Patrick writes:

> It appears that Monty Hall is a bit more generous with the odds, but no
>  less of a con man. It's still the same game.

As I understand the Monty Hall game, you have the opportunity to change
sides, thus increasing your odds to 2/3 after Monty has shown you one empty
window, with him having full god-like pre-knowledge. If that's so, he is
actually being quite generous.

A great many "games of chance" actually aren't games of chance at all, but
require a careful distortion of the facts (much like the current Republican
primaries :-).

Let me show you a game of chance that isn't -- and one that I've used over
the last 30 years to win dozens of steak dinners, beers, pizzas, etc. It is
100% guaranteed that you will win every time, but to find players, you have
to carefully and somberly lay out the conditions of the game.

Best of all, this game is actually based on a standard computer technology,
b-trees, so this thread may have some relevance after all.

The game is this: pick up a standard collegiate dictionary and mention that
there are 250,000 entries in the dictionary (go ahead, count 'em). What
you'll do is bet a steak dinner that you can find any word in the dictionary
in 20 guesses or less by asking only yes or no questions. Heck, just
calculate those odds: 20 chances out of 250,000. It doesn't matter if the
word they pick is "catatonic", "catastrophe", or just "cat." You wouldn't
haven't the slightest idea what their word would be.

After they've agreed and written their word down on a piece of paper, you
pick up the dictionary and open it up right at the middle -- and ask: "Does
the word you've chosen occur before (or is it) 'matrix'?", presuming that
"matrix" is the word at the top of the page that you opened. If the word does
occur before "matrix", you drop the last half of the dictionary (and in the
process throw away 125,000 words you now know don't qualify). At step 2, you
again split the remaining dictionary again right down the middle and ask the
same question regarding the new word at the top of the page. This process
repeats another 19 times, each time throwing away half of the remaining
dicitionary.

Because you get 20 tries at this -- and 2 to the 20th will separate out a
million entries down to 1 remaining word -- if you haven't found the word by
the 18th or 19th try, somebody merely screwed up. Otherwise, you are
absolutely 100% certain to win.

Who says computer theory can't be useful?

Wirt Atmar