Re: group-based judgement

From: the animated silicon love doll (cheshire@velvet.net)
Date: Thu May 30 2002 - 02:29:45 MDT


-----BEGIN PGP SIGNED MESSAGE-----
Hash: MD5

Thursday, May 30, 2002, 1:35:19 AM, you wrote:

> Just from a mathematical viewpoint, this system doesn't work. This is
> the statistical flaw with profiling. It sounds good to say one type is
> "twice as likely" to be bad as another type. But as shown in this
> example, using that as a basis for profiling doesn't provide reliable
> results. Any system that is only right 55% of the time is a failure.

While I do agree with you 100% about profiling not working, I believe
the point that Spike was making was more along the lines of "If you
have two options, and one is statistically twice as likely to be
harmful (for whatever reason), you'd be an idiot not to pick the one that's not as harmful."
This works fine with two mushrooms. It doesn't hold up very well with
two people, and breaks down completely with large groups of anything.

cheshire morgan. "Beer wants to be free."
                   -Bruce Sterling
PGP Key ID: 0x4193C6C7

-----BEGIN PGP SIGNATURE-----
Version: 2.6

iQCVAwUAPPXjAOfrS59H50XPAQGZpgP/QivQBBcWrLhEfjkjeRZoAhI3CYKv10/h
ekn5v/wLqnpEEQRtve0pSdvHXwBsI80XeTNrjj3iAgFA1S4USdrTji9JMsEkXpJo
OKTUHgxH67KE2SCAU0dnO+gCpPr6iez+BG8+aL8wXGq4fbyh9LDAPF+sQ9B1crzu
VlWVNegn/Iw=
=bjAi
-----END PGP SIGNATURE-----



This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 09:14:30 MST