Benjamin Goertzel ben at goertzel.org
Mon Apr 30 17:24:59 UTC 2007

```Hi,

ESP could provide a blackjack advantage quite differently from card
counting, yeah.

The essential approach to card counting consists of
-- playing the "basic strategy" for blackjack, which leave the house with a
roughly 0.5% edge over you
-- then, keeping track of how many cards have been dealt so far in the "2-6"
range, versus how many have been dealt so far in the "10's, face cards and
Aces" range.
-- Basically, you then want to bet more when a lot of "10's etc." have been
dealt, and bet less when a lot of "2-6"'s have been dealt.  Because you are
more likely to get screwed when hit with a "10 etc." than with a "2-6"
-- Also, of course, you can bias your play a bit (choices of when to take a
hit) based on whether you know there are very few "10s etc." left in the
deck or not

More sophisticated versions of card counting involve team play, where
different members of the team count the number of cards in different ranges,
and send each other messages involving the count via signals like the angle
of rotation of a chip sitting on the table.

If someone had ESP they wouldn't necessarily be seeing future patterns
involving:

Probability(next card is a 10, face card or ace) - Probability(next card is
a 2-6)

Rather, they might see, e.g., what the exact next card was going to be, but
with a certain (presumably high) degree of error.  Getting a noisy,
error-prone picture of the **exact next card** would give a quite different
pattern of behavior than card counting.

As a simple example, card counting gets more and more accurate as the deck
is dealt down so that fewer and fewer cards are left in the shoe.  I don't
see why ESP would be likely to obey this same pattern.

Anyway, blackjack is difficult to study as the records of games are stored
only on video.  If one wanted to look for evidence of psi in gambling, I
would focus on looking for evidence of precognition in slot machine or video
poker play.  Regular gamblers log onto the slot or poker machine by sliding
their "slot club card" into the machine, so there is may well be a recorded
log of the exact sequence of choices made by each player.  (For a slot
machine the choice is just how much to bet each time; poker machines are
more interesting because the player can make more choices.)  If one could
get ahold of these logs, one could test whether there were any particular
players who tended to be way luckier than would statistically make sense...

This would be a pretty straightforward data analysis job, but it would
require a casino owner being willing to share their data for this purpose.

-- Ben G

On 4/30/07, Damien Broderick <thespike at satx.rr.com> wrote:
>
> At 11:28 AM 4/30/2007 -0400, Ben Goertzel wrote:
>
> >if you're the dealer and YOU can count cards even primitively, you
> >can tell if a player is systematically tending to increase their
> >bets when the count is favorable and decrease their bets when the
> >count is unfavorable.  (But of course, a good counter will randomize
> >their bets to some extent, to make the recognition of their betting
> >pattern harder.)
>
> Suppose someone *did* have an ESP capacity applicable to such card
> games (leaving aside such miraculous powers as being able to change
> one card into another or derandomize the shuffle etc). Would it
> operate differently from card-counting? Would she make advantageous
> bets at times a card-counter would not be able to, or something? That
> is, is the only difference between using ESP and skill that the
> former might be done with the eyes closed?
>
> Damien Broderick
>
>
> _______________________________________________
> extropy-chat mailing list
> extropy-chat at lists.extropy.org
> http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20070430/b8cbcd48/attachment.html>
```