Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 5/26/83; site ihuxr.UUCP
Path: utzoo!linus!decvax!harpo!eagle!mhuxt!mhuxi!mhuxa!houxm!ihnp4!ihuxr!lew
From: lew@ihuxr.UUCP
Newsgroups: net.math
Subject: correction to posted puzzle answer
Message-ID: <588@ihuxr.UUCP>
Date: Sun, 28-Aug-83 22:44:07 EDT
Article-I.D.: ihuxr.588
Posted: Sun Aug 28 22:44:07 1983
Date-Received: Tue, 30-Aug-83 01:11:04 EDT
Organization: BTL Naperville, Il.
Lines: 15

Argh! I misstated the formula for P(M,N) in my posting. To repeat:

The probability of correctly "sticking" on the highest of N numbers
by picking M and then stopping on the next number which is the highest
so far is:

	P(M,N) = M/N * sum i = M, N-1 of 1/i

In my posting of this answer I had 1/N instead of M/N. My description of
the derivation was correct.

As long as I'm posting, note that P(N/2,N) -> log 2 / 2 (.346574) as
N -> infinity. P(5,10) = .3728, not much worse than P(3,10) = .3987

	Lew Mammel, Jr. ihuxr!lew