need jelly analysis 6/5 opener
Mark W. Lehman
Hi Folks. Was wondering if anyone could give jellyfish analysis
on the 6/5 opener. I know it's automatic to just run it but my
gut wrenches everytime I do. So often I have ended up trapped
with no home board developed or forced to play awkward in the
following rolls, while my opponent has the convenience of choosing
what type of game to play with plenty of time to develop. I know
that it is the *best* move but would like to see the comparison
between running vs. bringing a builder down with the 5 and pointing
on opponents bar point with 6. Thanks in advance.
BTW I am now comfortable with making 2 point with 6/4 rather
than the typical split that was once more comfortable on the
opening roll. :)
Sincerely, Mark Lehman
Daniel Murphy
mle...@omnifest.uwm.edu (Mark W. Lehman) writes:
>Hi Folks. Was wondering if anyone could give jellyfish analysis
>on the 6/5 opener. ... So often I have ended up trapped
>with no home board developed or forced to play awkward in the
>following rolls, while my opponent has the convenience of choosing
>what type of game to play with plenty of time to develop. ...
I suspect that, having "escaped" one checker with 65, you become too
reluctant to allow that checker to be sent back, and make awkward plays to
avoid losing your race lead.
With your rolls immediately following your opening 65, you need to make
natural plays: making points and bringing down builders. It is not a
great loss if your escaped checker gets sent back at this stage of the
game. If it is, you'll have a chance to make an advanced anchor and play a
game you may be more comfortable with. And even if you can't make an
advanced anchor, you'll have compensation for the race loss with the
positional strength you've been able to build on your side of the board.
If you play the rolls immediately following the 65 completely safe,
usually your position will become increasingly awkward and you will
eventually be forced to leave a blot which if hit will lose the game.
Better to leave blots (builders) early, not late, before your opponent
builds a board which prevents you from using the checker sent back
to make an anchor.
/Raccoon
--
_______________________
Daniel Murphy
San Francisco
rac...@cityraccoon.com
Jerry Schonewille
In article <5ac578$n...@omnifest.uwm.edu>, mle...@omnifest.uwm.edu (Mark W. Lehman) wrote:
>Hi Folks. Was wondering if anyone could give jellyfish analysis
>with no home board developed or forced to play awkward in the
>following rolls, while my opponent has the convenience of choosing
>that it is the *best* move but would like to see the comparison
>between running vs. bringing a builder down with the 5 and pointing
>on opponents bar point with 6. Thanks in advance.
>
>BTW I am now comfortable with making 2 point with 6/4 rather
>than the typical split that was once more comfortable on the
>opening roll. :)
>
I agree that running on an opening 6/5 is not pleasant but ...
JellyFish says the following
-0.089 24/13
-0.134 24/18 13/8
-0.156 13/7 6/1*
-0.169 13/7 13/8
-0.176 24/18 6/1*
best regards,
Jerry Schonewille
Stuart Katz, MD
Je...@polymore.com (Jerry Schonewille) wrote:
>In article <5ac578$n...@omnifest.uwm.edu>, mle...@omnifest.uwm.edu (Mark W. Lehman) wrote:
>>Hi Folks. Was wondering if anyone could give jellyfish analysis
>>on the 6/5 opener..... I know
>>that it [24/13] is the *best* move but would like to see the comparison
>>between running vs. bringing a builder down with the 5 and pointing
>>on opponents bar point with 6. Thanks in advance.
>>
To which Jerry responded:
>I agree that running on an opening 6/5 is not pleasant but ...
>JellyFish says the following
>-0.089 24/13
>-0.134 24/18 13/8
>-0.156 13/7 6/1*
>-0.169 13/7 13/8
>-0.176 24/18 6/1*
Something very basic is wrong here. I'm sure it is a misunderstanding
about the position to be analyzed since one can't legally hit on the
1-point on the opening shake. Perhaps this is the equity for a 65
response following an opening 65 played as a routine straight run to
the midpoint.
For an opening 65 at money play I get the following L7 2-ply
estimates:
24/13 +0.067
24/18 13/8 +0.019
13/7 13/8 -0.039
Stuart
Chuck Bower
In article <5aiv15$2...@dfw-ixnews4.ix.netcom.com>,
I agree with Stuart. Furthermore, here is JF level-6 rollouts
(432 trials each) say:
money money
play cubeless cubeless
equity total wins
24/13 +0.077(0.013) +53.9(0.5)%
24/18, 13/8 +0.031(0.012) +51.2(0.5)%
(numbers in parentheses are standard deviations)
Now for some commentary. Back in the 70's it became popular
to belittle the 65 opener played 24/13 (known as "lover's leap").
The macho players preferred slotting plays (for example 21 played
13/11, 6/5). The ultimate derision came in "Backgammon for Blood"
whose author (Bruce Becker) remained incognito, even after being
publically challenged to come forth by Lewis Deyong.
Computer players (EXBG, TDG, JF, Loner) have since unanimously(?)
agreed that not only is 24/13 the correct play, but 65 is by far one
of the best opening rolls!! Here is how JF level-6 rollouts (432
trials each) ranks the opening rolls:
money money
roll play cubeless rank cubeless rank
equity total wins
31 8/5,6/5 +0.162(0.014) 1 +55.0(0.5)% 1
42 8/4,6/4 +0.087(0.014) 2 +52.5(0.5)% 4
61 13/7,8/7 +0.085(0.014) 3 +52.8(0.5)% 3
65 24/13 +0.077(0.013) 4 +53.9(0.5)% 2
53 8/3,6/3 +0.060(0.014) 5 +52.3(0.5)% 5
NOTE THAT 65 RANKS SECOND IN TOTAL WINNING CHANCES! (A word of
caution, though: JF is playing for "money" which means gammons
count twice as much as simple wins, etc. If gammons don't count
--like at double match point--these results will likely be a bit
different.) The reason 65 is worse in total equity is because
it doesn't WIN as many gammons as the building plays. As far
as gammon LOSSES go, it is about the same as the building plays
(except 31, which is best all around).
Chuck
bo...@bigbang.astro.indiana.edu
c_ray on FIBS
P.S. Do you think Bruce Becker reads the newsgroup??
William C. Bitting
Jerry Schonewille (Je...@polymore.com) wrote:
: In article <5ac578$n...@omnifest.uwm.edu>, mle...@omnifest.uwm.edu (Mark W. Lehman) wrote:
: >Hi Folks. Was wondering if anyone could give jellyfish analysis
: >on the 6/5 opener.
: I agree that running on an opening 6/5 is not pleasant but ...
: JellyFish says the following
: -0.089 24/13
: -0.134 24/18 13/8
: -0.156 13/7 6/1*
: -0.169 13/7 13/8
: -0.176 24/18 6/1*
: best regards, : Jerry Schonewille
In the above list moves 3 and 5 are not the opening roll, obviously, but
the reply to the opening roll. It is not clear to me whether the other
rolls are all replies or not. The score seems to suggest that is the
case, that all are replies, not opening rolls. I assume the numbers would be
positive on opening rolls.
While 13/7 6/1* may not be a correct reply, 13/7* 6/1* was played by both
jellyfish and mloner each time it came up in their 300 match (1170 games)
series in the fall of 1995. wcb on FIBS
W1 scr R1 N1 move 1 W2 R2 N2 move 2 |mat
-- ---- -- -- ----------- -- -- -- ----------- ---
JF 1-0 62 sp 24-18 13-11 ML 65 h0 13-7* 6-1* |180
ML 2-0 62 sp 24-18 13-11 JF 65 h0 13-7* 6-1* |297
JF 2-4! 62 sp 24-18 13-11 ML 65 h0 13-7* 6-1* |69
ML 3-1 62 sp 24-18 13-11 JF 65 h0 13-7* 6-1* |30
ML 4-4! 62 sp 24-18 13-11 JF 65 h0 13-7* 6-1* |70
ML 1-0 63 sp 24-18 13-10 JF 65 h0 13-7* 6-1* |49
ML 1-2 63 sp 24-18 13-10 JF 65 h0 13-7* 6-1* |263
! post crawford game; sp split; hO double hit
Andrew Paik
Hi Everyone,
>>> Hi Folks. Was wondering if anyone could give jellyfish analysis
>>> on the 6/5 opener.
>> I agree that running on an opening 6/5 is not pleasant but ...
>> JellyFish says the following
>> -0.089 24/13
>> -0.134 24/18 13/8
>> -0.156 13/7 6/1*
>> -0.169 13/7 13/8
>> -0.176 24/18 6/1*
> In the above list moves 3 and 5 are not the opening roll, obviously, but
> the reply to the opening roll. It is not clear to me whether the other
> rolls are all replies or not. The score seems to suggest that is the
> case, that all are replies, not opening rolls. I assume the numbers would be
> positive on opening rolls.
Yes, I noticed that too. For Jellyfish, on level 7 with verify look-ahead,
I get the following for an OPENING 6-5.
.067 24/13
.019 24/18 13/8
-.039 13/7 13/8
-.085 13/2
-.125 24/18 8/3
Andy
pa...@crl.com
Richard McIntosh
I rolled this out last week for a game we're publishing. With the
following JF parameters:
* level 5
* truncated, horizon 7
* 7776 games
I got the following results:
24/13 +0.069
24/18 13/8 +0.024
Others have already pointed out the strength of 24/13 from the
bot-bot confrontations.
Regards,
Richard
Stuart Katz, MD wrote:
>
[... elision ...]