Arlington

  • Global Moderator
  • Status: Shakespeare
  • *****
  • Posts: 65435
  • Gender: Female
  • Arlington, VA
  • Laugh Loud. Run Fast. Be Kind.
    • View Profile
Re: Week 1: Intro and Chapter 1 discussion
« Reply #1 on: November 12, 2012, 07:56:02 AM »
The 'accuracy vs. precision' thing is something my physics prof in college drilled into us.  . . .  his point being that, when doing an experiment, even if all your data is consistent, if it's WAY outside what has been accepted for decades or more, it's probably smarter to consider that there might be something wrong with YOUR methodology or equipment -- maybe your thermometer is badly calibrated -- rather than assuming you've made some great new discovery -- it turns out water boils consistently at 105 C. :D 

My Kindles
Nautilus (Voyage)
Oasis 2016
AquaKindle
Godric's Hollow (Basic)
S7Edge (KApp for Android)
Galaxy (KApp for Android)
(Make your own reading bar) | Ann Von Hagel | Arlington, VA | kboards  MODERATOR

Offline The Hooded Claw

  • Status: Emily Dickinson
  • *******
  • Posts: 9493
  • Gender: Male
  • Oklahoma
  • Card-carrying melodrama villain
    • View Profile
    • All about me....<hee hee>
Re: Week 1: Intro and Chapter 1 discussion
« Reply #2 on: November 12, 2012, 02:10:50 PM »
I've never been part of one of these Klubs before, so I'm just winging this....

The thing that struck me about the start of the book is how much of this is related to stuff I've been reading lately.  In The Pleasure of Finding Things Out, one of the essays published was physicists Richard Feynman's "minority report" on the explosion of the Space Shuttle Challenger.  Feynman's critique pointed out that before the Challenger exploded, estimates of the risk of totally losing a space shuttle and crew ranged from as low as 1 in 100 to as high as 1 in 100,000.  The risky figures came mostly from engineers who presumably had the best scientific understanding of the risks, the higher figures came from managers who presumably had a less fine-tuned appreciation of risks, but did have a solid interest in having the shuttle program seen as safe and worth Congressional funding.  So guess which set of risk figures got reported to Congress and the press?   ::)

Also in The Pleasure of Finding Things Out, Feynman harped repeatedly in the different essays about the necessity to be able to entertain doubt and uncertainty in your mind without being so paralyzed by it that you are unable to continue.  Reminds me very much of the similar points Silver made.

It's a minor thing, but Silver mentioned "The Wicked Bible" as an example of a particularly problematic error, and that same volume was mentioned in "Wide as the Waters" about the history of the English-language Bible that I read earlier this year.  I'd never heard of The Wicked Bible before that, strange to have it pop up in two books on widely-differing topics read within a few months of each other.

Silver mentioned "The Schmakladic War" and I'd never heard of it.  And it sounded like a silly name of the sort you'd see used in a children's cartoon, but I looked it up in Wikipedia (using my PW, whee!) and it is indeed a real historical war.

Some recommended further-reading type ideas for anyone who wants to follow up on any of these things:

Extraordinary Popular Delusions and the Madness of Crowds--This book is over a century old, but is a fascinating read about investment fads that go disastrously wrong.  Very readable and interesting.  Covers stuff I've heard of but didn't really understand previously such as The South Sea Bubble and The Tulip Mania.

The Pleasure of Finding Things Out, by Richard Feynman--A collection of essays and presentations by Feynman.  Some verge into the geeky, but it is readable and interesting, and he emphasizes some ideas that are relevant to this chapter, notably that you've got to be able to admit the possibility you are in error.  One of the essays is his comments on the Challenger Disaster as I mentioned above, and it is extremely relevant to Silver's book.

A good practice I've read about (and tried to use myself) about checking to make sure you aren't ignoring inconvenient facts is to "look for disconfirming data."  Disconfirming sounds complicated, but isn't.  Our minds are programmed to instinctively search the sea of information for more information that confirms our biases or preferred solution.  One way to try to use this instinct in your favor is to stop and say "I am looking for stuff that will prove my choice wrong" and search the sea of data for things that will disprove, or at least argue against whatever belief or choice you are certain of.  Doing this yourself for your own choice supposedly reframes your bias so you "win" and are right if you can find info that debunks your ideas, but trying this for your own ideas is still awkward at best.  It's better if you can have an independent person with no emotional investment in the choice being investigated does that.  This is the idea behind "The Devil's Advocate" as supposedly used by Catholics to argue against making someone a saint.  I've actually tried this a couple of times at work, assigning someone to argue against a decision.  It is better than doing it yourself, but still tough for the person with the assignment, both because they probably have bought into the groupthink that lead to the decision, and in any case many people are nervous about going up against the group or especially the boss, even when they are told it is their job assignment to do so!

My mind was boggled by the figure of $20,000 in modern dollars for a handwritten book before printing was around!

I think Cathy is right about the column headings in figure 1-2 being switched.  I didn't catch that when I read the book.  But she's correct, if there's a five percent chance of failure, and the defaults are perfectly correlated, the chance of default should be five percent, and all five will default in that case.

I found another place where I thought editing was in error, though I may just not understand things.  At about location 190, there is a paragraph stating:

"The fashionable term is now 'Big Data.' IBM estimates that we are now generating 2.5 quintillion bytes of data each day, more than 90 percent of which was created in the last two years>"

To me, read literally, that means that of the 2.5 quintillion bytes generated on a day, 90% of that day's data was created in the last two years!  I'm probably being too picky, clearly he means that 90% of all the data currently around was created in the last two years.

Silver correctly points out in the chapter how it is hard to select out the key information in the huge amount of data available, so I was a bit surprised that he seemed to be chastising economists because a few analysts had correctly predicted the problems with the housing bubble.  Sure, those reports were in there, but how many reports said that things were great, and house prices would surely double in the next twenty years, or predicted that there was indeed a real estate bubble, but gave flawed or silly reasons for it?

Being reminded about the problems with house prices gave me a bit of a smug feeling....I am a lifelong renter, and have no interest in owning and maintaining my own home.  For years, I was lectured by many peers on how foolish I was to be renting, and that owning a house was a financial necessity because it was such an amazing investment!  That advice doesn't look so good now of course.  But I'd pass on the smug feeling if it would help the economy recover!

When Silver talks about humans having an extraordinary capacity to ignore risks that threaten their livelihoods, I was astonished he didn't quote Mark Twain:

"It is difficult to get a man to understand something when his job depends on not understanding it."

(Make your own reading bar)
"Once you can accept the universe as matter expanding into nothing that is something, wearing stripes with plaid comes easy."  - Albert Einstein

Offline drenfrow

  • Status: Scheherazade
  • *****
  • Posts: 1233
  • Gender: Female
  • Terlingua, Texas
    • View Profile
Re: Week 1: Intro and Chapter 1 discussion
« Reply #3 on: November 12, 2012, 02:24:52 PM »
I also remember learning accuracy and precision from a science lab course in college.  Interestingly, you can have "consistent error" and sometimes that can be okay.  I teach middle school science and we have low-quality thermometers; 10 thermometers can show a classroom temperature ranging from 19 to 26 degrees--very inaccurate!  However, if you are using one thermometer to measure temperature change, for example, it's actually not a big deal that it's inaccurate as long as it's precise (of course I doubt these thermometers are precise either, but they're good enough for what I use them for).  But Silver's point is well made that when precision is mistaken for accuracy dire consequences can result.


  • First off, I am fairly confident that there is a typo in Figure 1-2 (~loc 473). The general idea is that if you're betting that 5 mortgages won't default, there is a greater risk if they are connected or correlated in some way - such as if all 5 houses are located in the same town (which could get hit by a natural disaster and wipe out all 5, or a factory in the town could close and many people in that town are out of work). So the risk, the probability of losing the bet, is higher with correlation than without. I believe the column headings "if defaults are perfectly correlated" and "if defaults are perfectly uncorrelated" should be switched.


I thought the same thing--they seemed backward.

I highlighted several passages but I don't have my Kindle with me at the moment.  One idea I remember being struck by was risk vs. uncertainty.  That risk is something you can actually quantify, like odds of drawing a certain card in poker, but uncertainty is when you really have no idea of what's going on.  Unfortunately we often try to quantify uncertainty.

A good practice I've read about (and tried to use myself) about checking to make sure you aren't ignoring inconvenient facts is to "look for disconfirming data."  Disconfirming sounds complicated, but isn't.  Our minds are programmed to instinctively search the sea of information for more information that confirms our biases or preferred solution.  One way to try to use this instinct in your favor is to stop and say "I am looking for stuff that will prove my choice wrong" and search the sea of data for things that will disprove, or at least argue against whatever belief or choice you are certain of.  Doing this yourself for your own choice supposedly reframes your bias so you "win" and are right if you can find info that debunks your ideas, but trying this for your own ideas is still awkward at best.  It's better if you can have an independent person with no emotional investment in the choice being investigated does that.  This is the idea behind "The Devil's Advocate" as supposedly used by Catholics to argue against making someone a saint.  I've actually tried this a couple of times at work, assigning someone to argue against a decision.  It is better than doing it yourself, but still tough for the person with the assignment, both because they probably have bought into the groupthink that lead to the decision, and in any case many people are nervous about going up against the group or especially the boss, even when they are told it is their job assignment to do so!

This is similar to the use of the null hypothesis in scientific studies.  The null hypothesis is the opposite of what you think is going to happen and you try to disprove it.

BTW, I'm enjoying the book and I'm fine with the reading schedule.  Since it's non-fiction, there's no need to rush through it.
  


 
Woke up this morning with a terrific urge to lie in bed all day and read. -Raymond Carver

Offline hsuthard

  • Status: Arthur C Clarke
  • *****
  • Posts: 2334
  • Gender: Female
  • Honolulu, HI
    • View Profile
    • Suthard, Party of Five
Re: Week 1: Intro and Chapter 1 discussion
« Reply #4 on: November 12, 2012, 10:47:38 PM »
I'm enjoying this book more than I expected, non fiction is a welcome choice.

Like Claw, I was floored by the thought of a $20k book. I was also surprised that it took 300 years for the information to start flowing after the invention of the printing press. That is a looooong time to suppress technology.

I have to say, reading his explanation of the housing bubble and ensuing economic crisis isn't helping me get over lingering bitterness about our retirement portfolio losses.
-- Holly.

The more we live, the more beautiful we are.

Offline Eltanin Publishing

  • Status: Scheherazade
  • *****
  • Posts: 1262
  • Charlotte, VT
  • Cathy Ryan - Editor-in-Chief
    • View Profile
    • Eltanin Publishing
Re: Week 1: Intro and Chapter 1 discussion
« Reply #5 on: November 13, 2012, 05:59:46 AM »
Silver correctly points out in the chapter how it is hard to select out the key information in the huge amount of data available, so I was a bit surprised that he seemed to be chastising economists because a few analysts had correctly predicted the problems with the housing bubble.  Sure, those reports were in there, but how many reports said that things were great, and house prices would surely double in the next twenty years, or predicted that there was indeed a real estate bubble, but gave flawed or silly reasons for it?

This is a good point. His whole book talks about the signal and the noise - but also HOW to pull out the relevant info. Sure, the common person might not have realized there was going to be a housing bubble (burst) but these guys promoted themselves as people who could analyze such things. When everyone is talking about how insanely high housing prices are, it's only logical to assume they might come down.

I think the important take-away from this is to emphasize what Silver said about it being very difficult to remove your own biases from your analysis. The ratings companies were tasked with analyzing how much risk was in certain investments. But they also made more money the more investments got good ratings. They may have been knowingly giving the investments overly rosy ratings, but more likely, they simply couldn't put aside their own biases, beliefs, and hopes to see what they should have been able to see (and they should have been able to see it because they were supposed experts in this area).

I, too, am grateful that although my husband and I bought our condo (our first house) on what I believe was during the rising of the bubble (Jan 2005 - when, officially was the peak, or the bubble popping?), we did fine. The prices went up after we bought, and then came down a little, but they never came down to less than what we paid. They also didn't rise or fall as much as in other parts of the country. I'd say the prices went up about 15% after we bought, then came down about 7-10% (depending on how fast you want to be able to sell).

Offline Geoffrey

  • Moderator
  • Status: Edgar Allan Poe
  • *****
  • Posts: 6864
  • Gender: Male
  • Dallas, Texas
  • Live! Live! Live! Life is a Banquet!
    • View Profile
Re: Week 1: Intro and Chapter 1 discussion
« Reply #6 on: November 13, 2012, 02:46:09 PM »
Silver correctly points out in the chapter how it is hard to select out the key information in the huge amount of data available, so I was a bit surprised that he seemed to be chastising economists because a few analysts had correctly predicted the problems with the housing bubble.  Sure, those reports were in there, but how many reports said that things were great, and house prices would surely double in the next twenty years, or predicted that there was indeed a real estate bubble, but gave flawed or silly reasons for it?



There were a lot of reports along the lines of neverending happiness in the housing market and I remember seeing the long list of articles calling it a bubble before it popped, comparing our housing market to Japan and all that ... but therein lies one of the premises of this book - that we tend to ignore data that disagrees with out worldview.  I bought my house in '02 during the last recession and bought just enough house to fit our needs and to allow us to be able to pay for it so long as one of us is gainfully employed.  At times, I was tempted to take out a loan against the equity before that bubble burst - but I kept hearing comments about housing and how it can only ever go up and how this is a new economy and all that .... same as we heard before the Dotcom bust .... and that gave me pause.

I can't say I saw the meltdown in 08 coming or anything like that - but it didn't surprise me.  As a lay person in economics or Sports or political punditry or whatever, the hardest thing of all is to find the people who have analyzed the data with the least bias.  In general, though, I tend to ignore both those who are all roses and daffodils as well as those who are only gloom and despair.



Offline Ann in Arlington

  • Global Moderator
  • Status: Shakespeare
  • *****
  • Posts: 65435
  • Gender: Female
  • Arlington, VA
  • Laugh Loud. Run Fast. Be Kind.
    • View Profile
Re: Week 1: Intro and Chapter 1 discussion
« Reply #7 on: November 13, 2012, 06:25:35 PM »
Saw this on xkcd -- seems relevant:

http://xkcd.com/1132/


My Kindles
Nautilus (Voyage)
Oasis 2016
AquaKindle
Godric's Hollow (Basic)
S7Edge (KApp for Android)
Galaxy (KApp for Android)
(Make your own reading bar) | Ann Von Hagel | Arlington, VA | kboards  MODERATOR

Offline The Hooded Claw

  • Status: Emily Dickinson
  • *******
  • Posts: 9493
  • Gender: Male
  • Oklahoma
  • Card-carrying melodrama villain
    • View Profile
    • All about me....<hee hee>
Re: Week 1: Intro and Chapter 1 discussion
« Reply #8 on: November 13, 2012, 07:07:30 PM »
Why not phone someone on the other side of the world and ask them if the Sun has exploded?   ;D

Larry Niven actually wrote a short story about this, Inconstant Moon, where the Moon suddenly became VERY bright during the night, and communications were abruptly lost with Europe and Asia.

The $50 bet reminds me of the scientists before the first atom bomb test who made bets on whether or not it would set the world's atmosphere on fire.

(Make your own reading bar)
"Once you can accept the universe as matter expanding into nothing that is something, wearing stripes with plaid comes easy."  - Albert Einstein

Offline Eltanin Publishing

  • Status: Scheherazade
  • *****
  • Posts: 1262
  • Charlotte, VT
  • Cathy Ryan - Editor-in-Chief
    • View Profile
    • Eltanin Publishing
Re: Week 1: Intro and Chapter 1 discussion
« Reply #9 on: November 14, 2012, 05:51:43 AM »
Saw this on xkcd -- seems relevant:

http://xkcd.com/1132/



This will probably make more sense to me after we get to the Bayesian chapter...