I didn’t declare that it is overall growth for the market (as a whole) — just within that 750 items I’ve had access to.

Assuming that I continue to get the full list next year, THEN I’ll be able to determine something about the BookScan reporting venues overall growth.

“your conclusions about art-comics sales — a conclusion so thoroughly contradicted by the facts on the ground that one has to wonder about the factors that your analysis misses”

One of the reasons I write this thing each year is to try and get a better idea of what those factors are or might be.

I’m still less than assured that the “facts on the ground” say what you think they say (since they currently sit at “anecdote” stage, rather than facts, in my opinion), and I put an email into Eric to try and sort out some of the answers; once I’ve gotten his response, I’m going to pitch the same to both the Chris’, and we’ll see what we see.

-B

Actually, a statistician would call it “sampling” and a sample of size 750 is quite large (even assuming an infinite population) for estimating something like a growth rate. Of greater concern would be the possibility that the population of retailers that report to Bookscan is different than the overall population of retailers.

Again: You’ve first taken the top 750 out of some 13,000 units sold in bookstores last year, compared it to the top 750 out of an unknown number of books sold the year before — and declared a 2.3% overall growth for the market. That’s indefensible math solution #1.

You then compare this to Diamond’s yearly top 100 list, which — again — doesn’t give one anything remotely resembling a complete picture of the Direct Market. Brian, this isn’t “LITERALLY the ONLY comparison that CAN be made,” it’s a comparison that shouldn’t ever have been made to begin with.

Look, aside from the bit discussed above and your conclusions about art-comics sales — a conclusion so thoroughly contradicted by the facts on the ground that one has to wonder about the factors that your analysis misses — I’ve largely restrained myself to demonstrating that different conclusions can be drawn from the same circumstantial evidence, in as entertaining a fashion as I can. This isn’t a fight to the death, Brian. It’s a debate over what an analysis concerning information of limited reliability can actually tell one about various markets. That other people are arguing with your conclusions is just par for the course, not an attempt to castrate you in public. Calm down.

On the “75%” thing, outside of the introduction where I AM EXPLAINING BOOKSCAN’S FLAWS, it only ever gets mentioned ONCE, in like two sentences talking about the possible size of the Long Tail that BookScan isn’t tracking. Two sentences out of more than 11,000 words. Jinkies.

Further, RE “other numbers suggest 65% is closer to the fact”, only suggests that in terms of Dr. Phil diet books and Ben Franklin biographies — we have no ON THE RECORD source that has ever said any such thing in relationship to COMICS. In point of fact, I have more than one OFF the record source that tells me that BookScan is equivalent to NINETY percent of their bookstore sales. Given the lack of ON THE RECORD sources, I’ve chosen to go with Neilsen’s Official Party Line, because there really ISN’T any credible evidence otherwise…

On the “Top 100″ thing, there may be a miscommunication here because there are two different Top 100s — the monthly ones, and the end of the year one.

The end of the year one is ABSOLUTELY USELESS as a metric of ANYTHING. Why? Because it isn’t indexed to anything — there’s literally no way of telling anything about most of the books that appear on the year end.

The MONTHLY Top 100s *are* indexed, and can give specific numbers (within their remit)

The comparison to Diamond’s figures, and Diamond’s yearly growth in the book category (which, I should underline, is not *MY* calculation, but one performed by BOTH Comics Chronicles and ICv2… and they match!), is THE SUM of those twelve monthly Top 100s. In other words, there’s something on the order of 1100 data points. It is still WOEFULLY incomplete, yes, of course, but it is the ONLY set of data points that one can POSSIBLY reference.

I’ve never disagreed that it is NOT a “apples to apples” comparison, but it is LITERALLY the ONLY comparison that CAN be made, because of the way Diamond reports.

Either way, four-to-six sentences out of 11,000+ words.

Hope that makes some of this clearer?

-B