I cringed when I read the reference to "2000 different knots". I can see Dan Lehman rolling on the carpet laughing hysterically at these figures...
Not so fast: the difference between the usual blind assertion of “over 3,000” or such
and “2000” is nearly double; it implies that some effort to count was made, as opposed
to looking at the last image-reference # “3854” only. And if DFred is involved, the count
deserves careful consideration. One will get sizeable counts of “multi-strand knots”
and “button knots”; but 2,000 is still a lot (and might include splices). To the general
reader, it comes off as “a wayyyy lot”, and that’s fairly accurate; it is perhaps most
misleading in suggesting that one will find any knot of practical use, as there are
some common knots NOT included. (Charles Warner’s smaller book has around
500 knots in it, I believe–not sure how many might depend on distinctions of use
for a similar structure.)
As some separate thread, reporting on a background task, we might endeavor
to arrive at a real count by enumerating the particular knots in [i][u]ABOK,
w/some focusing on the practical, others on the decorative (incl. sennits etc.).