[mikeh]

Anyone who knows me knows that I do a lot of reading on a wide range of subjects, and one of my favourite topics is trying to understand, as best as a layperson can, how our minds operate. I have enjoyed several of Pinker's books, for example, and I know some others on this forum share this interest.

So I thought I'd share word about a fascinating book I am now reading, on the strong recommendation of a psychiatrist I had a business lunch with recently. I suspect some here are already familiar with it.

It is called Thinking Fast and Slow, and I got it on my kindle...it was published, I think, a couple of years ago. It will astound you and make you rethink any preconceptions you may have about how rational you think you may be.

We have 9 litigators in my firm and I am urging all of them to read this book. It will give you insight into how others think and, most importantly, how you think and will alert you to problems to which you would surely otherwise be completely blind. In short, I think it to be brilliant.

It is by a psychologist who is, I think, the only non-economist to win the Nobel Prize in economics, which will tell you something about his credentials.

[Winstonm]

Thanks, Mike. I know this stuff fascinates me.

[kenberg]

I may well give this a try. I think the interplay between logic and intuition is very challenging and I would be interested in seeing what he says about this and other issues. In social situations I think that I am often the first, or one of the first, to recognize when something is fishy, something isn't right. But I am not at all good at figuring just what it is. The mistakes that I most regret are those where I did not act on my intuition because I could not work it through in detail. I have learned that when my skin tells me that something is wrong, usually something is wrong.

It should make for interesting reading.

[mike777]

Daniel Kahneman and Tversky are very well known in the field of behavioral finance. I have read their stuff for decades.

[barmar]

I've mentioned this book on several occasions in other threads.

Another author I recommend is Dan Ariely and his book "Predictably Irrational".

[Fluffy]

I've always believed that we made 2 kinds of decisions, "digital" and "analogic", where analogic are made from mixing various weighted factors into a formula that we don't really comprehend (might have some voltaic issues)

Does this have anything to do with fast/slow?

[kenberg]

Fluffy, on 2014-July-29, 00:43, said:

I've always believed that we made 2 kinds of decisions, "digital" and "analogic", where analogic are made from mixing various weighted factors into a formula that we don't really comprehend (might have some voltaic issues)

Does this have anything to do with fast/slow?

I like this way of thinking about thinking. Both approaches are valuable, one can be a check on the other, and for significant issues it is a very good idea to bring both into play.

[barmar]

Fluffy, on 2014-July-29, 00:43, said:

I've always believed that we made 2 kinds of decisions, "digital" and "analogic", where analogic are made from mixing various weighted factors into a formula that we don't really comprehend (might have some voltaic issues)

Does this have anything to do with fast/slow?

It's similar, I think. The analogic decisions are the fast ones, made using the subconscious pattern-matching methods of the primitive brain. Digital decisions are the slow ones, made by the modern, contemplative brain. The way these two processes coexist is that the slow method usually acts as a check on the fast one.

[mikeh]

barmar, on 2014-July-29, 08:02, said:

It's similar, I think. The analogic decisions are the fast ones, made using the subconscious pattern-matching methods of the primitive brain. Digital decisions are the slow ones, made by the modern, contemplative brain. The way these two processes coexist is that the slow method usually acts as a check on the fast one.

I agree with this, in very broad terms. What I find most fascinating is just how bad the slow method is!

As the author says, it is lazy. When we are contemplating a question, even the slow method will usually readily answer a difficult question but will, instead, substitute an easier question, without our conscious awareness.

So 'system 2', the slow method or fluff's digital method, is not necessarily a conscious and definitely not a fully conscious way of thinking, even tho we may be under the impression that it is.

There is so much weirdness in the book, but all verified by studies, and often times I find myself recognizing the way I think or act...which is of course why this sort of book is so valuable.

[mike777]

mikeh, on 2014-July-29, 09:55, said:

I agree with this, in very broad terms. What I find most fascinating is just how bad the slow method is!

As the author says, it is lazy. When we are contemplating a question, even the slow method will usually readily answer a difficult question but will, instead, substitute an easier question, without our conscious awareness.

So 'system 2', the slow method or fluff's digital method, is not necessarily a conscious and definitely not a fully conscious way of thinking, even tho we may be under the impression that it is.

There is so much weirdness in the book, but all verified by studies, and often times I find myself recognizing the way I think or act...which is of course why this sort of book is so valuable.

valuable?

why?

How do you measure?

In 2014 I reread the book but only ask how do you measure and compare?
--------


let me put up something...something called heuristics

I suggest we call it trial and error, I suggest method call it trial d error or call it tinkering

[gwnn]

Fluffy, on 2014-July-29, 00:43, said:

I've always believed that we made 2 kinds of decisions, "digital" and "analogic", where analogic are made from mixing various weighted factors into a formula that we don't really comprehend (might have some voltaic issues)

Does this have anything to do with fast/slow?

I don't get it, what is the digital then? Don't we mix factors in all of our decisions? I'm not saying your distinction is wrong, just that I don't understand it.

[Fluffy]

digital, or logical, or heuristic if you want, just the contrary of analogic. In that thinking I normally use logic rules as X ->Y, the important thing is that the result is not a "feeling" of what its right, it is a yes or no.

A good example would be a bridge endind, or for a chess player a chess ending might be eeven better. On a 5 card endind you might picture your cards just the same as deep finesse (winners/losers) in the end. Assuming all the logic rules you deducted are right.

[gwnn]

Fluffy, on 2014-July-30, 03:12, said:

digital, or logical, or heuristic if you want, just the contrary of analogic. In that thinking I normally use logic rules as X ->Y, the important thing is that the result is not a "feeling" of what its right, it is a yes or no.

A good example would be a bridge endind, or for a chess player a chess ending might be eeven better. On a 5 card endind you might picture your cards just the same as deep finesse (winners/losers) in the end. Assuming all the logic rules you deducted are right.

I think most decisions will be somewhere on a spectrum between purely logical (I don't like the word digital since a lot of the time the range of possible outcomes is continuous and not discrete) and purely intuitive, that's why I find your distinction a bit strange. In a very limited range of situations such as solving a sudoku puzzle you can get by just with purely logical deductions but in real life you need to use less-than-exact data, working assumptions, and just common sense. I don't think there are a lot of purely logical decisions that we make in our day-to-day lives. Even in chess or in bridge you typically have a bunch of uncertainty and even if you had hours to compute various probabilities exactly, the underlying assumptions will be inexact/unknowable ("how often does he lead from Kxx?" even if you played against him countless times you can't know exactly as it depends on so many things).

[barmar]

mikeh, on 2014-July-29, 09:55, said:

I agree with this, in very broad terms. What I find most fascinating is just how bad the slow method is!

Or turn it around, what many find surprising is how good the fast method is. But if you think about it (using the slow method), it shouldn't be so surprising. The fast method is essentially the same method used by less "intelligent" animals, even many of the most primitive. It was honed millions of years ago by natural selection, and it's what has kept all animal species alive. The fast method is most successful at detecting and reacting to danger -- if our ancestors hadn't, they would have been eaten and not passed on their genes.

System 2 is much more recent, and we only have the luxury of using it in less critical situations. So it has not had as much selection pressure to be reliable, and hasn't had time to be improved as much as System 1. On the other hand, because System 1 is most useful for protecting us from death, it tends to be biased in the reactionary and risk-averse direction. This explains things like phobias -- if apes and primitive humans didn't avoid snakes and spiders they got bitten and died from the venom, because there was no such thing as medicine. As much as we know logically that we're not in as much danger now, the animal part of our brain is still hard-wired to consider them serious threats, and it takes work by System 2 to counteract it. It also explains why people react differently to equivalent situations depending on whether they're expressed in a negative or positive way: if you tell someone that there's a 30% chance of rain, they're probably more likely to take an umbrella than if you tell them that there's a 70% chance of good weather. This can have serious consequences: if a patient is contemplating an operation, they're probably more likely to go with it if the doctor says it has a 90% success rate than a 10% failure rate -- just mentioning failure triggers the primitive brain into protecting you at all costs.

[Fluffy]

gwnn, on 2014-July-30, 09:18, said:

I think most decisions will be somewhere on a spectrum between purely logical (I don't like the word digital since a lot of the time the range of possible outcomes is continuous and not discrete) and purely intuitive, that's why I find your distinction a bit strange. In a very limited range of situations such as solving a sudoku puzzle you can get by just with purely logical deductions but in real life you need to use less-than-exact data, working assumptions, and just common sense. I don't think there are a lot of purely logical decisions that we make in our day-to-day lives. Even in chess or in bridge you typically have a bunch of uncertainty and even if you had hours to compute various probabilities exactly, the underlying assumptions will be inexact/unknowable ("how often does he lead from Kxx?" even if you played against him countless times you can't know exactly as it depends on so many things).

mmm I think you are right, in university (informatics) I leart the so called Fuzzy Logic, from what I recall it consisted on applying logical rules to uncertain data. So if you had a variable such as West holds ♠K at 80% certainity, west has ♥Q at 50%, And then you can apply the rule If west has ♠K and ♥Q he is squeezed in the run of diamonds, and in the end you get a certainity of making the contract (40% I think)

I think the human brain is very good at calculating this things in an analogical way, however after various steps of logic, at least for me, there is a high tendency of ending around 50%. I tend to "lose" info by applying it at multiple levels.

But still applying pure logic with only certain/uncertain is possible, but it is tough!, For example years ago my declaring skills were lower than today, and I remember that I often started a difficuly han like this:

-This hand is too tough if trumps are 4-1, so I will just assume they are 3-2 and go from there.

After going down a couple of times in important hands when game was makeable despite the 4-1 break I changed the methodology


Another example, I was defending in biarrtiz against a 4♠ game, the key suits were as follows:

---

Declarer won the lead and run out of red cards before playing ♣A and another (blocking himself with ♣Q in dummy)

I went into the tank and realiced he must have this shape and deduced that tapping declarer would succed because he would next play ♠Q covered, a spade back, and when he finds trumps 4-1 he must play a club tu dummy but I pitch and win control.

All logical, except that I applied a total certainity to declarer playing a spade to hand after ♠Q-♠K-♠A
Instead declarer stopped to think, and reviewing the bidding and lead got to the conclusion that spades were 4-1, so he advanced ♣Q without a second round of trumps and made it.
I don't remember if I had a better defence, but you see the point.


On the other hand switching to an unbid suit through declarer were dummy has 3 small is pattern recognition which is very different from the former.

[mikeh]

Fluffy, on 2014-July-31, 03:03, said:

mmm I think you are right, in university (informatics) I leart the so called Fuzzy Logic, from what I recall it consisted on applying logical rules to uncertain data. So if you had a variable such as West holds ♠K at 80% certainity, west has ♥Q at 50%, And then you can apply the rule If west has ♠K and ♥Q he is squeezed in the run of diamonds, and in the end you get a certainity of making the contract (40% I think)

I think the human brain is very good at calculating this things in an analogical way, however after various steps of logic, at least for me, there is a high tendency of ending around 50%. I tend to "lose" info by applying it at multiple levels.

In fact, the evidence is exactly the opposite. Humans tend to be very bad at estimating probabilities, even when they are supposedly trained to do so. In the book, the author gives the example of study design....in which experimentalists have to determine the appropriate size of the study group needed to ensure that the resulting data is statistically reliable.

Almost all of those professional scientists tested, in terms of being asked, in a hypothetical situation, about the appropriate sample size grossly underestimated the number required.

For me, the book contains a lot of very useful ideas, some of which I intend to use when forming questions for expert witnesses (because I now have more insight into possible blind spots), and for making arguments to Judges and Juries (because I am now aware of such notions as anchoring, and of the 'what you see is all there is' issues with our mental processing, and for my own analysis of the evidence and the issues (because, I hope, I am now more alive to the misleading shortcuts that my own mind will be taking).

Now, in bridge, I expect the expert will get the odds roughly right most of the time, but I think that is more likely a residual effect of having really studied/thought about/practiced/ and seen situations many times and not because, as a human, the expert innately possessed any real talent. We can overcome the unconscious limits of our modes of thinking by practice...in essence, in some cases our perception of the odds becomes a 'system 1' process, just as riding a bicycle becomes (it started as system 2, which is why we usually need training wheels to start with )

In real life, where I notice the issues most right now is that I am taking a series of lessons designed to greatly change my golf swing. The result is that most times I am now over-thinking...using system 2 methods...and as a result am struggling.

[Fluffy]

I meant good in the sense of fast as opposed to the other way which for me its pretty slow

[Al_U_Card]

mikeh, on 2014-July-31, 09:45, said:

-(because I now have more insight into possible blind spots),

-I am now more alive to the misleading shortcuts that my own mind will be taking).

-just as riding a bicycle becomes (it started as system 2, which is why we usually need training wheels to start with )

-a series of lessons designed to greatly change my golf swing. The result is that most times I am now over-thinking...using system 2 methods...and as a result am struggling.

-Due to the integration of the viewpoints expressed by those other individuals.

-Shortcuts? Interference patterns rather.

-Not if you have the seat low enough to touch your toes to the ground. Balance and certainty (of not failing/falling) allows "feel" to guide the process.

-I also changed my swing radically over the winter (to a more Strickeresque square-to-square, upright swing plane) and limiting mental interference to rudimentary guides (such as a Mike Weir-like pre-swing waggle) and the admonition to start as "one-piece takeaway" removes the thought process and allows the body to function as it can on its own. The only problem is sore hands from the thousands of range balls so far... (I broke 80 recently and almost shot par on the back nine yesterday.)

Our internal processes are simple yet quite intriguing, to say the least. Why they are so is as interesting a subject as how to exploit them to our advantage.

[kenberg]

I can relate (a little) to that bit about golf. I mentioned elsewhere that when I was 12 or 13 I bought some clubs at the Salvation Army and I would get up early to bike out to the course before the groundskeepers would kick me of. I took a practical approach. The golf ball is a ball, I have a club, I whack the ball with the club. Not great but adequate. In my early 20s I took from golf lessons. The guy was highly regarded, I was told that Patty Berg (golf pro,no relation) would fly in for his guidance. The lessons totally ruined my game.I pretty much gave up the game, it wasn't even fun anymore.

[barmar]

kenberg, on 2014-August-02, 09:33, said:

The lessons totally ruined my game.I pretty much gave up the game, it wasn't even fun anymore.

In general, I think activities that require excessive use of System 2 are not "fun". It's a slow, exhausting method to do anything. But if you keep at it, eventually what you learn shifts over to System 1. Then the basic activity becomes automatic, and you can concentrate on the more interesting aspects that lead toward great expertise.

Watch a child first learning to read, sounding out every letter. There's little brain power left over to understand what they're reading, which is why reading primers use such simple sentences (e.g. "See Dick run."). It's a total chore at that stage, but eventually we internalize it, and when we can read easily we can then concentrate on what the words say, not how they say it.

Or watch novice bridge players. They don't have basic bidding sequences in System 1, they're constantly trying to remember what everything means and figure out what to do with common hands. This makes it difficult for them to plan ahead, or picture everyone's hands, etc. These are abilities that experienced bridge players take for granted, because all the mundane stuff has been shifted into System 1.

Every difficult activity, including sports, playing musical instruments, and learning a craft, involves an early, difficult learning stage like this. It's a wonder that we usually do manage to get past it. I guess it's a testament to the part of our intelligence that can see past the current difficulty to the rewards that will come down the line.