"Wisdom of Crowds" -- book

[andicap]

Has anyone heard of the book, "The Wisdom of Crowds," by James Surowiecki -- if so, do they think it has any application for handicapping or betting in general? Or trading exchanges/prediction markets?


http://www.randomhouse.com/features/wisdomofcrowds/

[twindouble]

Has anyone heard of the book, "The Wisdom of Crowds," by James Surowiecki -- if so, do they think it has any application for handicapping or betting in general? Or trading exchanges/prediction markets?


http://www.randomhouse.com/features/wisdomofcrowds/

Fear, Panic, Greed, uncertainty, irrational exuberance and the most important thievery. That's all you need to know. :D

[Cesario!]

I'll rarely put down a book after only two pages, but I did with this one. The author uses the show "Who Wants to Be a Millionaire" as an example of "crowd wisdom." In his example, he asserts that because the audience as a whole does better as a lifelife ("ask the audience") than the individual ("phone-a-friend") as introductory proof of his theory.

In making this claim, he explicity discounts any possibility that the questions asked of the audience are significantly easier than the questions asked of the friend as the reason for this discrepency. But, anyone watching the show even casually would know that this is indeed the case -- the audience is typically asked an early, and thus, question and the friend is asked one later in the game. At the very least, any argument asserting the "wisdom of crowds" would need to consider the difficulty of questions in making this conclusion.

For me, the failure to investigate this phenomonon a little bit deeper made me skeptical of the author's rigor and his ability to draw useful conclusions -- hence, the closing of the book so quickly. Proceed with caution -- it takes more than a catchy title to make something science...

Seth

[twindouble]

These intellegent answers tend to bug me for some reason. Good post Cesario!

[DJofSD]

andicap,

Ask him yourself on 9/7 -- he'll be in your city. See the first item. (http://www.randomhouse.com/features/wisdomofcrowds/meetauthor.html)

[kenwoodallpromos]

I thought that meant the crowd had a better collective memory when it came to lower value information, but when the big bucks is on the line it is all about who you know!LOL!!

[robert99]

Over 95% of the betting public regularly lose.
The media panders to the mob and puts them away.
Once the "mob has spoken", do the opposite - be a contrarian, as in the Sinatra song.

[Sly7449]

Greetings,

Is there a Updated Edition of this book and if so, are there any additions to the past reviews?

The initial thread is 4 years old but I see new interest generated by a well respected programmer.

Did the Crowd change during the initial writing/publication as opposed to the Now Crowd?

Sly

[Dave Schwartz]

No, but you can get some great information if you do a Google for "decision markets" and "prediction markets."

Check out Wiki as well.


Dave

[CBedo]

If you liked this book, you'd probably also like Crowdsourcing: Why the Power of the Crowd is Driving the Future of Business by Jeff Howe.

Lots of interesting topics that could potentially be applicable here.

(In fact, I think I brought up an example in Dave's inital thread that I thought was from Wisdom of Crowds, that I'm pretty sure is in Crowdsourcing instead.)

[bobphilo]

I'll rarely put down a book after only two pages, but I did with this one. The author uses the show "Who Wants to Be a Millionaire" as an example of "crowd wisdom." In his example, he asserts that because the audience as a whole does better as a lifelife ("ask the audience") than the individual ("phone-a-friend") as introductory proof of his theory.

In making this claim, he explicity discounts any possibility that the questions asked of the audience are significantly easier than the questions asked of the friend as the reason for this discrepency. But, anyone watching the show even casually would know that this is indeed the case -- the audience is typically asked an early, and thus, question and the friend is asked one later in the game. At the very least, any argument asserting the "wisdom of crowds" would need to consider the difficulty of questions in making this conclusion.

For me, the failure to investigate this phenomonon a little bit deeper made me skeptical of the author's rigor and his ability to draw useful conclusions -- hence, the closing of the book so quickly. Proceed with caution -- it takes more than a catchy title to make something science...

Seth

The author admits to this precise problem when he states:

“Now, the results of "Who Wants to Be a Millionaire?" would never stand up to scientific scrutiny. We don't know how smart the experts were, so we don't know how impressive outperforming them was. And since the experts and the audiences didn't always answer the same questions, it's possible, though not likely, that the audiences were asked easier questions. Even so, it's hard to resist the thought that the success of the Millionaire audience was a modern example of the same phenomenon that Francis Galton caught a glimpse of a century ago.
As it happens, the possibilities of group intelligence, at least when it came to judging questions of fact, were demonstrated by a host of experiments conducted by American sociologists and psychologists between 1920 and the mid-1950s, the heyday of research into group dynamics.”

http://tinyurl.com/2aqbq (http://tinyurl.com/2aqbq) Click Excerpts

In other words, while admitting the possibility of a confounding variable (different degree of difficulty of questions asked each group), the large effect size could very well be due to the “wisdom of the crowd” as observed earlier by Galton. Obviously, the Millionaire example is not offered as scientific proof but is merely an example using a TV program that the majority of his readers would be familiar with. He then goes on to cite academic studies performed with statistical rigor by sociologists and psychologists.

Even if he had only used an argument that failed to prove his hypothesis and left it at that in his introduction, it does not mean that it is disproved. One must never confuse a failure to prove a hypothesis with having disproved that hypothesis. I would at least give him a chance to do so in the text that follows. In reality he never claims the Millionaire example is scientific proof but does go on to cite studies that were performed with proper statistical rigor. Definetly worth giving him a chance to make his own argument in the rest of the book.

Bob

[Jackal]

Groups of people make good decisions when they have all the information at hand. The DRF or track program doesn't give you all the information - the crowd is "wrong" more often than usual.

[LutherCalvin]

The mind of all of us is better than the mind of any of us - that's common sense. The public choice (the favorite) wins consistently at a 33% clip, a hit rate that towers over most handicappers. But the key to the mint long term is evaluating the odds and hitting the overlays.

[Robert Goren]

This thread has got me wondering. Does favorite do better in the race with highest win pool that day compared to the other races of the day?

[TrifectaMike]

The mind of all of us is better than the mind of any of us - that's common sense. The public choice (the favorite) wins consistently at a 33% clip, a hit rate that towers over most handicappers. But the key to the mint long term is evaluating the odds and hitting the overlays.

How do you know when you have an overlay situation with any certainty?

Do you assign a probability to a perceived overlay?

Just asking.

Mike

[speculus]

If you liked this book, you'd probably also like Crowdsourcing: Why the Power of the Crowd is Driving the Future of Business by Jeff Howe.

Lots of interesting topics that could potentially be applicable here.

(In fact, I think I brought up an example in Dave's inital thread that I thought was from Wisdom of Crowds, that I'm pretty sure is in Crowdsourcing instead.)

And here (http://www.amazon.com/Why-Flip-Coin-Science-Decisions/dp/0471296457/ref=sr_1_1?ie=UTF8&s=books&qid=1263433855&sr=8-1) is another that shows how people (in groups, committees, govt, NOT exactly the crowd) "make a decision" as contrasted with an individual.

[Dave Schwartz]

Good to hear from you, Spec.

I have that book. Very basic stuff.

[bobphilo]

Groups of people make good decisions when they have all the information at hand. The DRF or track program doesn't give you all the information - the crowd is "wrong" more often than usual.

The public favorite wins more often than any other horse in the field. The 2nd choice wins the 2nd most races, the 3rd choice is the 3rd most common winner, and so on down the line. Furthermore, horses tend to win very closely in accordance with their post-time odds. Every study shows that the public is right more often than not. Recognizing the not too common exceptions are where the overlays come from.

Bob

[appistappis]

I wonder if these books are ripoffs of the original best book on this subject.

Extraordinary popular delusions and the madness of crowds.....by mackay 1841.

I was told to read this back in the early 70's before embarking on a career in the stock market and racetrack.

[bobphilo]

From my previous posts on this thread some may have drawn the conclusion that I agree with the "public wisdom effect". Not so. I'm merely pointing out the errors in some criticisms of the theory. Some of the arguments for the public widom effect are just as flawed. Personally, I feel that more research is needed to settle this important issue.

The"publics ability" to pick the most likely winner as well as the correct odds for most horses and outperform top newspaper handicapping "experts" has often been cited as proof for this position of "crowd wisdom". I've even heard these findings presented as proof that democracy is the best political system (it may well be but not for the reasons stated in the flawed argument) The flaw in this argument is that the pari-mutual system is not an example of the one man/woman = one vote condition required for democracy. The "vote" of the small $2.00 bettor only has a fraction of the clout of the thousands of dollars bet by the whales, who have a more disproportionate effect on the odds than that wagered by the smaller better. So we are mainly talkling about the opinions of bigger bettors when we talk about "public choices". Furthermore, when comparing the public choices with the media 'cappers, the media guys are under the disadvantage of having to make their picks the night before before scratches and track conditions at post-time are known.
Having said this, there seem to be other studies that point to the power of this "crowd effect" without these flaws. I'm really curious to see if Surowiecki's work is one of them. I can make no fair judgements without having read ithe book, but I'm curious to hear honest evaluations from my fellows here until I do.

Bob

[misscashalot]

I wonder if these books are ripoffs of the original best book on this subject.

Extraordinary popular delusions and the madness of crowds.....by mackay 1841.

I was told to read this back in the early 70's before embarking on a career in the stock market and racetrack.
http://www.econlib.org/library/Mackay/macEx.html

[macdiarmida]

Reading the book summary, it sounds similar to the Delphi Method (or Technique), so you might look up stuff online about that. Lots of stuff there. Delphi is supposed to be used for very complicated and therefore very subjective questions. Surowiecki talks about finding THE answer. Dunno about that; there's always the best estimate of a winner vs. the dumb-luck 150-1 stab in racing (and the market). "I'd rather be lucky than smart", as they say . . .

In both methodologies, each member of the group is supposed to work independently. No influencing others or allowing influence by others. No phoning a friend. Delphi is supposedly used by gummint and corporate folks to determine policy and help make Big Decisions, but if EVERYONE has an axe to grind or has their own agenda, guess what?? But you've probably put together P6 tickets with a group (usually with only one round of study), so you know what typically happens there. And if you get an answer that just about NOBODY likes, then the answer will just be ignored or buried. Reminiscent of racetracks and the answer they always get to their paid studies about what happens when you reduce takeout.

Tests with Delphi found that non-experts did only slightly worse than experts BTW. So you teachers out there, get your kids to work on the next P6 carryover!

The situations that are described by Mackay are more like reality IMO.

[Jackal]

The public favorite wins more often than any other horse in the field. The 2nd choice wins the 2nd most races, the 3rd choice is the 3rd most common winner, and so on down the line. Furthermore, horses tend to win very closely in accordance with their post-time odds. Every study shows that the public is right more often than not. Recognizing the not too common exceptions are where the overlays come from.

Bob

I agree with your concept but they don't win often enough to turn a profit. If you lose 2 of 3 at even money....

[Dave Schwartz]

I am familiar with the Delphi Methodology, which is essentially, an interactive collaboration model.

The wisdom of the crowds approach is similar except that there is no interactivity. It is similar to "round one" of the Delphi Method.

Some of the best answers come from non-experts. One of the issues at PA is that we are all so learned. I believe that getting the less-sophisticated of our members to vote would help this quite a bit. (That is one of the principles from the book.)


Dave

[Overlay]

I agree with your concept but they don't win often enough to turn a profit. If you lose 2 of 3 at even money....

I understand what you're saying with regard to favorites, but the public (despite its remarkable overall/aggregate accuracy in ranking horses and setting odds) regularly makes similar errors in judgment from race to race on individual longer-priced horses, where the odds are more forgiving as far as the winning percentage required to show a profit.

Looking at the winning chances of every horse in a race field (rather than just trying to narrow the field down to the one most likely winner) makes it possible to capitalize on such opportunities. The same logic also applies to exotic-wager combinations where probable payoff information is available.

[Dave Schwartz]

One thing to consider is that when the public wagers, it is not picking the best horse. It is trying to consider value as well as hit rate.

We need "voters" who just concentrate on hit rate.

[LutherCalvin]

How do you know when you have an overlay situation with any certainty?

Do you assign a probability to a perceived overlay?

Just asking.

Mike

Races with the largest fields and those with furious pace scenarios provide the greatest overlay opportunities for me.