Hurk4 wins the Smolin/string poll

[marcus]

Here are the forecasts we made in September about the Smolin/string salesrank ratio as of noon on 1 October (pacific time).

https://www.physicsforums.com/poll.php?do=showresults&pollid=1171

The prediction of Hurk4 (which seemed quite unreasonable to me at the time) turned out to be right after all!
On 1 October the actual ratio turned out to be 6.5, and he had predicted 6. Congratulations Hurk!

If the above link doesn't work, try
https://www.physicsforums.com/showthread.php?t=182206
which gives the poll thread and discussion.

Hurk lives in the Netherlands and says his interests are physics, music, theater,... family, friends, and nature. I understand he is retired from Philips Eindhoven (a solid-state physicist friend of mind worked there and met his wife in Eindhoven. He told me good things about the lab and the town.) Hurk has a quote from the French scientist Lavoisier that I like:

Nothing is created or destroyed---everything is process.

That must include you and me. We are each of us processes. And so is the universe. There are no fixed entities with fixed properties and qualities----this is an idea that arises from our language, I suppose, or from the way our brains have evolved.

Rien ne se cree, rien ne se perd, tout est transformation (Lavoisier)

https://www.physicsforums.com/member.php?u=30536
======================

Anybody, what do you think the salesrank ratio will be at the same time (noon pacific) on 1 November?

This is the ratio which compares the Amazon salesrank of The Trouble with Physics...and What Comes Next with the average rank of the five most popular string books at that moment.

I tracked this ratio February-May, for four months earlier this year to get a kind of baseline normal idea of it. It was fairly steady around its average value of 2.5 during those four months. That is "Trouble" was selling about 2.5 times better, judging by salesrank, than the average topfive string book, which served us as a kind of benchmark.

So in the poll, I guessed the ratio would be closest to 2, Arivero said 4, which I thought was already way too optimistic. Hurk4 said six---totally incomprehensible at the time. But that is how it turned out.

This month we have a new game. I am still betting that the ratio will settle down to 2 or 3, where it was much of the time earlier this year. That seems like a natural place to me. But you may think differently. What is your prediction?

 

[marcus]

Anyone who wants to try to guess, whether they register their guess in the poll or not, might like to have some idea of what the ratio is currently---as a point of departure.

It is currently (noon pacific, 9 October) 6617.6/1397 = 4.7

that is, the Smolin salesrank is 1397 and the average topfiver string salesrank is 6617.6, among all the books that Amazon sells.

In case anyone is curious, currently the five most popular stringy books are (with current salesranks)
Greene elegant 3608
Greene fabric 6107
Randall warped 6697
Kaku parallel 8086
Greene fabric 8590

Perhaps several of these might be considered 'string-inspired' instead of 'string theory' in a strict sense, but that's how the index has been computed all along and it helps to have a consistent benchmark. I treat a book as stringy if it seems more obviously string-inspired than otherwise, and average up those which are selling the best to provide something with which to compare Smolin TTWP sales.

To me it's quite remarkable, even bizarre, that this ratio is so high---4.7 is high relative to expectations, and what I saw earlier this year. Either this is just a fluke and will go away, or something is happening to the state of mind of the sector of the reading public that buys science books.

If anyone would like to check the standings, here is the link
https://www.amazon.com/gp/bestsellers/books/14545&tag=pfamazon01-20

 

[ccdantas]

Marcus,

I really do not understand what you are doing here, perhaps I'm misunderstanding it. But those comparisons do not seem to make much sense to me. You should not compare how Smolin's book is selling right now (after its one year-old release, iirc) with some other book that has already been released many years from now. You should at least compare how many books by Greene's The Elegant Universe (for instance) were sold in, say, 3 years after its release, with that of Smolin (after the same amount of time, that is, 3 years after its release). I assume here that books have a peak in their sales curve in their first years of release, and then decrease. The best thing would be to compare their sales curves in a fixed time span (after release), but I do not think such information is available.

Christine

 

[marcus]

Marcus,

I really do not understand what you are doing here, perhaps I'm misunderstanding it. But those comparisons do not seem to make much sense to me...

Hi Christine, thanks so much for asking! As I may have mentioned earlier, this ratio is one that I tracked on a regular basis for the four months February-May 2007 to establish a baseline.
And indeed the top five stringies have included many NEW books! Susskind Cosmic Landscape (2006), Randall Warped (2005, 2006), Steinhardt Endless (2007), Beck (2006), and doubtless others which I am forgetting.

At one point I counted the number of stringy books on Amazon and there were well over a dozen popularizations, and something like 20 if you include textbooks. But most of the popularizations never got much notice.

During that period, the average rank of the top five averaged over the month, was fairly steady around 5000.
Ranks of individual books tend to fluctuate. But for some reason the average of the top N books is fairly stable.

I don't care whether at any given time the members of the top five are old (like Greene's) or new (like Randall's book, Susskind's book, and Steinhardt's). When you sample the top five, whether they are new or old, you are in some sense sampling the whole population of stringy books. If there is a new book, then if it is popular enough, this will get it.

I would say that in a "live" subject one EXPECTS that books follow an arc or trajectory and that the old constantly fade away and are replaced by the new. I would expect sales of Smolin book to also arc like this.

But as long as string is a "live" subject, I would expect the top five average to remain steady, as new books come up to replace the old ones dying down. And I expect this string benchmark to continue to serve as a good point of comparison. So we will someday see the Smolin book decline relative to this standard.

Perhaps you don't realize the necessity to have some benchmark or baseline for comparison. The whole physics book salesrank picture shifts up and down in response to events outside the physics book market. Salesranks are based on other books that Amazon sells and other genres like religion, politics, self-help, Potter, etc can raise or lower physics salesranks. Also the university and school semester temporarily changes the ranks of popularizations versus textbooks. So the numbers bounce around in a crazy meaningless way.

What I have found is that the most steady way of gauging the impact of the Smolin book is to do so relative to this kind of standardized string impact indicator.

This does not prove anything about the relative TRUTH or VALUE of the books. Smolin book selling more or less does not prove that it is better or worse than the leading string book. What sales help to gauge is IMPACT on how the reading public is thinking and exposure to the book's message. If the book does not sell then its message does not reach those whose behavior and judgment might be changed by it.

So you see I am not trying to estimate the virtue of the book but rather its impact and usefulness relative to the problem of research monopoly and diversification which we face in the United States. If you can suggest any better ratio to use, please tell me! In any case thanks again for your reaction.

 

[ccdantas]

Hi Marcus,

Thank you for your clarification. I see better now what you are doing. Although I still do not completely agree, I do not have any other suggestion for the moment. I'm sure you put some thought on this more than I did, so I'll keep coming back to this thread and see where it leads... Thanks.


Christine

 

[arivero]

what about google trends? http://www.google.es/trends

http://www.google.es/trends?q=+stri...+Higgs+boson,+&ctab=0&geo=all&date=all&sort=0

 

[marcus]

what about google trends? http://www.google.es/trends

http://www.google.es/trends?q=+stri...+Higgs+boson,+&ctab=0&geo=all&date=all&sort=0

that's interesting. I hadn't used that before.
http://www.google.es/trends?q=string+theory&date=all&geo=usa&ctab=0&ctab=0&sa=N
In the US, it looks like a gradual decline in the rate of people searching "string theory" in google, since the beginning of 2005.
At present, the state where people search for it most often is Utah.

===================================

But what I'm interested in is the impact of Smolin's book----which is really too tiny to show up in Google Trends.
I curious what other people expect to be the trajectory of this book and what impact it could have on the state of mind in the US about science, if any.

where do you think this index is going to go? If you want to look for trends, here are noon readings for three recent days:

1 October 6.5
9 October 4.7
10 October 4.2
...
...
1 November ?

It looks like it is declining, but how far will it go? Will it get down to the kind of "plateau" of 2.5 it was on during February-May? Or will it find a new level?

 

[arivero]

that's interesting. I hadn't used that before.
http://www.google.es/trends?q=string+theory&date=all&geo=usa&ctab=0&ctab=0&sa=N
In the US, it looks like a gradual decline in the rate of people searching "string theory" in google, since the beginning of 2005.

It seems it is actually more useful to detect bursts that to detect trends. So probably the peak mid 2006 is associated to Smolin (September 19) & Woit (April 25) books. The other burst, in Xmas 2004, is less straightforward.

 

[marcus]

I've been expecting this Smolin impact rating to decline back down to the level where it plateaued during February-May, when I was watching it. Then it was fairly stead and averaged 2.5. So I would predict something like 2 or 3 for the first of November.

But at present it isn't acting like that. Where do you think this index is going? Here are noon readings for some recent days:


1 October 6.5
9 October 4.7
10 October 4.2
11 October 7.4
...
1 November ?

I was surprised yesterday (11 Oct) to see it at 7.4. Common sense says the book has been out for over a year---since beginning September 2006---and you'd think it would be tailing off already, unless it's destined to become a perennial favorite or classic of some kind.

Arivero, I'm glad you suggested the google search index. I wish the vertical scale was adjustable on those charts, or you could get actual numbers of searches. Their graphs are sketchy and hard to read. It's good to have objective ways to gauge what is going on: I hope we find some more.

 

[marcus]

In case anyone wants to try guessing where this index will be, a couple of weeks from now, here are noon readings for some recent days:


1 October 6.5
9 October 4.7
10 October 4.2
11 October 7.4
13 October 6.4
...
1 November ?

Common sense says the book has been out for over a year---since beginning September 2006---and you'd think it would be tailing off already, unless it's destined to become some kind of physics classic. That's why I predicted 3.

At noon (pacific) today "Trouble with Physics" salesrank was 1083 and those of the five currently most popular stringy books were:
Greene elegant 3427
Steinhardt endless 6993
Kaku hyperspace 7913
Randall warped 8126
Kaku parallel 8225
So the stringy topfive average was 6936.8 and the ratio 6936.8/1083 = 6.4

In other words, the Smolin book is currently selling some six times better, judging by salesranks, than the stringy average I'm using as a standard of comparison. Earlier this year when I was watching the ratio tended to stay around 2.5. So from the book's standpoint there has been, at least temporarily, a considerable improvement.

 

[MathematicalPhysicist]

so we got to the point that it doens't matter if the theory is valid, but that it's poular amongst laymen, correct?

 

[marcus]

so we got to the point that it doens't matter if the theory is valid, but that it's poular amongst laymen, correct?

I believe the contrary from what you say---namely I believe that what matters is that a theory is TESTABLE by making new predictions not already predicted by prior theory, so that it can be falsified. And then what matters is that it PASSES EMPIRICAL TEST.

What laymen find popular or appealing does determine the validity of the theory, in my view.

Why do you think otherwise?

 

[MathematicalPhysicist]

this thread of yours and others which all the time compare sales of pop sci books of theoriticians, why not compare ideas of theories discuss technical issues in the theories.
instead I am reading here some sort of gossip, i don't think it's a one case iv'e seen other posts of yours here of the same kind.

p.s
im as of yet not into the technical details of every theory but i guess there's more than comparing sales of books.

just my opinion.

 

[marcus]

Smolin's booksales are a way of predicting the impact of Smolin's book. Their being high gives reason to hope that his arguments for supporting more than one line of fundamental theory research in the US will influence opinion-leaders and decision makers in the US scientific establishment including funding agencies.

At the present time you and I cannot judge the validity of string thinking (there is so far no one testable string theory as such, rather more a philosophical framework or bunch of approaches) or the validity of nonstring quantum gravity approaches. That is not our job. Indeed I do not believe anyone can reliably judge the liklihood that one or the other of the approaches will be right or even contribute part of the solution.

That is WHY supporting a variety of research lines is important. We don't know which will pay off and to what extent. So the intelligent way is to spread our bets.

At present the situation in the US is extremely abnormal, by comparison with Europe, Canada and the UK. There is only one physics department in the US with a non-string QG group (two or more faculty). Postdoc support ordinarily goes where there is a group rather than solo faculty. Recent PhDs in nonstring QG HAVE TO LEAVE THE COUNTRY to continue doing nonstring QG. In the past couple of years they have gone to Europe, Canada, and the UK. None have found postdoc contracts in US physics departments (although one Baez PhD did get postdoc in a math department.)

String has a stranglehold monopoly on fundamental physics research in the US. If you are a graduate student who wants to study nonstring and write a thesis---there are scores who do want, judging from the turnout at Zakopane---you have almost no chance to do this in the US. Except at Penn State you will probably not get an advisor, or funding. After PhD at Penn State, you must look for a position at the dozen or so places outside US where they have nonstring groups.

(Perimeter-Waterloo, Western Ontario, Portsmouth, London, Nottingham, Utrecht, Mainz, Berlin-Potsdam, Tours, Lyon, Marseille, Montpellier, Chennai, Morelia,...)

The string monopoly in the US is clearly restricting the freedom of advanced graduate students to study what they want to study and pursue the line of research they want to pursue. I believe this monopoly or monoculture is harmful to US science and should be broken. So far the only book which has made the case for diversifying US theory research in this way is the Smolin book. So I am especially interested to gauge the impact of this book in whatever way I can see to do this.

If you can think of any other ways to judge its impact, please suggest them.

thanks

noon readings for some recent days:


1 October 6.5
9 October 4.7
10 October 4.2
11 October 7.4
13 October 6.4
14 October 4.7
...
1 November ?

at noon on 14th, Smolin rank was #1131, average topfive string rank was #5065, so ratio was 4.7
In case anyone is curious, today the five most popular stringy books and their ranks were
elegant 2189
parallel 2889
warped 5065
hyperspace 7261
fabric 9282

 

[MathematicalPhysicist]

but isn't smolin's book a popular book anyway, i.e it's not addressed spcifically to those who fund the research?
i don't see the correlation between popular books and the way they affect the funding of research in the us, care to elaborate on that?
i mean if he has some remarks he could give it in a conference or publish an article directed to the funding groups, which will affect them more seriously than writing it in a pop book, just my thoughts.
p.s
im not sure how is it even a pop sci book cause it seems he covers only the politics around the string and non string funding.
anyway, i don't think that selling more books than string theory books will change this matter of funding.

 

[marcus]

that is just what we are going to see, isn't it!
It is a remarkably intelligent book that has won the attention of an impressive lineup of reviewers and science journalists. I think it contains some of the best popular explanations I have seen so far, of the strengths and weaknesses of various approaches to unification and QG. Combines personal experience of the history of these fields with clear verbal imagery to deliver the mathematical sense without relying on equations.

The book is not only written for general audience but is also obviously aimed at policy makers, academic/government decision makers. It succeeds in being both about SCIENCE itself and about SCIENCE POLICY.
I think the twofold effectiveness is what makes this a remarkable and unique book. I can't think of any other book that excels it in those different dimensions. Can you?

I am beginning to suspect that my guess of 3, for the ratio as of the first of next month, was a mite conservative---here are noon readings for some recent days:


1 October 6.5
9 October 4.7
10 October 4.2
11 October 7.4
13 October 6.4
14 October 4.7
15 October 7.1
...
1 November ?

at noon on 15th, Smolin rank was #900, average topfive string rank was #6368, so ratio was 7.1
In case anyone is curious, today the five most popular stringy books and their ranks were
elegant 2188
fabric 5562
parallel 5647
warped 7699
hardbound elegant 10,744

============
One poster taking part in the thread may not realize that the merit of the book as arguing for a more diverse research strategy in the US is inseparable from its actual impact on public attitudes and institutional decision-making. Both must be considered and discussed in order to adequately handle either.

 

[MathematicalPhysicist]

so discuss the book, not its rankings in amazon.
just my humble thought about this post.

if you already did so, then i think you can keep track in the same thread instead of opening another thread.

 

[arivero]

On the contrary, the role of the threads is to keep people able to avoid what they are not interested on :-D

 

[josh1]

so we got to the point that it doens't matter if the theory is valid, but that it's poular amongst laymen, correct?

but isn't smolin's book a popular book anyway, i.e it's not addressed spcifically to those who fund the research?
i don't see the correlation between popular books and the way they affect the funding of research in the us, care to elaborate on that?
i mean if he has some remarks he could give it in a conference or publish an article directed to the funding groups, which will affect them more seriously than writing it in a pop book, just my thoughts.
p.s
im not sure how is it even a pop sci book cause it seems he covers only the politics around the string and non string funding.
anyway, i don't think that selling more books than string theory books will change this matter of funding.

Of course you are right loop quantum gravity! It's ridiculous that this same kind of completely unconstructive exchange has to take place over and over again!

 

[marcus]

Here are noon readings for some recent days:


1 October 6.5
9 October 4.7
10 October 4.2
11 October 7.4
13 October 6.4
14 October 4.7
15 October 7.1
16 October 4.4
...
1 November ?

at noon on 16th, Smolin rank was #1538, average topfive string rank was #6792, so ratio was 4.4
In case anyone is curious, today the five most popular stringy books and their ranks were
elegant 1831
fabric 2145
parallel 4092
warped 11,741
hyperspace 14,152

It's still a challenge, i think, to guess what the index will be at noon on 1 November. One would expect a downwards trend as the market for the book saturates and things return to normality. Anyone who still wants to hazard a prediction is welcome. Whatever you guess there's still a good probability, I'd say, that you'll be proven wrong.

 

[marcus]

1 October 6.5
9 October 4.7
10 October 4.2
11 October 7.4
13 October 6.4
14 October 4.7
15 October 7.1
16 October 4.4
18 October 6.5
19 October 6.3
...
1 November ?

at noon on 18th, Smolin rank was #1331, average topfive string rank was #8646.0, making the ratio 6.5
EDIT: noon numbers are now in for the 19th
and Smolin rank was #1085, while topfive stringie average rank was #6862.0, making the ratio 6.3
========================

From my perspective these numbers are remarkably high. Here are my February-May results from making the same measurement:
For February, the string topfive salesrank average was 5101.0 and the TWP figure was 1802.6, ratio 2.8
For March, the string average salesrank was 5785.5 and the TWP figure was 1967.7, ratio 2.9.
For April, the string average was 6741.3 and the TWP figure was 2443.8, ratio 2.8.
For May, string average 5107.4 and TWP was 2902.2, ratio 1.8
For the whole four month period the averages and their ratio were 5683.8/2279.1 = 2.5

So I would normally expect to see ratios like around TWO, like Smolin book should be selling about TWICE as well as the most popular five stringies average. In fact that is what i guessed in September for the 1 October figure---but it came out 6, as Hurk predicted. So this month I am guessing 3, for the 1 November.

But instead it keeps being, like six. It is really weird to be seeing numbers this large and we are already half way thru the month. It makes me suspect some change. The book has been out well over a year now and it is still having some impact. I heard Santa Barbara string-thinkers last year (November 2006) saying that the book would probably just go away and the whole thing (whatever they meant by that) quiet down. It is what a person might have reasonably expected to happen---and to have happened by now. But somehow it didn't.

 

[marcus]

1 October 6.5
9 October 4.7
10 October 4.2
11 October 7.4
13 October 6.4
14 October 4.7
15 October 7.1
16 October 4.4
18 October 6.5
19 October 6.3
20 October 5.2
21 October 5.0
...
1 November ?

at noon on 20th, Smolin rank was #1292, average topfive string rank was #6678.4, making the ratio 5.2. In case anyone is curious what today's top five string books were:
elegant 4155
fabric 4324
warped 6492
parallel 9002
becker text 9419

at noon on 21st Smolin rank #1306, topfive stringy avg. #6526.8, ratio 5.0
warped 2394
elegant 4657
fabric 5896
parallel 9351
endless 10,336

 

[marcus]

this ratio is behaving strangely. at 10 AM pacific today it was at 10.1
Smolin salesrank #738 and topfive stringy #7425.2.
Reminds me of a spike in the first half of September where it went up around 11.

One symptom of what's happening is that NEW stringy books aren't catching on as well as they used to in the past. New popularizations could be expected to come in and keep the average top five sales strong, unless the public's interest is drying up. The most recent new ones (Susskind "Cosmic Landscape: String Theory and the Illusion of Intelligent Design" and Steinhardt "Endless Universe") have actually not done so well. Both of them, judging by titles, were well calculated to appeal to the popular thirst for speculation about Origins----big bang creation drama and so on. And yet they didn't really go all that much.
Lenny Susskind is going to try the market again with somethng called "The Black Hole Wars: my battle with Stephen Hawking to make the world safe for quantum mechanics."
Hard to guess how that will go over. It has namedropping and "battle" dramatization going for it, and that box office Black Hole in the title. Who knows?

at 11 AM the ratio was 9.3 namely 8189.6/876. We'll see how it is at noon---that is the one I've been recording.
==================

I should make it clear (as I have several times in the past) that the Smolin book having a substantial impact would not prove it right---the book argues for spreading our research bets in a way that would make the situation in the US more like what we see in Europe and Canada---getting away from the string "monoculture" in US departments.
Whether that is right or wrong---a good or a bad policy suggestion---a valid argument to make or not----is not decided by sales index relative to some benchmark.

What it means if the book has a major impact is that more people (including opinion makers, university administration, government funding advisory panel personel) will read the book and there will be more liklihood that the monoculture pattern will be broken and that we will see a situation develop in the US that is more like it is abroad.

Sales do not prove correctness, obviously. They are a gauge of potential impact. E.g. whether we will get to where there is more than just a single university in the US (namely Penn State) with a nonstring QG group---whether grad students at several major US universities will have a choice---that's the kind of impact I'm trying to predict if it will happen or not.

1 October 6.5
9 October 4.7
10 October 4.2
11 October 7.4
13 October 6.4
14 October 4.7
15 October 7.1
16 October 4.4
18 October 6.5
19 October 6.3
20 October 5.2
21 October 5.0
22 October 7.3
...
1 November ?

Smolin book #1019, five most popular stringy average #7443
this time the five most popular were: warped, elegant, fabric, parallel, and the Susskind book about string landscape and the look of intelligent design

 

[marcus]

I was out yesterday and could not take the usual noon reading. May have to be on the road again today. However, just to keep us more or less up to date, at 11AM this morning the ratio was 6.1.
Smolin rank #998 and average of the five most popular stringies #6062.8.
Those five being
elegant 2825
warped 3779
fabric 7084
parallel 7298
hyperspace 9328

As it turned out, we didn't go out. The noon reading was nearly the same: 5.6


1 October 6.5
9 October 4.7
10 October 4.2
11 October 7.4
13 October 6.4
14 October 4.7
15 October 7.1
16 October 4.4
18 October 6.5
19 October 6.3
20 October 5.2
21 October 5.0
22 October 7.3
24 October 5.6
...
1 November ?

 

[marcus]

1 October 6.5
9 October 4.7
10 October 4.2
11 October 7.4
13 October 6.4
14 October 4.7
15 October 7.1
16 October 4.4
18 October 6.5
19 October 6.3
20 October 5.2
21 October 5.0
22 October 7.3
24 October 5.6
25 October 6.2
26 October 5.6
27 October 4.2
...
1 November ?

The ratios at noon on the 25th and 26th were 7496.8/1217 = 6.2 and
7033.4/1251 = 5.6 respectively. At noon on 27th, 6284.4/1514 = 4.2