Game #1
Open Division Final
San Mateo, California
15 September 1996
11 Point Match
Ron Karr (Black) vs. Robert Caradien (White)
Score: 3 - 0
Analysis by Steve Clark
This is the fourth game of the Open Division finals,
between Ron Karr and Robert Caradien.
The winner's circle was familiar territory to Robert Caradien.
In his sixth outing at Backgammon by the Bay,
this was his fourth trip to the championship round.
Ron Karr, not a stranger himself to the prize round,
came out swinging,
winning the first three games of the match.
The score is 3-0, as we pick up the action ...
| Play 1a |
|
|
Black to play 51
Candidate Plays Equities
13/8 24/23 +0.009
13/8 6/5 -0.030
24/18 -0.030
13/7 -0.067
|
|
SC:
|
Rollouts show the split 24/23 narrowly leading the slot of the 5 point.
Most experts still prefer to slot the 5.
Some still think slotting
is the better play.
Others believe the increased complexity of the game compensates for the slight loss of equity.
My own preference is the split
13/8 24/23.
|
| |
Black
|
White
|
|
1.
|
51: 13/8 6/5
|
|
|
| Play 1b |
|
|
White to play 62
Candidate Plays Equities
24/16 -0.171
24/18 13/11 -0.192
13/5 -0.194
13/7 13/11 -0.194
13/7 24/22 -0.216
|
|
SC:
|
13/5 was the favorite response for many years
but is now believed to be slightly inferior to
24/16,
or 24/18 13/11.
|
| |
Black
|
White
|
|
1.
|
... |
62: 13/5
|
|
|
| Play 2a |
|
|
Black to play 51
Candidate Plays Equities
13/8 6/5 +0.078
8/3 6/5 +0.057
|
|
SC:
|
Making the 5 point with the ace is clear.
Slotting plays such as 8/3 6/5 are not supported by the rollouts.
These slotting plays are made much less often than they used to be.
Further, rollouts show a fourth builder on the 8 point is quite useful.
|
| |
Black
|
White
|
|
2.
|
51: 8/3 6/5
|
|
|
| Play 2b |
|
|
White to play 31
Candidate Plays Equities
13/10 6/5 -0.040
24/21 6/5 -0.056
8/5 6/5 -0.067
8/5 24/23 -0.128
|
|
SC:
|
6/5 is the best way to make the 5 point
because of the stack of checkers on the 6.
The various possible 3's are about equal.
|
| |
Black
|
White
|
|
2.
|
... |
31: 13/10 6/5
|
|
|
| Play 3a |
|
|
Black to play 44
Candidate Plays Equities
13/9(2) 8/4(2) +0.056
13/5 8/4(2) +0.018
13/9 8/4(3) +0.013
|
|
SC:
|
After you make the 4 point with two of the 4's,
it seems clear to make the 9 with the remainder.
The 9 point will be a very useful point until the 3 point is made.
|
| |
Black
|
White
|
|
3.
|
44: 13/9(2) 8/4(2)
|
|
|
| Play 3b |
|
|
White to play 61
Candidate Plays Equities
24/17* +0.014
13/7 8/7 -0.044
|
|
SC:
|
Four-point primes are nice but hitting is a lot nicer.
|
| |
Black
|
White
|
|
3.
|
...
|
61: 24/17*
|
|
4.
|
65: Cannot move
|
52: 24/22*/17
|
5.
|
32: Bar/22 Bar/23
|
|
|
| Play 5b |
|
|
White to play. Double or roll?
Equity (center cube)
White 85.4% G 1.5% BG 0.1%
Black 14.6% G 0.7% BG 0.0%
Equity (White): +0.717
Equity (Black owns 2 cube)
White 58.4% G 22.2% BG 1.3%
Black 41.6% G 1.4% BG 0.1%
Equity (Black): -0.389
Equity (cubeless)
White 66.6% G 23.7% BG 1.4%
Black 33.4% G 7.5% BG 0.3%
Equity (White): +0.505
|
|
SC:
|
A clear double and take.
JellyFish settlement equities are not set correctly,
so it often is too conservative in these types of positions.
|
|
JF:
|
I would not have doubled (level 7 evaluation).
|
| |
Black
|
White
|
|
5.
|
...
|
Double -> 2
|
|
6.
|
Accept
|
|
|
| Play 6b |
|
|
White to play 51
Candidate Plays Equities
10/5 6/5 +0.380
8/3*/2* +0.363
13/8 6/5 +0.288
|
|
SC:
|
White rolls a poor number.
Hitting twice is not particularly attractive
because Black has the stronger board
and a blot on the 2 point is too far advanced.
It does have the advantage of completely stopping Black's development.
Moving two checkers to the 5 point is safe
but the men are not particularly well distributed.
In balance, neither play stands out.
|
| |
Black
|
White
|
|
6.
|
...
|
51: 8/3*/2*
|
|
7.
|
43: Bar/21 Bar/22
|
|
|
| Play 7b |
|
|
White to play 65
Candidate Plays Equities
13/2 +0.274
8/2 8/3* +0.217
|
| |
Black
|
White
|
|
7.
|
...
|
65: 8/2 8/3*
|
|
8.
|
66: Cannot move
|
|
|
| Play 8b |
|
|
White to play 43
Candidate Plays Equities
13/9 6/3 +0.699
10/3 +0.631
17/13 6/3 +0.603
|
|
SC:
|
The benefits of leaving blots in the outfield outweigh the risks.
Black will not be particularly happy to hit even if he does roll a joker.
|
| |
Black
|
White
|
|
8.
|
... |
43: 10/3
|
|
|
| Play 9a |
|
|
Black to play 21
Candidate Plays Equities
Bar/24 6/4 -0.707
Bar/24 9/7 -0.797
|
| |
Black
|
White
|
|
9.
|
21: Bar/24 6/4
|
|
|
| Play 9b |
|
|
White to play 51
Candidate Plays Equities
13/8 6/5 +0.586
13/7 +0.549
|
|
SC:
|
It is interesting that 13/7 is a candidate.
I would not have considered this play very long.
|
| |
Black
|
White
|
|
9.
|
... |
51: 13/8 6/5
|
|
|
| Play 10a |
|
|
Black to play 33
Candidate Plays Equities
24/21 9/3 9/6 -0.495
24/18 21/18 9/6 -0.496
24/21 13/10(2) 9/6 -0.516
24/21 13/7 13/10 -0.548
24/21 13/10(2) 4/1 -0.582
|
|
SC:
|
Black would rather defend from the 18 point than the 21.
Unfortunately he is then forced to leave a blot with 9/6,
as the alternative of 4/1 is very unappealing.
This brings the two plays into balance.
|
| |
Black
|
White
|
|
10.
|
33: 24/21 9/3 9/6
|
43: 17/13 8/5
|
|
|
| Play 11a |
|
|
Black to play 43
Candidate Plays Equities
13/9 6/3 -0.546
6/2 5/2 -0.583
5/1 4/1 -0.639
21/17 6/3 -0.684
6/2 24/21 -0.742
|
|
SC:
|
6/2 24/21 (played at the table) can't be right.
It weakens the defense while avoiding building a board.
13/9 6/3 would put Black's home board together while keeping his backgame points.
|
| |
Black
|
White
|
|
11.
|
43: 6/2 24/21
|
|
|
| Play 11b |
|
|
White to play 43
Candidate Plays Equities
5/1* 17/14 +0.737
5/1* 13/10 +0.675
|
|
SC:
|
5/1* 17/14 is best.
Extra blots in the outfield are not correct.
They just ask for some disaster sequence.
|
| |
Black
|
White
|
|
11.
|
...
|
43: 5/1* 17/14
|
|
12.
|
21: Bar/24* 4/2
|
|
|
| Play 12b |
|
|
White to play 32
Candidate Plays Equities
Bar/22* 13/11 +0.588
Bar/22* 5/3 +0.539
|
|
SC:
|
Here the risk of an outfield blot is less than on the last play,
because the blot on the 11 point is out of direct range.
Also, Black cannot hit two blots on this roll.
|
| |
Black
|
White
|
|
12.
|
...
|
32: Bar/22* 13/11
|
|
13.
|
65: Cannot move
|
|
|
| Play 13b |
|
|
White to play 53
Candidate Plays Equities
22/14 +0.771
6/1* 14/11 +0.722
22/17 14/11 +0.720
14/6 +0.700
|
|
SC:
|
The rollout results demonstrate the benefits of safety.
The hit on the 1 point is attractive, but Black does have a four-point board.
If he rolls a 1 from the bar, he could be turning the whole game around.
In contrast, if White plays safely,
Black has many awkward plays even if he does come in.
|
| |
Black
|
White
|
|
13.
|
...
|
53: 6/1* 14/11
|
|
14.
|
64: Bar/21 0
|
|
|
| Play 14b |
|
|
White to play 52
Candidate Plays Equities
22/15 +0.577
13/6 +0.536
11/6 3/1 +0.516
11/6 11/9 +0.497
13/8 13/11 +0.457
|
|
SC:
|
It seems right to try to escape if you can't cover the ace point.
|
| |
Black
|
White
|
|
14.
|
... |
52: 22/15
|
|
|
| Play 15a |
|
|
Black to play 61
Candidate Plays Equities
Bar/24*/18 -0.060
Bar/24* 21/15 -0.089
|
|
SC:
|
These plays look about equal to me.
24/18 leaves fewer blots.
21/15 gives more flexibility and more opportunities to hit (and be hit).
|
| |
Black
|
White
|
|
15.
|
61: Bar/24* 21/15
|
43: Bar/22 15/11
|
|
|
| Play 16a |
|
|
Black to play 31
Candidate Plays Equities
21/17 -0.711
24/21 2/1 -0.775
|
|
SC:
|
The rollout results show how grim it is to break up your own board.
The problem with 24/21 2/1
(the play made at the table)
is that Black no longer has a threat of hitting.
Leaving extra blot around is scary,
but it is important to retain an ability to strike back.
|
| |
Black
|
White
|
|
16.
|
31: 24/21 2/1
|
|
|
|
| Play 16b |
|
|
White to play 54
Candidate Plays Equities
22/13 +0.630
11/6 22/18 +0.621
11/6 5/1 +0.547
|
|
SC:
|
This rollout illustrates
the problems with the previous play by Black.
11/6 22/18, leaving a trailing blot, is almost as good as 22/13.
|
| |
Black
|
White
|
|
16.
|
...
|
54: 22/13
|
|
17.
|
21: 15/13 2/1
|
|
|
| Play 17b |
|
|
White to play 31
Candidate Plays Equities
13/10 11/10 +0.611
5/1 +0.572
|
|
SC:
|
Expert thinking used to be that you should not make this play.
Rather you should try to bring your checkers home without leaving a direct shot.
Thus the 5/1 play would have been favored.
The rollouts have demonstrated the greater flexibility
of providing stepping stones along the way.
The modern thinking recognizes
that White will be able to safely clear his position
with doubles a substantial part of the time.
It used to be thought that doubles for White were a rare and lucky break
rather than a common result.
|
| |
Black
|
White
|
|
17.
|
...
|
31: 13/10 11/10
|
|
|
| Play 18a |
|
|
Black to play 33
Candidate Plays Equities
21/18(2) 13/7 -0.417
21/18(2) 13/10(2) -0.493
21/18(2) 5/2(2) -0.530
|
|
SC:
|
Here are some doubles now, giving Black a much improved position.
|
| |
Black
|
White
|
|
18.
|
33: 21/18(2) 13/7
|
|
|
|
| Play 18b |
|
|
White to play 31
Candidate Plays Equities
5/1 +0.397
11/8 11/10 +0.332
|
|
SC:
|
Although playing behind Black's anchor weakens White's position,
the safe play is the best play here.
|
| |
Black
|
White
|
|
18.
|
...
|
31: 5/1
|
|
|
| Play 19a |
|
|
Black to play 31
Candidate Plays Equities
13/10 7/6 -0.406
13/9 -0.417
7/3 -0.432
|
|
SC:
|
This rollout result demonstrates
that Black really does not have to fear being hit.
Even if hit, he will have a strong return shot at the ace point.
|
| |
Black
|
White
|
|
19.
|
31: 7/3
|
|
|
|
| Play 19b |
|
|
White to play 62
Candidate Plays Equities
13/5 +0.377
11/5 11/9 +0.358
11/5 3/1 +0.341
11/3 +0.306
|
|
SC:
|
Minimizing shots is not the important factor here.
The checker moved to the 3 point is misplaced.
Also White will have to clear the 13 point soon in any case.
When clearing points,
especially when getting hit is not a total disaster,
you should clear the point that will cause the most difficulties in the future.
13/5 is correct.
|
| |
Black
|
White
|
|
19.
|
...
|
62: 11/3
|
|
20.
|
66: 13/1(2)
|
|
|
| Play 20b |
|
|
White to play 11
Candidate Plays Equities
13/12(2) 3/1 +0.752
13/11(2) +0.708
13/12(2) 11/10 2/1 +0.689
11/9 10/9(2) +0.689
13/12(2) 11/10 3/2 +0.675
|
|
SC:
|
This rollout demonstrates the value of a strong home board.
If Black could easily strengthen his own board,
safety would be more important.
But here, White should make his own board.
|
| |
Black
|
White
|
|
20.
|
...
|
11: 13/11(2)
|
|
|
| Play 21a |
|
|
Black to play 31
Candidate Plays Equities
6/3 6/5 -0.783
5/2 3/2 -0.788
4/1 4/3 -0.814
5/2 5/4 -0.827
5/1 -0.877
|
|
SC:
|
There is not really much to choose between the plays.
|
| |
Black
|
White
|
|
21.
|
31: 6/3 6/5
|
53: 11/3
|
|
22.
|
44: 21/13(2)
|
51: 11/6 11/10
|
|
| Play 23a |
|
|
Black to play 51
Candidate Plays Equities
13/8 13/12 -0.647
13/8 3/2 -0.664
13/7 -0.665
13/8 5/4 -0.668
13/8 4/3 -0.681
|
| |
Black
|
White
|
|
23.
|
51: 13/8 13/12
|
|
|
|
| Play 23b |
|
|
White to play 32
Candidate Plays Equities
10/5 +0.590
5/2 3/1 +0.575
6/3 6/4 +0.563
6/1 +0.562
5/2 5/3 +0.509
|
|
SC:
|
It is usually best in these circumstances
to move the third checker off the 10 point.
This has two advantages.
First, there is no wastage in the race.
Second, White can now clear the 10 point with rolls like 64.
White's overall chances of being hit are not so high
that he should give up the wastage of 6/1.
Even top flight experts spend too much time saving 6's to play.
|
| |
Black
|
White
|
|
23.
|
...
|
32: 6/1
|
|
|
| Play 24a |
|
|
Black to play 21
Candidate Plays Equities
5/2 -0.657
12/9 -0.673
8/6 12/11 -0.675
12/10 8/7 -0.676
8/6 5/4 -0.679
|
|
SC:
|
There are two rules to follow in this position:
First, slot points;
second, do not waste pips.
Slotting greatly improves your racing chances and increases your chances
of building a stronger board next roll.
This is particularly important when your opponent (as here)
will not be leaving a shot next roll.
Minimizing wastage
also improves your racing chances and increases your overall flexibility.
While your racing chances are not high here,
you should do everything to maximize the chances you have.
There are two plays which conform to the above rules: 5/2, and 8/6 12/11.
5/2 clearly is best for building the board and spreads you checkers properly
if it becomes a race.
It has the weakness of not going after the most important point
and also increases the gammon risk.
8/6 12/11 goes after the right point but leaves a third checker
on the 5 point while it clearly belongs on the 2.
Overall 5/2 seems best.
|
| |
Black
|
White
|
|
24.
|
21: 8/6 5/4
|
|
|
|
| Play 24b |
|
|
White to play 53
Candidate Plays Equities
10/5 6/3 +0.666
10/2 +0.652
6/1 6/3 +0.597
6/1 5/2 +0.576
|
| |
Black
|
White
|
|
24.
|
...
|
53: 10/2
|
|
|
| Play 25a |
|
|
Black to play 65
Candidate Plays Equities
12/1 -0.652
12/6 18/13 -0.689
18/12 18/13 -0.762
|
|
SC:
|
Just when Black has his shot opportunity coming, he gives it up.
It is often difficult to balance the opportunity lost in the race
with the opportunity lost from the hit.
Here Black goes wrong,
as we all do from time to time.
|
| |
Black
|
White
|
|
25.
|
65: 18/13 18/12
|
|
|
|
| Play 25b |
|
|
White to play
|
|
White won the race and (later) the match (11-3).
|
The game was recorded by Patti Beadles.
Rollouts were made by Richard McIntosh,
using
JellyFish Analyzer 2.02.
Rollout results show equities
for the player on move.
Candidate plays were better than or
within 0.100 equity of the actual plays,
evaluated at level 7.
Parameter
values for rollouts on moves were:
- level 5
- 1296 games (36x36)
- horizon 7
- seed 4577
Standard deviations of equity estimates were between 0.010 and 0.015.
Parameter values for rollouts on cube decisions were:
- level 5
- 23328 games (36x648)
- full game
- seed 4577
- settlement limit 0.570
Copyright © 1996-2010
BackGammon By the Bay