Strategy--Checker play

Forum Archive : Strategy--Checker play

How to trap an anchor

From:   Timothy Chow
Date:   27 April 2010
Subject:   How to trap an anchor
Forum: Forums

Mike Corbett's book gives some valuable tips on trapping but I have not
found any other good source of advice about how to trap. In the position
below, should I just shuffle my free checkers and wait for a better
opportunity, or should I try to trap now? If I try to trap, how should I do
it? Can bot rollouts be trusted for this kind of decision?

      24  23  22  21  20  19      18  17  16  15  14  13
 OO  | O   O               X |   |         X             |
 OO  | O   O                 |   |                       |
  O  | O   O                 |   |                       |
  O  |                       |   |                       |
  O  |                       |   |                       |
     |                       |   |                       |  X rolls 2-1
     |                       |   |                       |
     |                       |   |                       |
     |                       |   |         X             |
     |                 O   X |   | X   X   X   X   X     |
     |                 O   X |   | X   X   X   X   X     |
       1   2   3   4   5   6       7   8   9  10  11  12

Bob Koca  writes:

Here is another question to ask yourself now since it may influence several
plays down the line. Would X prefer to have one of the anchored checkers
on O's 2 point instead?

John O'Hagan  writes:

I think you need to get X off that anchor now. Since X has just a 2-
pt board, I guess I'd try 11/8. That pays off big time to D6 and, to a
lesser degree 65, but I'd take that chance. Maybe breaking a point from 7
through 10 is better since X has to play any two or higher from the
anchor and the lower double corresponding to the point you've opened isn't
as damaging as D6 is if you break the 11-point, plus there aren't any
double hitters. I don't think these considerations are enough to swing the
decision so I'd keep the 5 in a row and break from the back. X's weak 2-
point inner board makes leaving the blot on the 11-point worth the risk
IMHO. If X had a 3-point board, leaving the blot might be too risky.
Tough problem.

Stick  writes:

I think 8/7, 8/6 and I think it's clear. This puts max flex with no direct
shots and causes lots of breaking number [31 32 34 35 36] with four of
those coughing up two blots without either of them going much of anywhere.
I don't really care what a rollout says because later on in more 'should I
trap or not' situations will likely arise and the bot may not make my
choice there either.

Your opponent's board is as weak as it needs to be for you to trap and
better to do it sooner than later so you still have men back to pick up the
second checker should it hop out over the prime.

Matt-Cohn Geier  writes:

If you don't trap now then when the hell are you going to trap? Seems
pretty clear that you will lose the race otherwise.

I play 16/14, 7/6 here. 8/7, 8/6 is ok also but I definitely don't give
O a gin/TG 6-6.

Timothy Chow  writes:

I have a question about the merits of 16/14, 7/6 versus 8/7, 8/6, which is
also related to Bob Koca's comment. Is it typically right to leave a
slotted blot when it's important to try to pick up both checkers?

My reasoning would be that leaving a slotted blot gives more outfield
coverage to catch the opponent's escaping blot whereas lifting it ensures
that I can use both dice to try to attack the opponent's remaining blot. So
perhaps if O had fewer checkers off then 8/7 8/6 would be preferred
because I would want to make sure to contain at least one blot? And maybe
then I would want to bring more checkers up front before trapping? Or am I
not thinking about this right?

Fabrice Liardet  writes:

For me there are two main criteria for leaving a blot vs leaving a hole.

- The obvious number one is the size of the opponent's board. A two-point
  board means that I don't lose much by leaving a blot.

- The number two is outfield control. There is a plus side in leaving a
  blot for recirculation when one has advanced too much and the opponent
  can soon escape a checker completely when not hitting. That is not the
  case here, so I don't gain much either by leaving a blot.

The bottom line is that it does not to make a big difference in that
position, and although it might be slightly better not to leave a blot, I
would happily give the priority to tactical considerations.

Needless to say, I am also struggling with the general strategy for
containing two checkers. Even more so when the prime is farther from home
and the anchor is not even in the home board. I'll come back to that thread
in the hope to grasp one or two guidelines.

Petter Bengtsson wrote:

> I'd play 7/4 here. 8/6 8/7 is my second choice.

I didn't even think about it before reading you, but the more I look at it
the more I like 7/4. After all 2 is the smallest number that O can't
play on the other side. And starting the 4-point might reduce the chance
that O reanchors high after being hit.

An even harder question for me would me what should be done after the bar
point blot is missed. At what point does it pay to open the 8-point as

Neil Kazaross  writes:

8/7, 8/6 puts builders in a nice spot and causes O to puke if he rolls
any 3 except 3-3. We can decide whether to break another blocking point
soon if need be.

Casper van der Tak  writes:

I'd play 8/7, 8/6 as well, because it puts checkers in good positions,
traps O with single 3s, and does not leave jokers. Best to trap early
here, I think.

I don't know whether I'd trust bot rollouts from here -- would be
interesting to compare bot results, it would not surprise me if some bots
do a lot better than others in this particular case.

Timothy Chow  writes:

My play was 8/7, 8/6. I've rolled out only GNU's top 5 choices. Everything
else evaluates to +0.5 or below on 2-ply so, although they might rise in
the rankings upon rollout, I'm not inclined to spend the CPU time unless
someone persuades me it's worth doing so. The next highest prime-breaking
play (not shown below) is 7/4 according to 2-ply.

1.  16/15, 9/7     +0.752
2.  16/13          +0.745 ( -0.008)
3.  9/6            +0.744 ( -0.009)
4.  16/14, 7/6     +0.734 ( -0.018)
5.  8/7, 8/6       +0.707 ( -0.045)

Casper van der Tak  writes:

I played this out a number of times against GNU, and won every game with
the cube on 4 - including a few stone-cold passes. I don't trust the
rollout at all.

Stick  writes:

Here is an XG 4/4 rollout.

1.  9/6             +0.704
2.  16/13           +0.689 (-0.015)
3.  7/4             +0.681 (-0.023)
4.  8/7, 8/6        +0.664 (-0.040)
Did you find the information in this article useful?          

Do you have any comments you'd like to add?     


Strategy--Checker play

Avoiding major oversights  (Chuck Bower+, Mar 2008) 
Bearing off with contact  (Walter Trice, Dec 1999) 
Bearing off with contact  (Daniel Murphy, Mar 1998)  [Long message]
Blitzing strategy  (Michael J. Zehr, July 1997) 
Blitzing strategy  (Fredrik Dahl, July 1997) 
Blitzing technique  (Albert Silver+, July 2003)  [GammOnLine forum]
Breaking anchor  (abc, Mar 2004) 
Breaking contact  (Alan Webb+, Oct 1999) 
Coming under the gun  (Kit Woolsey, July 1996) 
Common errors  (David Levy, Oct 2009) 
Containment positions  (Brian Sheppard, July 1998) 
Coup Classique  (Paul Epstein+, Dec 2006) 
Cube ownership considerations  (Kit Woolsey, Apr 1996) 
Cube-influenced checker play  (Rew Francis+, Apr 2003)  [GammOnLine forum]
Defending against a blitz  (Michael J. Zehr, Jan 1995) 
Estimating in volatile situations  (Kit Woolsey, Mar 1997) 
Gammonish positions  (Michael Manolios, Nov 1999) 
Golden point  (Henry Logan+, Nov 2002) 
Hitting loose in your home board  (Douglas Zare, June 2000) 
Holding games  (Casual_Observer, Jan 1999)  [Long message]
How to trap an anchor  (Timothy Chow+, Apr 2010) 
Jacoby rule consideration  (Ron Karr, Nov 1996) 
Kamikaze plays  (christian munk-christensen+, Nov 2010) 
Kleinman Count for bringing checkers home  (Øystein Johansen, Feb 2001) 
Late loose hits  (Douglas Zare+, Aug 2007)  [GammOnLine forum]
Mutual holding game  (Ron Karr, Dec 1996) 
Pay now or pay later?  (Stuart Katz, MD, Nov 1997) 
Pay now or pay later?  (Stephen Turner, Mar 1997) 
Pay now or play later?  (Hank Youngerman+, Sept 1998) 
Play versus a novice  (Courtney S Foster+, Apr 2004)  [GammOnLine forum]
Playing doublets  (Grunty, Jan 2008) 
Playing when opponent has one man back  (Kit Woolsey, May 1995) 
Prime versus prime  (Albert Silver+, Aug 2006)  [GammOnLine forum]
Prime versus prime  (Michael J. Zehr, Mar 1996) 
Saving gammon  (Bill Riles, Oct 2009) 
Saving gammon  (Ron Karr, Dec 1997) 
Splitting your back men  (KL Gerber+, Nov 2002) 
Splitting your back men  (David Montgomery, June 1995) 
Trap play problem  (Brian Sheppard, Feb 1997) 
When in doubt  (Stick+, Apr 2011) 
When to run the last checker  (Stick Rice+, Jan 2009) 
When you can't decide  (John O'Hagan, Oct 2009) 

[GammOnLine forum]  From GammOnLine       [Long message]  Long message       [Recommended reading]  Recommended reading       [Recent addition]  Recent addition

  Book Suggestions
Computer Dice
Cube Handling
Cube Handling in Races
Extreme Gammon
Fun and frustration
GNU Backgammon
Luck versus Skill
Magazines & E-zines
Match Archives
Match Equities
Match Play
Match Play at 2-away/2-away
Opening Rolls
Pip Counting
Play Sites
Probability and Statistics
Source Code
Strategy--Bearing Off
Strategy--Checker play


Return to:  Backgammon Galore : Forum Archive Main Page