Game #10
Open Division Final
San Mateo, California
15 June 1997
11 Point Match
Julius High (Blue) vs. Steve Clark (White)
Score: 0 - 0
Analysis by Ron Karr
Two very special BGBB folks played in the June finals. Julius High,
one of our stalwart supporters, makes it to his first final.
Steve Clark, dear to our hearts for his generous sharing of the game
analysis duties, cooperates by playing in the finals on a month when Ron
Karr was scheduled to do the commentary.
|
RK:
|
Split or slot: the age-old question. Certainly the plays are close.
JellyFish
and the other 'bots vote clearly for splitting, indicating that the loss
of
racing equity when hit is not worth the extra chance of making the 5
point quickly. Some good players continue to slot, arguing that JellyFish
undervalues the backgames and deep anchor games that can occur after a
hit. If in doubt, slotting with 41 seems less desirable than with 21
or
51 because the builder on the 9 point provides excellent chances to
make
a good point next time anyway.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
13/9 24/23 -0.002 49.6% 12.6% 0.3% 50.4% 12.0% 0.3%
24/20 24/23 -0.018 49.4% 10.9% 0.3% 50.6% 11.7% 0.3%
13/8 -0.021 49.2% 11.9% 0.3% 50.8% 12.4% 0.3%
13/9 6/5 -0.033 48.8% 12.2% 0.3% 51.2% 12.8% 0.5%
24/20 6/5 -0.038 48.7% 10.9% 0.3% 51.3% 11.9% 0.4%
13/9 8/7 -0.089 46.6% 11.6% 0.3% 53.4% 13.4% 0.5%
24/20 8/7 -0.098 46.3% 10.5% 0.3% 53.7% 12.8% 0.4%
6/2 24/23 -0.103 46.1% 10.8% 0.3% 53.9% 13.2% 0.4%
8/4 24/23 -0.110 46.0% 10.9% 0.3% 54.0% 13.7% 0.5%
|
|
RK:
|
White hits, then has a close choice whether to stay on the 5 point or run
into the outfield. Running out minimizes return shots, desirable now
that he's ahead in the race. However, if he does get hit, it hurts
because he has few return shots, and Blue's builder is in a good
place.
On the other hand, if White stays put and Blue hits loose, White is a
favorite to hit back. And the play guards the outfield, making it
risky
for Blue to bring builders down, while retaining the possibility of
making the anchor.
White Blue
Candidate Plays Equity Win G/BG BG Win G/BG BG
13/8 24/20* +0.164 55.1% 15.7% 0.8% 44.9% 10.1% 0.3%
24/20*/15 +0.131 55.1% 13.2% 0.5% 44.9% 10.5% 0.3%
|
|
RK:
|
Attacking on the 5 point seems fairly automatic here. Since Blue already
has three back and is behind in the race, having a fourth checker hit is
not so bad. White has no board and very few prospects, so there's no
urgency to making a more forward anchor. Blue should have no trouble
making a more advanched anchor later if that turns out to be necessary.
It would seem the potential gain from making the 5 point outweighs the
potential loss of racing equity when hit.
Somewhat surprisingly, JellyFish rollouts slightly prefer the conservative play
of making the 22 point. In hindsight, one can see some merits to this:
- Blue strips the 8 point by hitting loose. This reduces his possible
covers for the 5 point as well as removing a builder for the bar point.
He would have preferred to hit from the 6 point.
- One of the ways that hitting often gains is by preventing the opponent
from making new points, and that isn't much of a factor here, since White
has no extra builders. In addition, White has no really awkward numbers,
which can sometimes result from getting hit. For example, if White's
back checker were on the 23 point, a 62 from the bar would force him to
expose a new blot. Here, if White doesn't hit back, he simply brings
down a builder or moves a back checker.
- Making the 22 point anchor does have value, even if it's not urgent.
It reduces the value of White's 53, 33 and 55, and reduces White's
overall priming potential. It increases coverage of the outfield. If
Blue ends up with a backgame, any additional anchors he makes will go
well with the 3 point, while combinations that include the ace point may
be less effective (particularly the 14 and 15 games).
- If Blue anchors, he still leaves a 4-shot, but White also needs 4s to
make the 5-point anchor. Plus, he avoids getting hit twice with a number
like 54.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
Bar/22 24/22 -0.182 43.3% 9.8% 0.2% 56.7% 14.3% 0.5%
Bar/23 8/5* -0.192 44.6% 10.1% 0.3% 55.4% 17.9% 1.1%
Bar/23 24/21 -0.237 42.3% 9.3% 0.2% 57.7% 17.1% 0.7%
Bar/20 -0.239 42.2% 9.1% 0.2% 57.8% 16.8% 0.8%
Bar/23 9/6 -0.252 41.0% 8.6% 0.2% 59.0% 15.5% 0.6%
Bar/22 13/11 -0.261 42.0% 9.2% 0.2% 58.0% 18.5% 1.1%
Bar/22 6/4 -0.267 42.1% 8.7% 0.2% 59.3% 17.2% 0.9%
Bar/22 9/7 -0.271 41.1% 8.1% 0.2% 58.9% 16.8% 0.8%
Bar/22 8/6 -0.276 40.7% 9.0% 0.2% 59.3% 17.2% 0.9%
|
|
RK:
|
After entering, White has two attractive uses for two of the 2s:
- hit the blot on the 5 point, gaining further in the race and
reducing
Blue's offensive potential, or
- make his own 4 point, increasing the pressure on Blue's three
back checkers
The rollouts indicate a definite edge for hitting on the 5 point.
Despite White's unformed position, he gains more by preventing Blue from
doing anything. White should be able to make some points eventually.
White Blue
Candidate Plays Equity Win G/BG BG Win G/BG BG
Bar/21 24/20* +0.264 57.4% 19.1% 1.2% 42.6% 8.4% 0.2%
Bar/23 24/20* 13/11 +0.259 56.9% 19.8% 1.2% 43.1% 8.7% 0.3%
Bar/23 24/20*/18 +0.233 56.7% 18.1% 1.0% 43.3% 9.0% 0.2%
Bar/23 13/11 6/4(2) +0.217 57.0% 17.8% 0.8% 43.0% 10.7% 0.3%
Bar/23 24/20* 8/6 +0.217 55.3% 19.3% 1.2% 44.7% 9.1% 0.3%
Bar/23 24/20* 6/4 +0.192 55.7% 16.4% 0.9% 44.3% 9.2% 0.3%
Bar/23 24/22 6/4(2) +0.187 56.4% 16.1% 0.7% 43.6% 10.6% 0.3%
Bar/23 8/4 6/4 +0.184 55.8% 17.2% 0.8% 44.2% 10.9% 0.3%
Bar/21 6/4(2) +0.183 55.9% 16.5% 0.8% 44.1% 10.6% 0.3%
|
|
RK:
|
The further he falls behind, the clearer it is for Blue to play
positionally and battle for the 5 point rather than buttoning up.
Also, he gets to hit loose from the 6 point, which is an efficient use of
checkers.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
Bar/22 6/5* -0.267 44.0% 8.7% 0.3% 56.0% 22.0% 1.7%
Bar/22 23/22 -0.301 41.2% 7.8% 0.2% 58.8% 19.4% 1.1%
Bar/22 9/8 -0.345 39.5% 7.5% 0.2% 60.5% 20.2% 1.0%
Bar/21 -0.352 40.1% 7.6% 0.2% 59.9% 21.8% 1.4%
Bar/22 24/23 -0.359 39.9% 7.6% 0.2% 60.1% 23.3% 1.2%
|
|
RK:
|
Blue can't do much constructive here, so the best approach is to "put your
checkers where you want them" and come out to the opponent's bar point.
This may leave a few more shots than some of the alternatives, but it's
worth it since Blue is already far behind.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
Bar/18 -0.392 40.7% 6.8% 0.2% 59.3% 25.5% 2.1%
Bar/24 22/16 -0.418 40.1% 6.8% 0.2% 59.9% 26.7% 2.3%
Bar/24 9/3 -0.459 38.0% 6.6% 0.2% 62.0% 26.6% 2.2%
Bar/24 13/7 -0.475 39.6% 6.6% 0.2% 60.4% 30.1% 3.4%
|
|
RK:
|
White has a big racing lead, but must try to get something going
offensively. He could make the anchor and hit on the bar point. But
hitting two checkers is much stronger than hitting only one; despite
the
direct return shots on the 3 point, it gives White fewer total return
shots.
Which 4 is better? Hitting on White's bar brings a builder closer to
the
home board action, but hitting on 16 gains more in the race, and
retains
a triple shot at the other blot.
White Blue
Candidate Plays Equity Win G/BG BG Win G/BG BG
20/16* 6/3* +0.417 59.8% 26.8% 2.5% 40.2% 7.0% 0.2%
11/7* 6/3* +0.401 59.6% 26.0% 2.4% 40.4% 7.2% 0.2%
11/7* 23/20 +0.289 57.3% 20.3% 1.2% 42.7% 7.0% 0.2%
|
|
RK:
|
When I was first reviewing this game, I noticed that JellyFish 2.0,
Level
7, wanted to double, and put White's equity at +.499 cubeless. I
thought,
"no double. White has no board yet (although he'll be making a point
this
roll). Blue has an easy take, with all his checkers in play and good
chances of making two (or more) points in White's board. JellyFish tends to
undervalue positions with many checkers back." I thought JellyFish 3.0
might do better, since it's supposed to have better evaluations of
backgames. Lo and behold, JellyFish 3.0 Level 7 says the equity is +.477: no double.
However, the rollouts show a cubeless equity of +.512! And the full
rollouts
imply that White should double, since the center cube equity of +.713 is
less than twice that of Blue owning the cube (+.375).
So what is going on? First, it's possible that JellyFish makes more errors
for
Blue than for White when playing it out.
The Level 6 rollouts, which look ahead 1-ply, and
therefore play more accurately, indicate that this may be the case.
On the other
hand,
White has a clear edge, and no matter what happens, Blue won't be
turning
the game around quickly. Does White have any crushing sequences that
would
justify turning the cube now? Probably 44, 33 or 22 followed by a fan
would be serious market-losers. Is this enough to double? Could be.
But
in practice, it wouldn't be terribly surprising to find players taking
later in some positions where you technically have lost your market.
One thing that's certain: it's a take.
|
|
JF:
|
I would not have doubled (level 7 evaluation).
JF 2.02 would have doubled (level 7 evaluation).
White Blue
Level 7 Evaluation Equity Win G/BG BG Win G/BG BG
JellyFish 3.0 +0.477 61.0% 29.8% 2.9% 39.0% 6.6% 0.3%
JellyFish 2.02 +0.499 62.6% 29.9% 2.3% 37.4% 7.2% 0.3%
Level 5 Rollouts Equity Win G/BG BG Win G/BG BG
Cube centered +0.713 84.7% 1.9% 0.4% 15.3% 0.4% 0.0%
Blue owns cube +0.375 52.2% 29.1% 4.7% 47.8% 0.7% 0.0%
Cubeless +0.512 61.2% 30.9% 4.8% 38.8% 6.5% 0.3%
Level 6 Rollouts Equity Win G/BG BG Win G/BG BG
Cubeless +0.434 59.3% 28.2% 3.8% 40.7% 6.9% 0.4%
|
|
RK:
|
Making the 2 point was not what White had in mind, but he makes what he can.
White Blue
Candidate Plays Equity Win G/BG BG Win G/BG BG
20/14 13/7 8/2(2)* +0.460 59.3% 31.4% 2.9% 40.7% 6.8% 0.2%
13/7(2) 8/2(2)* +0.398 57.0% 30.4% 3.1% 43.0% 7.5% 0.3%
20/2* 8/2 +0.379 57.3% 28.6% 2.5% 42.7% 7.5% 0.2%
20/2* 13/7 +0.361 57.8% 26.1% 1.9% 42.2% 7.3% 0.2%
|
|
RK:
|
Hitting seemed clear here, and I was surprised that anchoring came out as
close as it did. Sending another checker back is a nice gain, especially
while making the 9 point as well, and preventing White from covering the
3 point. There should be plenty of opportunity to make the 3-, 4- or
5-anchor later on.
On the other hand, some arguments in favor of anchoring:
- Even after hitting, White is unlikely to be able to achieve a quick
turnaround. Without an anchor, he will be reluctant to attack
aggressively.
- White does have the better board. Despite being ahead in the race,
White is a bit awkward after the 66, so he'll be looking for
opportunities to attack to help him bring his position home. Anchoring
now prevents this, and guards against immediate jokers such as 22, 33,
44.
- White may not be able to safety the blots on 14 or 3, so Blue could
anchor and then hit later.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
Bar/22* 13/9 -0.336 43.1% 8.0% 0.3% 56.9% 26.1% 2.0%
Bar/21 24/21 -0.349 44.0% 6.8% 0.2% 56.0% 28.0% 1.9%
Bar/21 13/10 -0.389 43.0% 6.6% 0.2% 57.0% 29.4% 2.3%
Bar/22* 24/20 -0.393 41.8% 7.4% 0.2% 58.2% 28.0% 2.6%
|
|
RK:
|
Since White is way ahead in the race, one approach would be to minimize
contact with Bar/23 14/11. The problem is that it leaves an inflexible
position with two checkers on the 23 point, and relatively few chances to
improve the front position.
White's position is a bit awkward after the 66, so he's right to take
advantage of his better board to attack, attempting to make additional
points to go with the 2 point. White can withstand a few hits at this
stage and still be OK.
White Blue
Candidate Plays Equity Win G/BG BG Win G/BG BG
Bar/22 6/4* +0.320 56.9% 25.0% 2.1% 43.1% 8.6% 0.3%
Bar/20 +0.274 55.3% 23.7% 1.6% 44.7% 8.2% 0.2%
Bar/23 6/3* +0.267 55.3% 23.8% 1.7% 44.7% 9.0% 0.3%
Bar/22 23/21 +0.262 54.8% 23.9% 1.7% 45.2% 8.7% 0.3%
Bar/23 14/11 +0.218 53.2% 23.1% 1.5% 46.8% 9.0% 0.3%
|
|
RK:
|
Blue hits and finally gets his second anchor.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
Bar/21* 22/21 -0.211 45.7% 9.0% 0.3% 54.3% 20.6% 1.2%
Bar/21*/20 -0.248 44.8% 8.9% 0.3% 55.2% 21.9% 1.5%
|
|
RK:
|
With White on the bar, board-building is the top priority. Advancing the
anchor would be too conservative.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
9/5 6/5 -0.112 48.9% 10.9% 0.4% 51.1% 19.1% 1.3%
24/20 21/20 -0.191 45.7% 8.3% 0.2% 54.3% 18.2% 0.8%
24/20 6/5 -0.218 46.5% 8.5% 0.3% 53.5% 22.2% 1.5%
|
|
RK:
|
Coming in and anchoring, then swinging the other blot around, seems like
a natural play. However, it has some drawbacks. It leaves Blue 6s
(plus 51 & 54) to send a fourth checker back behind his growing blockade,
and 4s to cover the 9 point. If Blue misses the shot, White still needs
to safety it, while also needing to get the back checkers moving, which
may not be easy.
By coming in on the 21 point instead and playing 14/11, White leaves only
six numbers to get hit on his side of the board. In addition, he gains more
flexibility in Blue's home board. Blue may try to attack, but
probably won't have enough ammunition to do serious damage.
White Blue
Candidate Plays Equity Win G/BG BG Win G/BG BG
Bar/21 14/11 +0.073 50.0% 18.2% 1.2% 50.0% 11.7% 0.5%
Bar/18 +0.055 49.6% 17.6% 1.1% 50.4% 12.0% 0.5%
Bar/22 14/10 +0.007 47.6% 16.8% 1.0% 52.4% 12.0% 0.4%
Bar/21 6/3 -0.095 45.4% 14.8% 1.0% 54.6% 15.1% 0.9%
|
|
RK:
|
A great number, hitting and making a five-prime.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
21/15*/9 13/7(2) +0.560 71.7% 20.2% 0.8% 28.3% 8.2% 0.4%
21/15*(2) 13/7(2) +0.339 63.9% 17.6% 0.8% 36.1% 11.7% 0.7%
21/15*/9(2) +0.278 61.4% 16.0% 0.6% 38.6% 11.2% 0.5%
21/15* 21/9 13/7 +0.177 58.5% 15.2% 0.6% 41.5% 14.2% 0.9%
21/15* 13/7(3) +0.147 57.4% 15.6% 0.7% 42.6% 15.2% 1.1%
21/15*/9 8/2*(2) +0.141 55.0% 16.4% 0.7% 45.0% 12.6% 0.6%
21/15*(2) 8/2*(2) +0.096 53.7% 15.9% 0.7% 46.3% 13.7% 0.6%
|
|
RK:
|
Should Blue consider doubling here? After all, he has just made a
five-prime and has four checkers trapped, and Blue has enough flexibility that
there shouldn't be much trouble keeping this prime for a while. On the
other hand, White still has the potential of containing some of Blue's
back checkers, and Blue is unlikely to ever make the sixth point in the
prime. His main threat is for White's position to gradually
self-destruct. Therefore, unless White will pass, there's not much
reason for Blue to double yet.
Blue White
Level 7 Evaluation Equity Win G/BG BG Win G/BG BG
Cubeless +0.277 61.7% 14.5% 0.4% 38.3% 10.1% 0.4%
Level 5 Rollouts Equity Win G/BG BG Win G/BG BG
Cube centered +0.634 81.9% 0.3% 0.0% 18.1% 0.7% 0.1%
White owns cube +0.459 63.1% 20.5% 0.7% 36.9% 1.5% 0.1%
Cubeless +0.548 70.9% 21.5% 0.8% 29.1% 8.7% 0.7%
Level 6 Rollouts Equity Win G/BG BG Win G/BG BG
Cubeless +0.554 71.5% 21.1% 0.9% 28.5% 9.2% 0.5%
|
|
RK:
|
Re-establishing a good anchor.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
24/21 +0.472 69.2% 17.0% 0.6% 30.8% 8.6% 0.3%
24/22 21/20 +0.374 66.1% 16.4% 0.6% 33.9% 11.1% 0.6%
|
|
RK:
|
13/7 13/8 is a natural-looking play. The problem is that it concedes the
outfield and leaves four checkers behind a five-prime, a difficult position to
play from. JellyFish prefers the alternative of freeing one back checker.
This improves White's timing and maintains the midpoint. The remaining
blot is safe for the moment, since Blue has few numbers to attack with,
and White can remake the forward anchor with 1s. These gains outweigh
the risk of having the runner re-hit.
White Blue
Candidate Plays Equity Win G/BG BG Win G/BG BG
21/10 -0.504 29.7% 7.8% 0.2% 70.3% 17.3% 0.6%
21/15 13/8 -0.558 28.1% 7.4% 0.2% 71.9% 18.9% 0.7%
13/7 13/8 -0.597 25.9% 6.4% 0.2% 74.1% 17.5% 0.5%
|
|
RK:
|
Blue certainly has a big positional advantage, with a five-prime with four
checkers behind it. He has the spare on the midpoint to play with, and
may be able to escape one of the checkers on the 24 point to give himself
more time. Even if he has to break the 21 point first, it will be
difficult for White to form a winning counter-prime, with two checkers
already deep on the 2 point. A double certainly seems called for here.
After all, is it clear that White will (or should) take?
|
|
JF:
|
I would not have doubled (level 7 evaluation).
Blue White
Level 7 Evaluation Equity Win G/BG BG Win G/BG BG
Cubeless +0.488 69.8% 15.8% 0.4% 30.2% 6.7% 0.2%
Level 5 Rollouts Equity Win G/BG BG Win G/BG BG
Cube centered +0.884 94.3% 0.2% 0.0% 5.7% 0.3% 0.0%
White owns cube +0.686 74.2% 20.9% 0.7% 25.8% 1.3% 0.1%
Cubeless +0.744 78.7% 21.3% 0.7% 21.3% 4.9% 0.2%
Level 6 Rollouts Equity Win G/BG BG Win G/BG BG
Cubeless +0.776 79.6% 22.9% 0.6% 20.4% 4.9% 0.1%
|
|
RK:
|
An optimistic take. White will have to roll quite well to fill in his
5
and 3 points while not exposing more blots, as well as escaping some
back
checkers while maintaining an anchor. Blue's timing is not bad:
there's
the checker on the midpoint to play with, plus the chance of escaping
from
the 24 point. Worst case, he may have to break from the 21 point
first,
but then White still has to win the timing battle.
The rollouts indicate a double and a clear pass. Why does JellyFish Level 7
say
"no double"? My guess is it's overvaluing White's two anchors. JellyFish has
learned that in most situations with 4 checkers back, having two anchors is
better than having one, but in this case White's timing is such that
he's
desperate to break one of them, preferably the 22 point, and he'll have
to incur some risks doing so. Level 7, doing only a 2-ply lookahead,
doesn't "see" far enough in advance to predict that White's position is
likely to collapse.
|
|
JF:
|
I would have accepted (level 7 evaluation).
White Blue
Level 7 Evaluation Equity Win G/BG BG Win G/BG BG
Cubeless -0.488 30.2% 6.7% 0.2% 69.8% 15.8% 0.4%
Level 5 Rollouts Equity Win G/BG BG Win G/BG BG
Cube centered -0.884 5.7% 0.3% 0.0% 94.3% 0.2% 0.0%
White owns cube -0.686 25.8% 1.3% 0.1% 74.2% 20.9% 0.7%
Cubeless -0.744 21.3% 4.9% 0.2% 78.7% 21.3% 0.7%
Level 6 Rollouts Equity Win G/BG BG Win G/BG BG
Cubeless -0.776 20.4% 4.9% 0.1% 79.6% 22.9% 0.6%
|
|
RK:
|
A good number. Blue's main asset is his five-prime and he wants to do
everything he can to maintain it. He also prefers to keep the
21 point,
since it interferes most with White's offense, and is willing to give
up
the 24 point to do so. This also improves his timing by giving him an
extra free checker to play with. A "safe" play like 13/5 should not be
considered since it presents real danger that Blue could be forced to
break his five-prime.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
24/16 +0.687 77.4% 18.5% 0.5% 22.6% 5.1% 0.1%
21/13 +0.571 73.3% 17.4% 0.5% 26.7% 7.2% 0.2%
13/8 24/21 +0.569 72.7% 17.7% 0.5% 27.3% 6.6% 0.1%
|
|
RK:
|
Pointing on the 1 point is tempting while Blue has two outfield blots.
If Blue fans and White escapes and hits another blot, he could turn the
game around. However, that would be quite a parlay, and the play
increases the chances that White will remain stuck behind the prime and
be forced to crash his board. The rollouts find that the purer play of
21/10 is best. Despite its apparent looseness, this play keeps all
checkers in play and reduces White's chances of crashing. If hit,
White may be able to remake the 21 point and just continue from there.
White Blue
Candidate Plays Equity Win G/BG BG Win G/BG BG
21/10 -0.675 22.9% 4.9% 0.1% 77.1% 17.6% 0.5%
21/15 6/1* -0.721 21.9% 4.9% 0.1% 78.1% 20.3% 0.7%
21/15 8/3 -0.765 20.8% 4.5% 0.1% 79.2% 21.8% 0.8%
7/1* 6/1 -0.774 19.3% 4.5% 0.1% 80.7% 20.1% 0.5%
21/15 7/2 -0.781 19.5% 4.4% 0.1% 80.5% 20.9% 0.6%
7/1* 8/3 -0.826 18.7% 3.8% 0.0% 81.3% 23.0% 0.7%
8/2 7/2 -0.853 16.9% 3.2% 0.0% 83.1% 21.8% 0.6%
7/1* 7/2 -0.873 16.7% 3.2% 0.0% 83.3% 23.3% 0.7%
|
|
RK:
|
After entering, Blue does best to take further outfield control with 24/14.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
Bar/14 +0.808 81.9% 20.1% 0.5% 18.1% 3.5% 0.0%
Bar/20 13/7 +0.769 80.0% 20.9% 0.5% 20.0% 4.5% 0.1%
Bar/20 16/10 +0.744 79.7% 20.1% 0.5% 20.3% 5.5% 0.1%
Bar/20 21/15 +0.686 78.8% 18.9% 0.5% 21.2% 8.1% 0.2%
|
|
RK:
|
White can't escape, so he trades an outer point for an inner one.
White Blue
Candidate Plays Equity Win G/BG BG Win G/BG BG
8/3 7/3 -0.876 16.0% 2.9% 0.0% 84.0% 22.0% 0.6%
7/2 7/3 -0.882 15.9% 2.5% 0.0% 84.1% 21.9% 0.5%
8/3 6/2 -0.928 14.5% 2.1% 0.0% 85.5% 23.3% 0.6%
|
|
RK:
|
Blue creates some builders to attack with, in case White is able to run.
There are a bunch of plays that do this and they're all fairly close.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
14/6 13/5 +0.841 82.5% 21.5% 0.6% 17.5% 3.0% 0.0%
16/8 14/6 +0.836 82.6% 20.8% 0.5% 17.4% 2.9% 0.0%
16/8 13/5 +0.830 82.3% 20.9% 0.6% 17.7% 3.1% 0.0%
16/12 14/6 13/9 +0.825 82.3% 20.3% 0.5% 17.7% 3.0% 0.0%
16/8 14/10 +0.809 81.9% 20.2% 0.5% 18.1% 3.4% 0.0%
14/2 13/9 +0.804 81.2% 21.3% 0.6% 18.8% 3.8% 0.0%
16/12 14/10 13/5 +0.802 81.8% 19.9% 0.5% 18.2% 3.8% 0.0%
16/12 14/2 +0.790 80.9% 20.8% 0.5% 20.1% 4.2% 0.1%
|
|
RK:
|
This seems like a good time to get an extra checker on the launching pad
with 22/21, as long as White can safety the other blot. White has no
time to maintain two anchors, so he'd clearly prefer keeping the 21-point
if he runs with a 6. He's not too worried about being attacked, since he
has an anchor and a better board.
White Blue
Candidate Plays Equity Win G/BG BG Win G/BG BG
7/3 22/21 -0.789 19.0% 2.7% 0.0% 81.0% 19.2% 0.5%
7/2 -0.835 17.3% 2.6% 0.0% 82.7% 20.2% 0.5%
7/3 2/1 -0.877 15.9% 2.1% 0.0% 84.1% 21.0% 0.5%
8/3 -0.921 14.8% 1.9% 0.0% 85.2% 23.0% 0.6%
7/3 6/5 -0.926 15.3% 2.1% 0.0% 84.7% 24.5% 0.7%
6/2 6/5 -0.937 14.8% 1.9% 0.0% 85.2% 24.5% 0.7%
6/2 8/7 -0.946 14.1% 1.7% 0.0% 85.9% 23.9% 0.7%
|
|
RK:
|
Should Blue hit loose on the 3 point? Playing 14/9 has the disadvantage
of allowing White to escape with any 6. On the other hand, it could hurt
to get hit back while White has a better board. The checker on the
3 point is primed, and by not hitting, most non-6s cause White to break
the 8 point and possibly leave a shot.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
14/9 +0.796 81.1% 19.5% 0.5% 19.9% 2.5% 0.0%
8/3* +0.782 80.5% 20.3% 0.5% 19.5% 3.5% 0.0%
14/10 8/7 +0.780 80.5% 19.8% 0.5% 19.5% 3.3% 0.0%
5/1 8/7 +0.773 79.8% 20.7% 0.5% 20.2% 3.4% 0.0%
5/1 14/13 +0.769 79.7% 20.6% 0.5% 20.3% 3.4% 0.0%
21/16 +0.764 80.9% 18.1% 0.4% 19.1% 3.9% 0.0%
|
|
RK:
|
Again, Blue has no need to hit. 21/16 gives him a shot at any checker
that escapes, while guarding the inner board in case White has to break
the 6 point. Because of White's better board, the more conservative
play of staying on the 17 point (getting out of the way of 66) is
equally good, while improving the builder 8/7 or duplicating 1s with
21/20.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
21/17 8/7 +0.839 83.4% 18.5% 0.4% 16.6% 1.7% 0.0%
21/17 21/20 +0.834 83.4% 17.9% 0.4% 16.6% 1.6% 0.0%
21/16 +0.832 83.1% 18.5% 0.4% 16.9% 2.0% 0.0%
21/17 9/8 +0.831 83.1% 18.4% 0.4% 16.9% 1.8% 0.0%
9/5 21/20 +0.829 82.7% 19.0% 0.4% 17.3% 1.9% 0.0%
9/5 8/7 +0.814 81.8% 19.7% 0.5% 18.2% 2.2% 0.0%
5/1 21/20 +0.808 81.7% 19.4% 0.4% 18.3% 2.3% 0.0%
5/1 8/7 +0.804 81.1% 20.3% 0.4% 18.9% 2.4% 0.0%
8/3* +0.799 81.4% 19.3% 0.4% 18.6% 2.6% 0.0%
5/1 9/8 +0.788 80.5% 20.2% 0.4% 19.5% 2.7% 0.0%
|
|
RK:
|
Moving to the edge of the prime with the ace is clear, giving White an
additional 6 with which to escape.
White Blue
Candidate Plays Equity Win G/BG BG Win G/BG BG
6/2 22/21 -0.805 17.3% 1.2% 0.0% 82.7% 16.0% 0.3%
6/1 -0.890 14.7% 1.2% 0.0% 85.3% 19.2% 0.4%
6/2 3/2 -0.892 14.6% 1.2% 0.0% 85.4% 19.2% 0.4%
|
|
RK:
|
Blue wants to stay back with both rear checkers, hoping that White is
forced to break his 6 point and expose a checker, or hoping to hit him if
he escapes with a 6. There are numerous plays that do that, and they're
all close, but it makes sense to simply slot the next point he wants
(the 3 point).
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
9/6 5/3 +0.812 82.5% 17.3% 0.3% 17.5% 1.4% 0.0%
5/2 9/7 +0.809 82.3% 17.6% 0.3% 17.7% 1.7% 0.0%
8/3 +0.808 82.4% 17.1% 0.3% 17.6% 1.5% 0.0%
5/2 8/6 +0.807 82.4% 17.0% 0.3% 17.6% 1.5% 0.0%
9/6 8/6 +0.799 82.2% 16.4% 0.3% 17.8% 1.2% 0.0%
9/6 16/14 +0.791 82.2% 15.8% 0.3% 17.8% 1.3% 0.0%
8/5 9/7 +0.791 82.0% 16.1% 0.3% 18.0% 1.3% 0.0%
5/2 16/14 +0.791 82.0% 16.2% 0.3% 18.0% 1.5% 0.0%
16/13 5/3 +0.790 82.0% 16.0% 0.3% 18.0% 1.4% 0.0%
16/13 8/6 +0.783 81.9% 15.6% 0.3% 18.1% 1.2% 0.0%
16/13 9/7 +0.782 81.8% 15.6% 0.3% 18.2% 1.3% 0.0%
8/5 16/14 +0.780 81.8% 15.5% 0.3% 18.2% 1.3% 0.0%
16/11 +0.773 81.6% 15.1% 0.3% 18.4% 1.2% 0.0%
21/18 5/3 +0.766 81.8% 13.9% 0.2% 18.2% 1.2% 0.0%
21/18 9/7 +0.759 81.6% 13.5% 0.2% 18.4% 1.1% 0.0%
21/18 8/6 +0.755 81.5% 13.4% 0.2% 18.5% 1.0% 0.0%
|
|
RK:
|
Blue can cover the 3 point and move the outfield checker out of the way of
a shot -- although the more aggressive play of staying on 16 would be OK
too.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
6/3 16/14 +0.632 76.3% 12.5% 0.2% 23.7% 2.0% 0.0%
8/3 +0.623 75.6% 13.2% 0.2% 24.4% 2.4% 0.0%
8/5 16/14 +0.616 75.6% 12.8% 0.2% 24.4% 2.5% 0.0%
16/11 +0.616 75.6% 12.5% 0.2% 24.4% 2.4% 0.0%
16/13 8/6 +0.604 75.0% 12.7% 0.2% 25.0% 2.6% 0.0%
16/13 3/1 +0.581 74.0% 12.7% 0.2% 26.0% 2.8% 0.0%
21/16 +0.579 74.6% 9.9% 0.1% 25.4% 1.3% 0.0%
8/5 3/1 +0.575 73.6% 13.5% 0.2% 26.4% 3.4% 0.0%
21/18 16/14 +0.542 73.0% 10.9% 0.2% 27.0% 2.9% 0.0%
21/18 8/6 +0.531 72.6% 11.7% 0.2% 27.4% 4.0% 0.0%
|
|
RK:
|
White has to leave a shot, so he might as well hit,
right? Maybe not. What does White gain by hitting? It's very unlikely
that White can keep Blue on the bar while he escapes everybody and wins.
White's main concern is really to not get any more checkers sent back
and maintain some racing equity. It's 11 shots either way, but by
hitting, White actually is more likely to get hit back. If Blue misses,
he may remain on the bar to get more chances.
White Blue
Candidate Plays Equity Win G/BG BG Win G/BG BG
10/5 6/5 -0.694 21.1% 1.3% 0.0% 78.9% 12.5% 0.2%
10/4* -0.701 21.3% 2.3% 0.0% 78.7% 14.7% 0.3%
10/5 2/1 -0.730 20.2% 1.2% 0.0% 79.8% 14.3% 0.3%
10/5 3/2 -0.730 20.2% 1.2% 0.0% 79.8% 14.3% 0.2%
|
|
RK:
|
After hitting, either 3 that doesn't move the back checker is OK.
Blue wants to pick up the other blot if possible.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
Bar/21* 8/5 83.8% 18.1% 0.5% +0.850 16.2% 1.2% 0.0%
Bar/21* 14/11 83.8% 17.9% 0.4% +0.848 16.2% 1.2% 0.0%
Bar/21* 21/18 81.5% 14.3% 0.3% +0.764 18.5% 1.1% 0.0%
|
|
RK:
|
Blue correctly stays back, still hoping to hit another blot. The pure
play of 11/2, despite the shot, appears at least as good as 8/3 11/7,
since it makes it easier to make the board in order. Again, getting hit
is not a big concern here, since it actually increases the chances of
picking up another checker.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
11/2 +0.830 82.2% 19.9% 0.6% 17.8% 1.8% 0.0%
8/3 11/7 +0.824 82.1% 19.0% 0.6% 17.9% 1.4% 0.0%
21/16 11/7 +0.811 82.3% 17.1% 0.5% 17.7% 1.1% 0.0%
11/6 21/17 +0.801 82.1% 16.3% 0.5% 17.9% 0.9% 0.0%
8/3 21/17 +0.762 80.9% 15.1% 0.5% 19.1% 1.3% 0.0%
21/12 +0.732 80.5% 13.0% 0.3% 19.5% 1.2% 0.0%
|
|
RK:
|
White is forced off the anchor, so it's logical to minimize the damage by
playing 5/1.
White Blue
Candidate Plays Equity Win G/BG BG Win G/BG BG
21/15 5/1 -0.963 16.0% 2.6% 0.0% 84.0% 29.9% 1.0%
21/11 -1.103 13.4% 2.3% 0.0% 86.6% 37.0% 2.5%
|
|
RK:
|
An awkward roll with a few possible choices. Attacking seems right
because of White's three blots. Blue could simply make the 2 point, but
that would give White a lot of freedom while depriving himself of useful
builders. After hitting, simply lifting the blot is best despite the
apparent awkwardness. White's better board deserves some respect, and
Blue still has good chances to continue the attack after the safer play.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
9/4*/3 +0.860 82.4% 22.7% 0.7% 17.6% 2.1% 0.0%
7/2 3/2 +0.844 81.6% 22.1% 0.7% 18.4% 1.7% 0.0%
9/4* 9/8 +0.799 79.6% 23.3% 0.8% 20.4% 3.5% 0.0%
9/4* 7/6 +0.782 78.7% 25.4% 0.9% 21.3% 5.3% 0.1%
9/4* 21/20 +0.760 78.4% 24.1% 0.8% 21.6% 5.5% 0.1%
|
|
RK:
|
An interesting choice. Blue would like to stay back to prevent the
loose checker from getting home easily, but then the 6 is a bit awkward
(9/3 or 7/1). By playing 21/15 he makes it easy for White to hit or run
by, but at least keeps all his checkers in play. The rollouts show a tiny
preference for staying back despite the awkwardness.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
7/1 9/8 +0.710 80.8% 10.2% 0.2% 19.2% 1.0% 0.0%
9/3 7/6 +0.703 80.9% 9.2% 0.2% 19.1% 0.8% 0.0%
21/15 9/8 +0.699 81.0% 8.5% 0.2% 19.0% 0.8% 0.0%
9/3 21/20 +0.694 80.5% 9.0% 0.2% 19.5% 0.9% 0.0%
7/1 21/20 +0.692 80.2% 9.9% 0.2% 19.8% 1.4% 0.0%
21/15 7/6 +0.689 80.5% 9.1% 0.2% 19.5% 1.3% 0.0%
21/14 +0.643 78.4% 9.2% 0.2% 21.6% 1.9% 0.0%
|
|
RK:
|
One of the few rolls that doesn't hit or get by.
White Blue
Candidate Plays Equity Win G/BG BG Win G/BG BG
15/12 2/1 -0.767 15.7% 0.4% 0.0% 84.3% 8.4% 0.1%
15/12 3/2 -0.769 15.6% 0.4% 0.0% 84.4% 8.3% 0.1%
15/11 -0.769 15.5% 0.4% 0.0% 84.5% 8.2% 0.1%
15/12 6/5 -0.801 14.5% 0.5% 0.0% 85.5% 9.4% 0.2%
6/3 15/14 -0.848 12.9% 0.4% 0.0% 87.1% 10.8% 0.2%
6/2 -0.851 12.9% 0.4% 0.0% 87.1% 11.1% 0.2%
6/3 2/1 -0.856 12.9% 0.4% 0.0% 87.1% 11.7% 0.2%
6/3 6/5 -0.858 12.8% 0.4% 0.0% 87.2% 11.7% 0.2%
|
|
RK:
|
After hitting, 8/5 is a tad better than 14/11 because of the extra builder.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
8/5 15/14* +0.925 89.3% 13.9% 0.3% 10.7% 0.4% 0.0%
15/14*/11 +0.914 89.2% 13.1% 0.3% 10.8% 0.4% 0.0%
7/4 15/14* +0.877 86.1% 16.4% 0.4% 13.9% 1.3% 0.0%
|
|
RK:
|
Is it right to hit loose in positions like this? It doesn't hurt to get
hit back since Blue has a prime. On the other hand, since White owns
the 23 point, it doesn't gain much unless he fans for several rolls,
since his chances of rolling a 2 are just as great as rolling a 1 if he's
on the ace point. Blue's best way to increase his gammon chances is to
try to close out one or more blots. He has a fairly effective way to try
for that by simply breaking the 8 point. This forces White to break
anchor if he rolls a 6 before a 1. Blue will then be in good position
to attack the remaining blots and possibly close the board.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
8/6 8/7 +1.128 94.0% 24.5% 0.5% 6.0% 0.2% 0.0%
3/1* 11/10 +1.087 92.8% 22.9% 0.5% 7.2% 0.3% 0.0%
8/6 11/10 +1.087 94.1% 20.2% 0.4% 5.9% 0.1% 0.0%
8/5 +1.079 92.4% 22.9% 0.5% 7.6% 0.4% 0.0%
11/9 8/7 +1.067 92.3% 22.0% 0.5% 7.7% 0.4% 0.0%
7/5 7/6 +1.063 92.9% 20.3% 0.5% 7.1% 0.2% 0.0%
3/1* 8/7 +1.055 92.5% 20.6% 0.5% 7.5% 0.5% 0.0%
3/1* 7/6 +1.050 91.6% 22.1% 0.5% 8.4% 0.9% 0.0%
11/8 +1.049 93.1% 18.5% 0.4% 6.9% 0.2% 0.0%
|
|
RK:
|
It seems straightforward to clear from the rear while continuing to block
White from escaping with 5s. However, Blue has a problem with 5s too:
they don't allow him to clear the bar point, and 65 or 54 force a shot.
Clearing both the 8 and 7 points now lets Blue start bearing off sooner
and increases his gammon chances more than trying to block 5s.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
8/5(2) 7/4(2) +1.060 95.4% 15.1% 0.2% 4.6% 0.0% 0.0%
10/4 8/5(2) +1.039 95.0% 13.9% 0.2% 5.0% 0.1% 0.0%
7/1(2) +1.026 93.7% 15.1% 0.2% 6.3% 0.1% 0.0%
10/4 7/4(2) +1.023 94.0% 14.1% 0.2% 6.0% 0.1% 0.0%
10/1 7/4 +0.986 91.7% 15.1% 0.2% 8.3% 0.1% 0.0%
10/1 8/5 +0.975 91.3% 14.7% 0.2% 8.7% 0.1% 0.0%
10/4 6/3(2) +0.934 91.1% 11.2% 0.1% 8.9% 0.2% 0.0%
|
|
RK:
|
2/1 slightly reduces the chances of crashing further by preserving the
maximum number of 2s.
White Blue
Candidate Plays Equity Win G/BG BG Win G/BG BG
2/1 0 -1.085 4.1% 0.0% 0.0% 95.9% 16.4% 0.2%
3/2 0 -1.092 4.1% 0.0% 0.0% 95.9% 17.1% 0.2%
|
|
RK:
|
Since White's board has already crashed, Blue would have little to gain by
maintaining his five-prime. Bringing two in allows him to start taking off
checkers in hopes of a gammon.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
7/3 7/6 +1.124 97.5% 17.1% 0.2% 2.5% 0.0% 0.0%
5/1 7/6 +1.045 93.9% 16.6% 0.2% 6.1% 0.0% 0.0%
5/1/5/4 +1.038 94.5% 14.7% 0.1% 5.5% 0.1% 0.0%
5/1 4/3 +1.037 94.6% 14.3% 0.1% 5.4% 0.1% 0.0%
7/3 5/4 +1.022 93.1% 16.0% 0.2% 6.9% 0.1% 0.0%
|
|
|
White Blue
Candidate Plays Equity Win G/BG BG Win G/BG BG
3/2 0 -1.213 2.1% 0.0% 0.0% 97.9% 25.0% 0.4%
2/1 0 -1.214 2.0% 0.0% 0.0% 98.0% 25.0% 0.4%
|
|
RK:
|
Blue increases his gammon chances by yanking two off. In some
situations, strippage on an inner point can be dangerous, but here
stripping the 4 point isn't a problem since Blue has convenient 4s and
3s to play elsewhere.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
6/Off 4/Off +1.230 97.9% 26.7% 0.5% 2.1% 0.0% 0.0%
6/Off 5/1 +1.204 97.8% 24.5% 0.4% 2.2% 0.0% 0.0%
|
|
RK:
|
White continues to make accurate technical plays. Why is 3/2 better now?
At this point it's unlikely to make a difference in terms of crashing,
so White wants to have the best possible distribution for bearing off.
In the (unlikely) event that White hits and gets into racing contention,
not being stripped on the 2 point could gain a roll.
White Blue
Candidate Plays Equity Win G/BG BG Win G/BG BG
3/2 0 -1.335 2.0% 0.0% 0.0% 98.0% 36.7% 0.8%
2/1 0 -1.343 2.0% 0.0% 0.0% 98.0% 37.5% 0.9%
|
|
RK:
|
Two off is clearly best to increase gammon chances. Even if it results
in a tad more shots than clearing the 6 point, getting hit is hardly
fatal and the extra checkers off help if hit too.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
5/Off 3/Off +1.369 98.1% 39.7% 1.1% 1.9% 0.0% 0.0%
6/1 6/3 +1.284 98.3% 31.4% 0.5% 1.7% 0.0% 0.0%
|
|
|
White Blue
Candidate Plays Equity Win G/BG BG Win G/BG BG
23/15 -1.316 2.1% 0.0% 0.0% 97.9% 34.9% 0.8%
23/17 3/1 -1.346 2.0% 0.0% 0.0% 98.0% 37.6% 0.9%
|
|
RK:
|
Two off is still clear, despite the risk of a double shot with 64.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
5/Off 3/Off +1.380 98.3% 40.2% 1.1% 1.7% 0.0% 0.0%
6/1 6/3 +1.289 98.4% 31.6% 0.5% 1.6% 0.0% 0.0%
|
|
RK:
|
White plays efficiently to the 6 point. But this could cost if he rolls
a number that can't escape next time. Running one checker avoids this,
and is fairly safe since Blue has no spares, although some small doubles
could hurt.
White Blue
Candidate Plays Equity Win G/BG BG Win G/BG BG
23/18 15/11 -1.348 2.5% 0.0% 0.0% 97.5% 38.9% 0.9%
23/14 -1.351 2.4% 0.0% 0.0% 97.6% 39.1% 0.9%
15/6 -1.359 1.7% 0.0% 0.0% 98.3% 37.8% 1.4%
|
|
|
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
6/Off(2) +1.527 98.8% 52.7% 2.4% 1.2% 0.0% 0.0%
4/1(2) 3/Off(2) +1.411 98.1% 43.7% 1.3% 1.9% 0.0% 0.0%
|
|
|
White Blue
Candidate Plays Equity Win G/BG BG Win G/BG BG
23/12 -1.408 1.5% 0.0% 0.0% 98.5% 42.6% 1.2%
23/17 6/1 -1.550 1.2% 0.0% 0.0% 98.8% 55.1% 2.4%
|
|
RK:
|
Leaving a voluntary shot when bearing off is rarely correct, so Blue
makes the play that would be right in most situations. But here, Blue's
main goal is to win a gammon. His chances of losing are quite small even
if he gets hit, since White only has a 3 point board, with a couple of
blots around, and Blue already has 8 off. (If Blue were to leave a
shot and White hit it
and
covered the 6 point, White's winning chances
would still be around only 9%).
What are the gammon chances? White needs a minimum of 23 pips to get both his
outside checkers home, and then needs to bear one off. Considering wastage,
this takes something over 3 rolls on average. After 4/1 4/3, Blue figures to
be off in 4 if he misses only once. Since White is on roll, he is a definite
underdog to be gammoned (24.4%, according to JellyFish).
So what's the best play? Continue with the theme of yanking two checkers
off? 3/off 1/off would be reasonable, since it would potentially gain a
roll for the gammon and would further reduce Blue's losing chances if
hit. Plus White may not be able to stay around for any future shots.
However, the best play may be to put White on the bar with 5/2*/1!
(JellyFish loves these kinds of plays--and, unlike us, never overlooks
one.) This gains significantly when White dances (25% of the time).
Even if White hits, Blue may be able to hit back somewhere and continue
the gammon quest. This play increases the gammon chances to 32.6% while
giving White only 2.9% wins.
Blue White
Candidate Plays Equity Win G/BG BG Win G/BG BG
5/2*/1 +1.276 97.1% 32.6% 0.8% 2.9% 0.0% 0.0%
3/Off 1/Off +1.268 97.9% 30.3% 0.7% 2.1% 0.0% 0.0%
4/1 4/3 +1.238 99.6% 24.4% 0.2% 0.4% 0.0% 0.0%
|
|
RK:
|
Play efficiently to avoid the gammon.
White Blue
Candidate Plays Equity Win G/BG BG Win G/BG BG
12/6 23/19 -1.156 0.0% 0.0% 0.0% 100.0% 15.6% 0.0%
23/13 -1.163 0.0% 0.0% 0.0% 100.0% 16.4% 0.0%
23/17 12/8 -1.173 0.0% 0.0% 0.0% 100.0% 17.4% 0.0%
|
|
RK:
|
Not enough! 53, 52 or 51 would have been better.
Blue went on to win the match, 11-2.
|
The game was recorded on tape and transcribed by Richard McIntosh.
Rollouts were made by Richard McIntosh,
using JellyFish Analyzer 3.0.
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
- 7776 games (36x216)
- horizon 7
- seed 615
Standard deviations of equity estimates were between 0.002 and 0.007,
typically 0.003.
Parameter values for level 5 rollouts on cube decisions were:
- level 5
- 23328 games (36x648)
- full game
- seed 615
- settlement limit 0.550
Parameter values for level 6 rollouts on cube decisions were:
- level 6
- 216 games (36x6)
- full game
- seed 615
Copyright © 1996-2010
BackGammon By the Bay
