OOP length

[davewitley]

--- In HPFGU-OTChatter at y..., "cindysphynx" <cindysphynx at c...> wrote:

> So my guess is . . . 728.
> 
> Cindy (who thinks one of the math majors ought to do some cool 
> projection of the length of Book 7 based on the length of the first 
> four books correlated against the number of months JKR took to 
write 
> each book, and who is guessing the projection would put Book 7 at 
> 1,421 pages)

OK, I took Devin's figures, bunged 'em in Excel, plotted as a 
function of book number (not date) and tried various trendlines as I 
don't have time to work them out myself.

Quadratic fit looks very good, and predicts about 1100 pages, but I 
find it hard to think of a rational basis for a quadratic law.

Linear fit is poorer, and predicts 800 pages.  You could just about 
believe that there is a consistent increase in length as she gains 
confidence and gets more verbose - I felt there was a bit of that in 
GOF.

Fancier stuff (exponential, power, log) has even less common sense 
basis.

However, given that the broad outlines of the plot were worked out 
from the beginning (this is why I think dates are bad - they reflect 
the vicissitudes of her life and rewrites, not book length), I would 
expect the best predictor to be the average of all the books so far - 
about 450.  Remember, this is Devin's American editions for 
comparison.

Not bad for ten minutes work, (including writing this) though I say 
it myself.

David