interesting anecdote
John S Mamoun
I played a backgammon match last year with an expert, and he
played an X-1 opening roll by slotting on the five point,
instead of splitting the back men. I asked him, "Hmmm...are
you sure that's the best move? Jellyfish says its wrong."
He says, "Yes, I know that, and Jellyfish is a great player,
but its not God."
I've spent a year thinking about who is right, JF or the human.
I'm still not quite sure what the answer is...
I suppose I'd tilt towards the bot, but Is JF God or is it not?
Fredrik Dahl
It's not, but its rollout results are pretty Godish, except in very
extreme positions.
The rollout results are almost dead even for 21, but with 41 and 51 it
can't be right to slot.
--
- Fredrik Dahl
Brian Sheppard
John S Mamoun <js...@mail2.sas.upenn.edu> wrote in article
<5v51n9$4il$1...@netnews.upenn.edu>...
> I played a backgammon match last year with an expert, and he
> played an X-1 opening roll by slotting on the five point,
> instead of splitting the back men. I asked him, "Hmmm...are
> you sure that's the best move? Jellyfish says its wrong."
> He says, "Yes, I know that, and Jellyfish is a great player,
> but its not God."
>
> I've spent a year thinking about who is right, JF or the human.
> I'm still not quite sure what the answer is...
>
> I suppose I'd tilt towards the bot, but Is JF God or is it not?
Rollouts are pretty good, if repeated often enough to reduce
statistical noise.
My judgment is that splitting with 4-1 and 5-1 is correct, and
slotting is equal to splitting with 2-1.
But I still slot, because the differences are small enough that I
am willing to indulge in playing positions that I am more comfortable
with.
Brian
Daniel Murphy
On Wed, 10 Sep 1997 10:13:21 -0700, Fredrik Dahl <fre...@sn.no>
wrote:
>John S Mamoun wrote:
>>
>> I played a backgammon match last year with an expert, and he
>> played an X-1 opening roll by slotting on the five point,
>> instead of splitting the back men. I asked him, "Hmmm...are
>> you sure that's the best move? Jellyfish says its wrong."
>> He says, "Yes, I know that, and Jellyfish is a great player,
>> but its not God."
>
>It's not, but its rollout results are pretty Godish, except in very
>extreme positions.
>
>The rollout results are almost dead even for 21, but with 41 and 51 it
>can't be right to slot.
>
'Can't be right' -- assuming two good and evenly matched players.
Slotting is an attempt to avoid the type of game in which both players
make advanced anchors and hope for better rolls. Such mutual holding
games are not always easy to play but tend to be simpler than mutual
priming games or back games. By slotting you hope to quickly build an
overpowering front position. This tends to work best against timid
opponents -- the kind who sit back on the ace point waiting to roll
the right double while you gammon him.
While slotting with 41 and 51 may still be wrong in match play, I
suspect slotting is more appetizing in money play using the Jacoby
rule, which encourages earlier doubling to activate gammons. The idea
here is to quickly build a position in which you can give your
opponent a nearly optimal double, or a double in a volatile, gammonish
game, and gammon him. If the plan fails, you pass. Simplistically,
your goal is to win four points, or lose one. (Plus, as any money
player can tell you, typical opponents, having forced you into a back
game, tend to overestimate their chances, double early, misplay the
position and never drop a redouble.)
Unfortunately, there aren't as many timid or weak players as there
used to be. However, slotting can still work even when it fails.
Often such games turn into complicated positions with one or both
players having several checkers back. Such positions naturally favor
the more skillful player. Besides, I think it's generally
acknowledged that JellyFish tends to undervalue back games. And it
would certainly be an interesting endeavor to attempt to quantify
exactly how much JellyFish undervalues the opening slotting plays
because of its underevaluation of the back games that slotting
sometimes lead to.
__________________________________________________
Daniel Murphy Denmark rac...@cityraccoon.com
Backgammon on-line: telnet://fibs.com:4321
In San Francisco: http://www.backgammon.org
MelRae
Daniel Murphy wrote:
I agree very much with what you have to say here, Daniel (for what my
opinion is worth!). I have been both a slotter and a splitter in my
time and have reached the conclusion that slotting is better for the
reasons you explain. I find that attacking aggressively by building
points in order, and hitting subsequently, is most profitable.
Splitting seems to result in positions where I struggle to make an
anchor and survive my opponents attack, rather than being in a position
to attack him/her. I tend to split the back men for coverage so that I
can hit returns; it doesn't matter how strong an opponent might be,
since if I am able to keep taking half a roll from him he is not going
to be able to build a board (except with some market losing doubles
perhaps) and I am unconcerned at the number of men I might have sent
back.
I don't know what Jellyfish might be doing to produce "false" results
for the slot/split debate, but I too suspect it is something to do with
human weaknesses in cubing decisions! That is, in practice, perhaps
humans are taking too many dangerous cubes against a slotter. And if
splitters tend to reach more holding and racing positions (which players
are generally good at analysing quite accurately) then the resulting
cube plays will tend to be much more efficient, leading to less
significant errors.
Don Rae
Morten Wang
[MelRae <mel...@netvigator.com>]
> I don't know what Jellyfish might be doing to produce "false" results
> for the slot/split debate, but I too suspect it is something to do with
> human weaknesses in cubing decisions!
These JellyFish rollouts of opening positions, how old are they, and
what version was used to do them? Now we have JellyFish 3.0 which
should play backgames better, will it make a difference? If so, how
about setting up a new set of rollouts and get the results?
Morten!
--
"God does not deduct from our alloted life span
the time spent playing backgammon."
--> Morty on FIBS
--> Backgammon homepage: http://home.sn.no/~warnckew/gammon/
Philippe Michel
In article <3417905...@news.businessnet.dk>,
Daniel Murphy <rac...@cityraccoon.com> wrote:
> [...] in money play using the Jacoby
> rule, which encourages earlier doubling to activate gammons.
Does the Jacoby Rule really have a noticeable influence ?
Could someone show an early- to middle-game position that is a reasonably
clear double with the Jacoby Rule in effect and a reasonably clear
non-double without it ?
Tapio Palmroth
js...@mail2.sas.upenn.edu (John S Mamoun) wrote:
>
>
> I played a backgammon match last year with an expert, and he
> played an X-1 opening roll by slotting on the five point,
> instead of splitting the back men. I asked him, "Hmmm...are
> you sure that's the best move? Jellyfish says its wrong."
> He says, "Yes, I know that, and Jellyfish is a great player,
> but its not God."
>
> I've spent a year thinking about who is right, JF or the human.
> I'm still not quite sure what the answer is...
>
> I suppose I'd tilt towards the bot, but Is JF God or is it not?
just one big question :
if you are doing something , do you know what you are doing and
to what your doings are leading ?
okay , jf says play x is better than play y and you find similiar
situation and you play x-way.you never wonder why jf says so , or you
just think : its jf, must be good :)
of course jf is good , but is it good because it doesn't make major
mistakes as we all humans tend to do from time to time ?
so if you know what you are doing and to where it may lead the game ,
it sure is better to do that even though jf and some experts consider
it as a mistake or something.
but they don't know whats happening in your mind , how you want to
play your games and this situation .are you more happy in some sorts
of games your opponent don't want to be or is it just your weird
sense of humor which is leading you ?
whatever it is , try to be happy every time you disagree with jf or
experts , because it is the sign that you are in the right course
to learn to play this bigfungame.
you have to think with your own brains , even though they be smaller
than somebody else's , but they are yours.
so :)
tapio
Herbert Kruse
Fredrik Dahl wrote:
>
> John S Mamoun wrote:
> >
> > I played a backgammon match last year with an expert, and he
> > played an X-1 opening roll by slotting on the five point,
> > instead of splitting the back men. I asked him, "Hmmm...are
> > you sure that's the best move? Jellyfish says its wrong."
> > He says, "Yes, I know that, and Jellyfish is a great player,
> > but its not God."
> >
> > I've spent a year thinking about who is right, JF or the human.
> > I'm still not quite sure what the answer is...
> >
> > I suppose I'd tilt towards the bot, but Is JF God or is it not?
>
> extreme positions.
>
> The rollout results are almost dead even for 21, but with 41 and 51 it
> can't be right to slot.
>
> - Fredrik Dahl
Thats a good joke (41 and 51 are clear).
In my eyes it seems, that bots have problems with these slotting plays.
But how can anyone say, that this or that is definitly right??
Its much more important to do the right play for YOUR style, so test it
YOURSELF and do the best YOU can!
Btw: my results with slotting are VERY good against all bots, thats the
reason i love to play them; beside the reason of non-psychologic doubles
of the most bots (understand your opponents doubling strategie and use
the knowledge!).
- Herbert Kruse
lee
In article <5ve7q1$d...@tron.sci.fi>, Tapio Palmroth <her...@sci.fi>
writes
>js...@mail2.sas.upenn.edu (John S Mamoun) wrote:
>>
>>
>> I played a backgammon match last year with an expert, and he
>> played an X-1 opening roll by slotting on the five point,
>> instead of splitting the back men. I asked him, "Hmmm...are
>> you sure that's the best move? Jellyfish says its wrong."
>> He says, "Yes, I know that, and Jellyfish is a great player,
>> but its not God."
>>
>just one big question :
>if you are doing something , do you know what you are doing and
>to what your doings are leading ?
>okay , jf says play x is better than play y and you find similiar
>situation and you play x-way.you never wonder why jf says so , or you
>just think : its jf, must be good :)
>of course jf is good , but is it good because it doesn't make major
>mistakes as we all humans tend to do from time to time ?
>so if you know what you are doing and to where it may lead the game ,
>it sure is better to do that even though jf and some experts consider
>it as a mistake or something.
>but they don't know whats happening in your mind , how you want to
>play your games and this situation .are you more happy in some sorts
>of games your opponent don't want to be or is it just your weird
>sense of humor which is leading you ?
>whatever it is , try to be happy every time you disagree with jf or
>experts , because it is the sign that you are in the right course
>to learn to play this bigfungame.
>you have to think with your own brains , even though they be smaller
>than somebody else's , but they are yours.
>
Nice one!
JF may be the best so far, but it also helps to use your own brains.
Until jf xx is released God is still a long way ahead.
Lee
--
----------------------------------------------
l...@infoplus.demon.co.uk
Web site: http://www.ibmpcug.co.uk/~oak/info/
Email for links to/from this site
----------------------------------------------