Red State, Blue State sales are a factor of 2^100 lower than they should’ve been

In his forthcoming book, Albert-László Barabási writes, “There is a theorem in publishing that each graph halves a book’s audience.” If only someone had told me this two years ago!

More seriously, this tongue-in-cheek theorem, if true, defines an upsetting paradox. As we discussed at the beginning of the Notes section of Red State, Blue State, we structured the book around graphs because that seemed to be the best way to communicate our findings. Tables are not a serious way of conveying numerical information on the scale that we’re interested in, and, sure, we could’ve done it all in words (even saying things like “We ran a regression and it was statistically significant”), but we felt that this would not fully involve readers in our reasoning. The paradox–or maybe it’s not such a paradox at all–is that graphs are grabby, they engage the reader, but this makes reading the book a slower, more intense, and more difficult endeavor.

P.S. Barabási apparently believes the theorem himself. His research publications are full of graphs, but his book has none at all (and no tables either). Well, it has one diagram, I guess. He may very well be making the right call on this one. People who want to see the graphs can follow the references and look up the scientific research articles that underlie the work described in the book.

8 thoughts on “Red State, Blue State sales are a factor of 2^100 lower than they should’ve been

  1. (1) If it's indeed a theorem in publishing, one would think you'd find it on Google. But A search for "each graph halves a book's audience" pulls up only 7 references (including this one!) each of which mentions Barabasi's book.

    <a href="; rel="nofollow"&gt <a href="http://;” target=”_blank”>;

    (2) Like Cook and Fellows, I've heard this about equations.

    (3) What does "theorem in publishing" even mean?

  2. Zbicyclist:

    1. Actually, it's only one reference on Google. The others are just blogs that have an autolink to this one.

    2. The bit about equations seems more plausible, at first. But maybe Barabási has a point. Equations can be skipped, but graphs invite scrutiny and thus effort on the part of the reader.

    3. A theorem in publishing is a statement that can be derived from axioms in publishing, of course.

  3. Interesting that he calls this a theorem, not a law. My publisher may not have heard this since they encouraged me to include diagrams so I made a few charts and a few tables. But I made the diagrams independent of the text so nothing is lost by skipping them.

    I do subscribe to the hypothesis about equations and therefore I included none in my book. There are authors who try to sneak equations past readers by writing sentences like: "the standard error is the standard deviation divided by the square root of the sample size." The reason for this I've never understood.

    The SuperFreaks apparently believe that equations sell books. In Chapter 1, they manufactured this: RIMPACT > PIMPACT. (the Real estate agent's effect is larger than the Pimp's effect). And since they sold a lot of books, maybe we should take note!

  4. Doesn't it depend on what the alternative is? Replace graphs with equations and you may well cut the number of readers. But its not clear to me that replacing graphs with prose is an improvement. I have been recently reading one of those pop mathematics on Poincare's conjecture. The author decided to forego any equations or diagrams. But trying to describe topology just in words is a waste of time & much of the book is incomprehensible, except perhaps to the people who would understand the equations anyway.

Comments are closed.