|
Hello Folks,
I saved all of the replies to the original post back in December by Alan
Webb. I have read and reread it and have felt that all together these
responses constitute nothing short of another chapter in Magriel.
I have taken the liberty to do some minor editing (deleted headers and
corrected some minor punctuation) and compiled all of it together in a one
text document which is attached.
If anyone is interested in this document and has difficulty downloading
it, please contact me at:
divdesman@home.com
and I will be happy to e-mail it to you. It is about 10 pages of text so
I felt it inappropriate to post the whole thing. If there is sufficient
interest I will be happy to post it here in its entirety.
ENJOY!
Have a Happy Day,
Richard
--------------
Subject: Holding Games?
Date: Mon, 21 Dec 1998 12:00:20 +0100
From: vsg-isg@t-online.de (VSG)
Newsgroups: rec.games.backgammon
I recently started a thread which referred to Back games. I have to say
the response was more than I hoped for so I thought Id push my luck and
ask another question :)
What is a holding game?
Ive read Bill Roberties "Backgammon for Winners" and "Backgammon for
Serious Players" both of which do not seem to address the concept and
strategy of Holding games.
My impression is that it is the making of an advanced anchor, which
should not be broken, so that if you find yourself behind in the game,
it acts to slow down your opponents bearing in and gives you the
opportunity to build your board and hope for a hit.
Surely that cant be it?
I mean it seems to me that it shouldn't be too difficult to bear in as
you can easily move behind? I think there must be something I'm missing
here :) and I have the feeling its to do with WHEN to break up the
advanced anchor. Problem with that is if you move one out your likely
to get blitzed (which was covered in the books) as your opponent will
have a whole bunch of builders ready to zap you.
I guess its a question of timing, which is likely only to come with
experience, or is there a pip count rule.. i.e. under so many percent
RUN RABBIT RUN!! :)
I hope this question provides food for thought and look forward to any
Replies here or per email on vsg-isg@t-online.de
Thanks in advance
Alan Webb
PS. thx once again for the backgame replies. BTW "2-3, 1-3, 2-4" seemed
to be the most popular, and another thing i learned was.. "Backgames as
a last resort.." pleased to see this newsgroup caters for keen learners
:) and isn't full of "I know that..huh!" attitudes
**********************************************************************
Brian Sheppard
Date: Tue, 22 Dec 1998 13:06:57 GMT
You nailed it: a holding game involves a high defensive anchor that
provides some chance of getting a shot.
The distinction between "high" and "low" defensive anchors is important
because of cube-handling. If you hold the 18 or 20 point you can almost
always take a double, but the 21 point does not always provide enough
equity to justify taking and it gets still worse as you go still lower.
I would put the dividing line for holding games at the 22 point, calling
Games with lower anchors "ace-point" or "deuce-point" games.
Brian
***************************************************************************
Donald Kahn
Date: Tue, 22 Dec 1998 13:44:21 GMT
Pretty close. These checker play in these games is not among the most
intricate. I don't pose as an expert on the matter, but here are a
few random observations.
If opponent has three points inside or better, breaking the point can
be dangerous. You tend to wait for a "convoy" (set of doubles).
Otherwise, expect to be "worked over", as my old teacher Art Dickman
put it. Remember if he hits you loose, you are not a favorite to
survive.
Certainly never run off voluntarily when you are behind in the race.
It can't be right to race, when you are losing the race.
If he still has the midpoint to clear, you can use some sets of
doubles to move into the outfield to "oppose" it.
The real art, from both sides, is the cube element. This separates
the men from the boys. If you have a 20-pt anchor, what racing lead
and how many outfield points does the racing leader need to double
you? When is it a take?
Used to be you had to have played a **lot** of backgammon to have your
feet on the ground on this kind of stuff. But with such as JellyFish,
you set up variation after variation, ask the evaluation, roll it out
if it's close, and deduce your own set of rules for active play.
Deekay
*******************************************************************
David Montgomery
Date: 22 Dec 1998 11:12:28 -0500
The term isn't really well-defined. There are positions that everyone
would call a holding game, but there are also a lot of positions that
some would call holding games, while others would call them something
else. I call a lot of positions anchor games that others call holding
games.
The fundamental idea is maintaining contact in order to be able to
hit the opponent while they are coming home (bearing in or clearing
their last few outside points).
Usually you maintain this contact with a point, although often a blot
a ways behind the point provides additional contact, making it harder
for the opponent to play safely behind you.
The point(s) held usually include the 22, 21, 20, or 18 points.
However, you could also have a 15 point vs 15 point holding game,
and a midpoint holding the 18 point isn't that uncommon.
I tend to think of a position as a holding game only when there is
a credible chance of getting a shot while the leader is bearing home.
So I wouldn't call holding the 22 (or 21 or 20) point against a six prime
a holding game -- I would call it an anchor game. However, if I held
the 22 point against the 4, 5, 6, and 9 points, it looks more like a
holding game.
If your main holding point is an anchor, you may need a second point
in the outfield to have a good chance of a shot. Let's say your
opponent is bearing in and needs to clear the mid and 8 points.
Holding the 22 point alone, I would tend to think of this as an
anchor game. With the 22 point and the midpoint (or some other
outside point bearing on the opponent's mid) I would call it a
holding game.
Holding games are distinguished from backgames primarily by when you
intend to hit your opponent. In a backgame, often the plan is not
to hit until your opponent is bearing off. In a holding game, usually
you will hit as soon as possible, as long as your offense isn't a
shambles. (However, in advanced backgames, like a 54 or 53 backgame,
you usually do intend to hit while the opponent is bearing in.)
David Montgomery
monty@cs.umd.edu
monty on FIBS
********************************************************************
Julian Hayward
Date: Tue, 22 Dec 1998 17:04:14 +0000
Not quite. The point is not to slow down your opponent's bearing in but
to hinder him from clearing points to bear in safely. A typical position
along these lines is:
24-23-22-21-20-19----18-17-16-15-14-13
o o o x o o o x
o o x o o o x
o o
x
x x x x x o
x x x x x o
1--2--3--4--5--6-----7--8--9-10-11-12
The aim is to make life as difficult for O to clear the three remaining
outfield points. In this position X is not badly placed - there are
plenty of rolls for O which force him to leave 17 rather than 12 in
order to avoid leaving shots, and 6-3 forces him off 18 (even better for
you).
The critical aims for you are:
(i) You hold points that cover more or less all the outfield space that
O has to play through, so if O leaves shots they will be direct ones,
(ii) You have a relatively good home board *now*, so that O is under
pressure to play safely,
(iii) O's position is relatively stripped, so he is likely to be forced
off his outfield points very soon.
(iii) is very important - if you move O's two spares from 21/22 to 19/18
he has three, maybe four rolls to get a 65 or any double. That
flexibility means the game will turn out to be little more than a race
in which you are far behind.
Another common position is for you to hold 13 and 18, in which case O
reduces to two men on his midpoint, then often gets forced to play one
of them off with 6-x.
If O sees this sort of situation developing, he can try to defeat it in
several ways - use good numbers to clear points at the back rather than
build the home board - accept an early shot to clear an awkward point
*before* X has built up his prime - avoid building outfield points that
don't extend the prime in front of X's back men without good reason.
Hope that's helpful!
Julian
**********************************************************************
Andrew Bokelman
Date: Tue, 22 Dec 1998 14:33:56 -0500
> VSG writes:
> I mean it seems to me that it shouldn't be too difficult to bear in as
> you can easily move behind? I think there must be something I'm missing
> here :)
Your opponent can move behind you but the moves are more awkward. With
each move your opponent is trying not to leave an exposed checker for
you to hit, and sometimes this can lead to piling up chips on a few points
and/or leaving blots inside.
Scenario 1.
While your opponent is trying to bear in, you are building a home board
with even distribution in case you hit. You don't hit, but when you run
your board is set up better for bearoff.
Scenario 2.
Your home board breaks during the holding game and your opponent is not
in as much danger bearing in. But because your home board broke you are
now playing checkers to the inner most points. When you finally run, your
board is in even better shape for bear-off than in Scenario 1.
Scenario 3.
Your opponent has played checkers behind you but there are gaps and blots.
And you have built a strong home board. So you can consider running
one checker from your holding point and continue to hold with the
other one for maybe a couple of rolls. This may make the difference in
putting you ahead in the race.
Scenario 4.
Because you have left checkers back to hold, you are able to hit when your
opponent is forced to break an outside point. This either puts you
ahead in the race, or your home board is strong enough that you
opponent gets stuck behind it and you are able to pull ahead in the race.
***************************************************************************
Robert-Jan Veldhuizen
Date: 23 Dec 98 07:23:54 +0100
Andrew Bokelman wrote:
> Scenario 2.
>
> Your home board breaks during the holding game and your opponent is not
> in as much danger bearing in. But because your home board broke you are
> now playing checkers to the inner most points. When you finally run,
> your board is in even better shape for bear-off than in Scenario 1.
Well, it depends on what you mean by "better shape". Absolutely
speaking, lower points in your inner board represent less pips and so
they are usually quicker to clear. But relatively speaking, moving
checkers from your 6, 5 and 4 points to your 3 and especially 2 and 1
points, is putting you behind in the race because you will probably be
wasting more pips during the bear-off.
So, if you have a choice between (safely) leaving your anchor and start
racing with an (almost) equal or even lower pipcount than your opponent,
I would prefer that to breaking my board.
--
Zorba/Robert-Jan
***************************************************************************
Chuck Bower
Date: 22 Dec 1998 21:56:40 GMT
David Montgomery wrote:
> VSG writes:
> > What is a holding game?
(David responds:)
> The term isn't really well-defined.
Excellent point.
> (snip)
> Holding games are distinguished from backgames primarily by when you
> intend to hit your opponent. In a backgame, often the plan is not
> to hit until your opponent is bearing off. In a holding game, usually
> you will hit as soon as possible, as long as your offense isn't in a
> shambles. (However, in advanced backgames, like a 54 or 53 backgame,
> you usually do intend to hit while the opponent is bearing in.)
Again, I agree. Note that some people call the 20,21 and 20,22
point backgames (David's more common notation '54 or 53 backgame(s)')
"holding games", which further emphasizes his first comment about holding
games (and backgames, for that matter) not being so well defined.
I think David's sentence: "In a backgame, often the plan is not
to hit until your opponent is bearing off" is worth elaborating upon.
Here I three positions to ponder:
+24-23-22-21-20-19-+---+18-17-16-15-14-13-+
| O O | | O O O |2
| O | | O |
| O | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| O X X | | X |
| O O X X X | | X X O |
| X O O X X X | | X X X O |
+-1--2--3--4--5--6-+---+-7--8--9-10-11-12-+
O on to play 62.
Here O is not ready to hit X, since his/her board is in such a
premature state. (Also, X will maintain a nice 5-prime, possibly
choking O. Lastly, giving up one of the two anchors this early is
clearly suicidal.)
Now look at:
+24-23-22-21-20-19-+---+18-17-16-15-14-13-+
| O O O O | | O O |2
| O O O | | O O |
| | | |
| | | |
| | | |
| | | |
X| | | |
X| X | | |
X| X | | |
X| X | | |
X| X O O X | | |
X| X O O X X X | | |
+-1--2--3--4--5--6-+---+-7--8--9-10-11-12-+
REW NOTE: Corrected later to read:
O on Roll (**X on roll**). Cube decisions by both sides?
Now O's side of the board is ready! X will be almost defenseless
if O can hit. 27/36 numbers hit, and most of those will be game winners.
One thing which should be clear: X has a take. If O misses X can
probably cash, but even if O will have a subsequent take the rewhip to
8 has to be juicy. And even if hit, X will scrape out a few wins.
(I would pass as O after missing the triple shot.) Does O have a redouble
in the illustrated position? I'm not sure if there are sufficient market
losers but against most human opponents this is a nice pressure redouble.
I would cube.
Finally look at:
+24-23-22-21-20-19-+---+18-17-16-15-14-13-+
| O O O O | | O |2
| O O O O | | O |
| | | |
| | | |
| | | |
| | | |
| X | | |
| X | | |
| X X | | |
| X X X | | |
| X O O X X | | |
| X O O X X X | | X O |
+-1--2--3--4--5--6-+---+-7--8--9-10-11-12-+
X on roll. Cube decisions by both sides?
X is smarting after rolling 65. Now a take looks misguided
mostly because X has failed to bear off a checker, in contrast to the
previous position. I suppose a case can be made for O being too good
but I think a redouble is O's best move against fallible opposition.
(There are probably some 'steamers' who would take.)
The fact that I used the 32 backgame (23,22 from O's POV) is no
coincidence. These seem to produce a reasonable fraction of the last
type of position illustrated (gammonee turning into the gammonER!).
Again, I point out that David was aware of these exceptions (he said
"...often...", not "...always..."). I just thought his post was a
nice lead-in to touting one of the occasional bonuses of playing
backgames. As far as the downsides.... Well, I think most of us are
all too painfully aware of those.
Chuck
bower@bigbang.astro.indiana.edu
c_ray on FIBS
***************************************************************************
Harald Retter
Date: Wed, 23 Dec 1998 13:02:55 GMT
Chuck Bower wrote:
> <SNIP>
> +24-23-22-21-20-19-+---+18-17-16-15-14-13-+
> | O O O O | | O O |2
> | O O O | | O O |
> | | | |
> | | | |
> | | | |
> | | | |
> X| | | |
> X| X | | |
> X| X | | |
> X| X | | |
> X| X O O X | | |
> X| X O O X X X | | |
> +-1--2--3--4--5--6-+---+-7--8--9-10-11-12-+
>
> X on roll. Cube decisions by both sides?
>
> Now O's side of the board is ready! X will be almost defenseless
> if O can hit. 27/36 numbers hit, and most of those will be game winners.
Not if you cubed before. And even if, I disagree.
> One thing which should be clear: X has a take. If O misses X can
> probably cash, but even if O will have a subsequent take the rewhip to
> 8 has to be juicy. And even if hit, X will scrape out a few wins.
> (I would pass as O after missing the triple shot.)
The pass is mandatory, I am more curious, if it would be too good, note
that X wins a pretty ammount of Backgammons!
> Does O have a redouble
> in the illustrated position? I'm not sure if there are sufficient market
> losers but against most human opponents this is a nice pressure redouble.
> I would cube.
With 6 chechers born off? Chuck be serious, X has plenty of live after a
hit, even after a double hit! And don't forget, there is only one side,
which can be gammoned. I wouldn't turn it, even if it was an initial
double and even against the majority of human players (Yes, there are
exceptions). Never turn a cube that looks like serious blunder when taken
(/beavered? ;-)), is a good advice to follow. Hit and look what happens,
O will have plenty of Take-potential after a lot of hitting sequences.
And even if your market is lost, it will *never* be lost by far, not in
a single sequence of 1296!
Regards, Harald Retter
***********************************************************************
Chuck Bower
Date: 23 Dec 1998 16:56:27 GMT
haraldretter@my-dejanews.com wrote:
> Chuck Bower) wrote:
> > +24-23-22-21-20-19-+---+18-17-16-15-14-13-+
> > | O O O O | | O O |2
> > | O O O | | O O |
> > | | | |
> > | | | |
> > | | | |
> > | | | |
> > X| | | |
> > X| X | | |
> > X| X | | |
> > X| X | | |
> > X| X O O X | | |
> > X| X O O X X X | | |
> > +-1--2--3--4--5--6-+---+-7--8--9-10-11-12-+
> >
> > X on roll. Cube decisions by both sides?
(Hopefully most of you have realized I meant "O on roll.")
> > Now O's side of the board is ready! X will be almost defenseless
> > if O can hit. 27/36 numbers hit, and most of those will be game
> > winners.
(snip)
> > Does O have a redouble
> > in the illustrated position? I'm not sure if there are sufficient
> > market losers but against most human opponents this is a nice pressure
> > redouble. I would cube.
Harald corrected:
> With 6 chechers born off? Chuck be serious, X has plenty of live after
> a hit, even after a double hit! And don't forget, there is only one side,
> which can be gammoned. I wouldn't turn it, even if it was an initial
> double and even against the majority of human players (Yes, there are
> exceptions). Never turn a cube that looks like serious blunder when taken
> (/beavered? ;-)), is a good advice to follow.
> Hit and look what happens, O will have plenty of Take-potential after a
> lot of hitting sequences. And even if your market is lost, it will
> *never* be lost by far, not in a single sequence of 1296!
Well, I guess I went swimming in the r.g.bg sea without a (robot)
life-preserver (again) and got attacked by a shark! I'm bleading
profusely, but the beach is in sight....
I suppose I could argue (like some on this newsgroup would) that "most
human opponents" are so bad that it would be a good cube against THEM. You
know, all the millions of people who've played BG a few times; long enough
to know the rules (and long enough to realize they don't like the game...).
OTOH, those kind of players NEVER drop a cube! I guess I'm the one left
defenseless.
It would be interesting to know, following A) a single hit this time
and B) a double hit this time what the equities look like. If O can close
out ONE checker I estimate s/he will be about a 2/1 favorite cubeless
(which isn't as strong here since X will be owning the cube). Obviously
getting a second checker back will be a LOT better. The problem doesn't
appear to be getting checkers back there, but KEEPING them there (and, when
two are back, keeping them from anchoring). O has to get checkers around
to complete the 6-prime. X tries and nuzzle up to the edge of the 5-prime
to hop it.
Can we trust the robots here? David Montgomerey has found weaknesses
in SOME of the robots when it comes to getting a second checker back--the
weakness is that they don't try hard enough to accomplish this. I don't
know if he has had time to investigate this with the latest technology.
I have noticed myself that SOMETIMES the deepest lookahead evaluations can
be CONSIDERABLY OFF compared to rollouts in these late game positions.
I guess I should do some rollouts, see if they agree with me, and if
they do, post that result and say emphatically "...END OF STORY". If they
don't then I waffle about robots not knowing how to finish backgames, play
a (statistically insignificant) number of me.vs.robot games to get a feel
of how to play the position (and where the robots are screwing up ;) and
form a private opinion, all the while dodging the sharks who want to play
this position against me as a prop. Or I could just admit I'm wrong. But
it's too early for that. Some baywatch-babe-lookalike lifeguard is
swimming towards me now. Maybe Harald will go after her and I can get
away!
Chuck
bower@bigbang.astro.indiana.edu
c_ray on FIBS
|
![[Long message]](pics/Long.gif)
![[GammOnLine forum]](pics/GammOnLine.gif)
![[Recommended reading]](pics/Star2.gif)
Recommended reading
![[Recent addition]](pics/New.gif)
Recent addition