Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

History of the Crawford rule?

7 views
Skip to first unread message

Jason A. Dahlin

unread,
Jan 18, 2001, 9:35:45 PM1/18/01
to
I've been able to find it was named after John Crawford, but don't the story
behind who he was...

I understand that it stops somone from saying "I have nothing to lose, so I
may as well double." If I lose the game, I still lose the match. If I win,
I'm that much closer to ctahcing up (or win, depending on the race
situation).

Again... just looking for some back-story.

Jason
jda...@carolina.rr.spam.com


Andrew Grant

unread,
Jan 19, 2001, 6:45:22 AM1/19/01
to

> I understand that it stops somone from saying "I have nothing to lose, so
I
> may as well double." If I lose the game, I still lose the match. If I
win,
> I'm that much closer to ctahcing up (or win, depending on the race
> situation).

Only for one game. After that, if you win, you're free to double straight
away in all subsequent games.

Nor does it apply in situations where, for instance, your opponent needs two
points and turns the cube to 2. If you accept this double it then becomes
correct to redouble to 4 immediately, for the same sort of reason.

Andrew Grant.


Julian Hayward

unread,
Jan 19, 2001, 3:09:19 PM1/19/01
to
In article <5KN96.1865$Kl5.3...@typhoon.southeast.rr.com>, Jason A.
Dahlin <bl...@carolina.rr.com> writes

>I've been able to find it was named after John Crawford, but don't the story
>behind who he was...
>
>I understand that it stops somone from saying "I have nothing to lose, so I
>may as well double." If I lose the game, I still lose the match. If I win,
>I'm that much closer to ctahcing up (or win, depending on the race
>situation).

The most significant consideration is that it stops a game played with
both players 2 away from victory becoming meaningless - you could
otherwise let the opponent take a point then double the next game
immediately, nullifying his advantage. It has a similar effect at other
scores, making the game that takes the leader to 1 from winning closer
in value to the rest.

--
Julian Hayward 'Booles' on FIBS jul...@ratbag.demon.co.uk
+44-1480-210097 http://www.ratbag.demon.co.uk/
---------------------------------------------------------------------------
"A witty saying proves nothing" - Voltaire
---------------------------------------------------------------------------

Jason A. Dahlin

unread,
Jan 19, 2001, 8:22:38 PM1/19/01
to
I understand how it applies, when it doesn't apply and the general purpose
behind it...

My question was "what is the history behind it" "why is it named after some
guy named Crawford"?

Jason


"Julian Hayward" <Jul...@ratbag.demon.co.uk> wrote in message
news:Vb1MVCAv...@ratbag.demon.co.uk...

Michael Strato

unread,
Jan 20, 2001, 3:10:21 PM1/20/01
to
Hi Jason,

The Crawford rule was invented by John Crawford.

Here I quote a paragraph from "The Backgammon Book" by Oswald Jacoby and
John R. Crawford. Copyright 1970 by Oswald Jacoby and John R. Crawford.
Viking Press ISBN 670144096:

"The Crawford Rule

At the Bahamas and Las Vegas tournaments it was felt that this doubling
privilege was unfair. One proposed solution was to forbid doubles once a
player got within one point of victory, but it was agreed that this might
lead to an interminable series of dull single games in the event that one
man had something like a twenty-to-ten lead in a twenty-one-point match.

John Crawford, who is always chairman of the tournament committee at these
events, devised a rule providing that once a player gets within one point of
victory, there can be no double in the first game after that; however,
doubling would be allowed in the match from then on. This gives the player
who has come so close to victory a one-game safety in which his opponent
cannot double immediately and then happen to win a lucky gammon or
backgammon to snatch away the victory."

<end quote>

There's your answer Jason. Kind of interesting to note that we often hear
people suggesting rules or variants that would reduce the luck factor in
Backgammon. It appears that this is one that does just that. I say if the
leading player was lucky or skillful enough to get so close to victory then
he doesn't need this advantage. I think it is fair that the trailing player
should maintain the right to double for he is the one that could do with
some leverage at this point in the match.

Michael


"Jason A. Dahlin" <bl...@carolina.rr.com> wrote in message
news:yL5a6.9788$Kl5.1...@typhoon.southeast.rr.com...

Art Grater

unread,
Jan 20, 2001, 4:59:06 PM1/20/01
to
Besides giving the leader an advantage, it introduces additional complexity
into match-score strategy.

Fairness is not really an issue, since the Crawford advantage is available
to both players when the match begins. In any event, excitement trumps
fairness!

Art

Michael Strato

unread,
Jan 20, 2001, 5:50:06 PM1/20/01
to
Hi Art,

You have a very good point there.

I forgot that I do like the "additional complexity into match-score
strategy."

Thanks for the reminder,

Michael


"Art Grater" <a...@NNOSPAMdoggy.com> wrote in message
news:94d158$3au$1...@slb6.atl.mindspring.net...

Stein Kulseth

unread,
Jan 26, 2001, 6:36:53 AM1/26/01
to

Art Grater wrote in message <94d158$3au$1...@slb6.atl.mindspring.net>...

>Besides giving the leader an advantage, it introduces additional complexity
>into match-score strategy.

Does it? Or does it reduce the complexity of match-score strategy?

The match equity table with the crawford rule is quite "well-ordered" in
the sense that each point you win from the current score until you win
is more or less of equal value. Without the crawford rule the match equity
table would seem to become less "well-ordered".

Say you're trailing -3, -2, When should you double?
Well if there are no gammons you shouldn't double at all - the extra point
gained
is worthless to you, but valuable to your opponent. And how does this affect
correct cubehandling at previous scores.

It is in any case not obvious to me that overall cube strategy would be very
much easier
without the Crawford rule.

Stein


0 new messages