View Single Post
Old 2005-09-08, 19:07   #1
ewmayer
2ω=0
 
ewmayer's Avatar
 
Sep 2002
Rep├║blica de California

2×13×443 Posts
Exclamation Proposal for new subforum

Inspired by the latest crank-maths thread, I would like to suggest a new subforum specifically for these kinds of threads. I propose to title it "Cranks, Kooks and Miscellaneous Malfeasants". This would serve as a home for threads which in the opinion of the moderators are either certain or highly likely to be wastes of time if taken seriously (or even if not, if they lead to flame wars, as crank and troll threads so often do.)

This is not meant to stifle discussion, but rather to encourage *useful* discussion, by explicitly contrasting it with useless (or extremely likely to be so) discussion, and thus providing a mechanism for forewarning readers and newbies as to the distinction.

Properties which make a thread a prime candidate for the CKMM subforum could include (but need not be limited to) the following:

1) Obvious trolls, e.g. "I hav ritten a pogrom for my cellfone that is 10x faster than Prime95 and used it to find the first 10-million-digit prime, but I ain't tellin what it is until i collect the prise $$$. so keep searching, u losers."

2) Threads in which the author makes wild unsupported claims which are either provably false or extremely likely to be so. The aforementioned thread is a classic example of this, as is e.g. this one:

http://www.mersenneforum.org/showthread.php?t=3965

Another example of this would be someone who admits they have little number-theoretic experience but claims to have discovered an elementary proof of, say, the Riemann hypothesis or Fermat's last theorem.

3) Threads in which the author makes wild claims which are not provably false but still extremely likely to be so (e.g. "I have proved there are infinitely many Fermat primes"), which the author supports via obscure (but possibly quite-mathematical-sounding) argumentation. In my opinion this is the most insidious form of crankery, because while it is generally just as useless as its more-blatant brethren, because it is couched in specious (and usually byzantine) mathematical "reasoning' it generally require time-consuming analysis by knowledgeable people to find the worm in the core of the apple, as it were. An example of this:

http://www.mersenneforum.org/showthread.php?t=676


Your thoughts?
ewmayer is offline