Ok thanks
El lun, 17 oct 2022 14:38, Øystein Schønning-Johansen
escribió:
> It's just an option for use with initial positions, such that the first
> roll is never a double. Nice if you want to roll out a match equity table.
> I'm not sure how it handle the turn, when this option is used. I
It's just an option for use with initial positions, such that the first
roll is never a double. Nice if you want to roll out a match equity table.
I'm not sure how it handle the turn, when this option is used. I think I
have to read the code to answer that.
-Ø
On Mon, Oct 17, 2022, 14:22
Hi, I can't understand what the "Rollout as initial position" option is for.
and seed, the result is the
same for any numbers of threads.
I haven't looked in a long time, but at one time weren't all rollouts
done via Mersenne independent of what RNG is used for in game dice
generation?
There are separate RNGs for play and rollouts and the latter is Mersenne
by default a
is the
> same for any numbers of threads.
>
I haven't looked in a long time, but at one time weren't all rollouts
done via Mersenne independent of what RNG is used for in game dice
generation?
___
Bug-gnubg mailing list
Bug-gnubg@gnu.org
https://lists.gnu.org/mailman/listinfo/bug-gnubg
On Wed, 1 Feb 2017, Øystein Schønning-Johansen wrote:
The rollout code in GNU Backgammon is of course threaded. How is this
threaded actually? Is it threaded on each sample (aka each game). One
thread pr. game? Or one thread pr. evaluation? Or?
I'm not sure I really understand how it is done,
Hi!
It's a long time since I browsed the GNU Backgammon code now. :-)
The rollout code in GNU Backgammon is of course threaded. How is this
threaded actually? Is it threaded on each sample (aka each game). One
thread pr. game? Or one thread pr. evaluation? Or?
I just wrote a code that does a
, but I do think it's much
better than what we've got now. Having looked any many rollouts at
bgonline.org, I find XG's style much easier to read than gnubg's or Snowie's.
It presents the data in an order much more suited to the thought process I am
interested in.
There is always a tradeoff being
: Listing JSDs in rollouts
Hi Ian,
I'm with you on making the rollout reports easier to read. For those who prefer
the old reporting, there could be an option for users to turn on if they
wanted, with the default being the new easier-reading variety.
What would you think of also converting
-sysware.com; bug-gnubg@gnu.org
Date: Tue, 31 May 2011 12:32:51 +
CC: leobu...@usa.net
Subject: Re: [Bug-gnubg] Fwd: Listing JSDs in rollouts
Since no-one else has commented I'll chime in here:
Adding the JSDs is a great idea, if it can be done tidily.
Now I'm going to hijack
On Tue, 31 May 2011, Ian Shaw wrote:
Gnubg has a larger problem in it's rollout reporting.
The original is on the left. Look how messy it is.
On the right I have tweaked it to reduce redundant information, and
include the suggested JSD information. I think the primary data - the
rollout
Wondering whether y'all have decided on the suggestion that the rollout
results indicate the plays' JSDs. Note that because we can manually stop
rolling out specific plays it is not clear in the outoput listing whether a
rollout was stopped manually or whether the JSD limit was reached. In
Hi
If I try to continue cmarked rollouts done on an older build with the latest
build, it crashes.
Another user has reported the same problem to me.
Michael
_
Send us your Hotmail
On Sat, 16 Jan 2010, Christian Anthon wrote:
As usual an example makes the problem easier to track...
On Sat, Jan 16, 2010 at 1:47 PM, Michael Depreli
michaeldepr...@hotmail.co.uk wrote:
Since updating to build 20091230 I'm having problems with cmarked rollouts
stopping
short of the set
...
On Sat, Jan 16, 2010 at 1:47 PM, Michael Depreli
michaeldepr...@hotmail.co.uk wrote:
Since updating to build 20091230 I'm having problems with cmarked
rollouts
stopping
short of the set trials for some moves.
I'm using 1296 trials minimum 720 if jsd 2.33 is reached.
I'm seeing
On Sun, 17 Jan 2010, Philippe Michel wrote:
[...] I get the same kind of behaviour all the time [...] It doesn't
particularly shock me [...]
On the other hand, there are other problems in the stop on JSD code :
This :
/* it was stopped, catch it up to the other moves and resume
Since updating to build 20091230 I'm having problems with cmarked rollouts
stopping
short of the set trials for some moves.
I'm using 1296 trials minimum 720 if jsd 2.33 is reached.
I'm seeing that one move has reached 1296 but the others seem to be stopping at
anything between 1295 and 1285
As usual an example makes the problem easier to track...
Christian.
On Sat, Jan 16, 2010 at 1:47 PM, Michael Depreli
michaeldepr...@hotmail.co.uk wrote:
Since updating to build 20091230 I'm having problems with cmarked rollouts
stopping
short of the set trials for some moves.
I'm using 1296
from srpm to get the to work on your fedora system.
Christian.
On Wed, Dec 16, 2009 at 2:56 AM, Timothy Y. Chow tc...@alum.mit.edu wrote:
I've installed a version of gnubg from early August 2009 on a
multi-processor machine running Fedora 10, and the way I do long rollouts
is to type something
I've installed a version of gnubg from early August 2009 on a
multi-processor machine running Fedora 10, and the way I do long rollouts
is to type something like
gnubg -t inputfile outputfile
This works fine, but I would like to take advantage of the multiple
processors to get
The numbers will be different because of
a) cube dependent moves
b) truncation of the game at double/pass
Christian.
On Wed, Oct 28, 2009 at 9:18 PM, Adi Kadmon adi.kad...@gmail.com wrote:
Hi,
I've noticed that the data in the rubric Cubless in a cubeful rollout
are different, sometimes
-- Forwarded message --
From: Adi Kadmon adi.kad...@gmail.com
Date: Wed, Oct 28, 2009 at 10:18 PM
Subject: Different cubeless data between CL and CF rollouts
To: bug-gnubg@gnu.org
Hi,
I've noticed that the data in the rubric Cubless in a cubeful rollout
are different, sometimes
Hi,
I've noticed that the data in the rubric Cubless in a cubeful rollout
are different, sometimes very differnet, from the data in the rubric
Cubeless in a cubeless rollout of the same move/s.
I'd like to ask:
(a) What is the reason for that difference?
(b) Which one of the above two groups of
? rollouts, red font is not always
on the actual move made.
From: christian.ant...@gmail.com
To: misja.a...@gmail.com
CC: bug-gnubg@gnu.org
Examples, as precise as possible, makes this much easier to track.
Christian.
On Sat, Oct 10, 2009 at 11:20 AM, Misja Alma misja.a...@gmail.com wrote:
I
remained marked as doubtful.
Misja
2009/10/10 Michael Depreli michaeldepr...@hotmail.co.uk:
I think this may have been reported before.
After Cmarked rollouts, sometimes the red font is not on the actual move
made.
This can be cured by re-analyzing the match. Rollouts appear to be
preserved
I think this may have been reported before.
After Cmarked rollouts, sometimes the red font is not on the actual move made.
This can be cured by re-analyzing the match. Rollouts appear to be preserved.
Michael
, but then in the analysis window gnubg's originally preferred
move became highlighted in red, and probably because of that the move
in the moves explorer remained marked as doubtful.
Misja
2009/10/10 Michael Depreli michaeldepr...@hotmail.co.uk:
I think this may have been reported before.
After Cmarked rollouts
Also reported by another user on the BGO forum.
Perform some CMark rollouts.
Increase the rollout trial numbers and select a few moves.
Gnubg crashes when you hit the rollout button.
Michael
by another user on the BGO forum.
Perform some CMark rollouts.
Increase the rollout trial numbers and select a few moves.
Gnubg crashes when you hit the rollout button.
Michael
New! Receive and respond to mail from other email accounts from within Hotmail
Find out how. http
Sorry I think you're right. I was using my second PC and forgot to check the
version (it was 20091001)
20091002 seems to be OK.
Date: Sun, 4 Oct 2009 04:12:36 -0600
Subject: Re: [Bug-gnubg] Extending CMarked rollouts crashes gnubg
From: mpe...@capp-sysware.com
To: michaeldepr...@hotmail.co.uk
That makes sense. I am trying to reproduce the other bug reported yesterday
on BGO (in the same thread) that was posted regarding rollouts (As shown
below). Has anyone been able to reproduce this behavior?
Posted By: Iancho Hristov backgammon.hris...@gmail.com
Date: Saturday, 3 October 2009
This was posted to BGO by W Womack:
Anyone else having a problem with build 20091001 crashing when you try to
do a rollout? XP machine, worked fine with previous builds.
Has anyone seen this?
___
Bug-gnubg mailing list
Bug-gnubg@gnu.org
Michael Petch wrote:
This was posted to BGO by W Womack:
Anyone else having a problem with build 20091001 crashing when you try to
do a rollout? XP machine, worked fine with previous builds.
Has anyone seen this?
More importantly, can anyone confirm this?
Christian.
I just tried it. I took a random position (tried this twice on 2 different
ones) that were previously analyzed. selected the top 3 moves from the
analysis pane and then hit the rollout button. I don't have debug info but
it seems to crash libgtk-win32-2.0-0.dll (I know that's not really helpful)
I can produce a crash this way:
- start a match, when it's your turn, before rolling, click on Analysis/Rollout
- I get a pop up with the following message:
This action will roll out the current position before the dice are
rolled. It is
not possible to extend nor save this rollout ?
And just
Here it is (from 20091002 code).
I start a match, wait for my turn, before rolling click on analysis/rollout,
get the dialog with the strange message, click on OK, crash.
MaX.
D:\Documents\gnubggdb gnubg-gui-py-sse2-wmt-O3.exe-debug.exe
GNU gdb 6.8
Copyright (C) 2008 Free Software Foundation,
Yes it crashes. As far as I can see with whatever rollout.
My home brewed Windows also crashes on both Vista machines.
N.
On Fri, Oct 2, 2009 at 8:54 AM, Christian Anthon
christian.ant...@gmail.com wrote:
Michael Petch wrote:
This was posted to BGO by W Womack:
Anyone else having a problem
On Fri, Oct 2, 2009 at 12:09 PM, motiv4u moti...@gmail.com wrote:
Yes it crashes. As far as I can see with whatever rollout.
My home brewed Windows also crashes on both Vista machines.
N.
Home brewed latest CVS (this morning) seems to be OK so far.
N.
On Fri, Oct 2, 2009 at 8:54 AM,
Likely I did something wrong with the stop all button for cmarked
rollouts. I'll take a look tonight.
Christian.
motiv4u wrote:
On Fri, Oct 2, 2009 at 12:09 PM, motiv4u moti...@gmail.com wrote:
Yes it crashes. As far as I can see with whatever rollout.
My home brewed Windows also crashes
I am committing the fix Christian.
On 02/10/09 4:17 AM, Christian Anthon christian.ant...@gmail.com wrote:
Likely I did something wrong with the stop all button for cmarked
rollouts. I'll take a look tonight.
Christian.
motiv4u wrote:
On Fri, Oct 2, 2009 at 12:09 PM, motiv4u moti
Michael Petch wrote:
I am committing the fix Christian.
Thx. :)
Christian.
___
Bug-gnubg mailing list
Bug-gnubg@gnu.org
http://lists.gnu.org/mailman/listinfo/bug-gnubg
Committed to cvs, if someone can test it on Windows and let me know I'd
appreciate it.
On 02/10/09 4:25 AM, Michael Petch mpe...@capp-sysware.com wrote:
I am committing the fix Christian.
___
Bug-gnubg mailing list
Bug-gnubg@gnu.org
I will let you guys decide how you wish to word it. I only fixed the crash.
On 02/10/09 1:57 AM, Massimiliano Maini maxma...@gmail.com wrote:
First, the message is in poor english and unclear.
Second, there should probably be also a cancel button.
You may use the cmark functionallity in recent versions to mark your
moves and cubes. Analyse-rollout-match will then roll-out the marked
choices.
Christian.
___
Bug-gnubg mailing list
Bug-gnubg@gnu.org
http://lists.gnu.org/mailman/listinfo/bug-gnubg
Hello,
I wanted to ask if there's a way to give GNU in advance (like programing)
several psitions to rollout (say, the proper cube decision in each), so that
when I come home after a long day I can find 2 or more different positions
rollouted already?
Thanks,
-- Adi
00:00:41
Thr = number of threads
XGR = Extreme Gammon XGRoller emulation: 0 ply rollouts truncated at 6
moves, first two cube decisions at 1-ply.
Extreme Gammon does not use VR for XGRoller, but I have repeated the
test with VR on.
It is clear that the speed hit is enormous when using VR for 0-ply
Are you sure that for 0-ply switching off VR is more accurate in a given time?
Unless I'm doing something wrong I get the opposite results.
Subject: RE: [Bug-gnubg] Feature Request: Gnu to use rollouts for playing
Date: Fri, 18 Sep 2009 13:12:12 +0100
From: ian.s...@riverauto.co.uk
the masses it might be optimal to
choose no VR and lots of trials in a short time.
-- Ian
Subject: RE: [Bug-gnubg] Feature Request: Gnu to use rollouts
for playing
Date: Fri, 18 Sep 2009 13:12:12 +0100
From: ian.s...@riverauto.co.uk
To: bug-gnubg@gnu.org
play comparison rollouts. (Relative equity rather than absolute)
I haven't checked out cube decisions but it did occur to me that it wasn't
necessary to have the
same settings for cube and checker rollouts if it could be shown that different
settings were better for each.
Subject: RE: [Bug
Ideal would be just to implement the feature and make it configurable.
Gnubg should be able to play (and do rollouts ?) at a configurable
rollout level.
Establishing what level (where to truncate, ply during play, how to
stop , VR on/off)
is good enough may be a personal matter. If the feature
I'm sure it's obvious but I couldn't find it..
How do I stop the Cmarked rollouts?
If I hit STOP it just moves on to the next one ( a bit problematic when I have
over 100)
Thanks
_
Save time by using Hotmail to access your
, michaeldepr...@hotmail.co.uk
michaeldepr...@hotmail.co.uk wrote:
I'm sure it's obvious but I couldn't find it..
How do I stop the Cmarked rollouts?
If I hit STOP it just moves on to the next one ( a bit problematic when I have
over 100)
Thanks
OK
I think the Cmark feature is excellent and should be enhanced.
I already mentioned on another post about having the ability to scan the files
by Cmark with a toolbar button
and also to be able to clear CMark rollouts without clearing the whole analysis.
I'm planning on using the Cmark
threading support all rollouts will end with a trial
count for each move of 2592. If the JSD for a move falls back below the 3.1
it does properly restart the rollout afrom the point where it originally
stopped on a move. The other curiousity is that although the trial count
continues to rise
-Original Message-
From: Christian Anthon [mailto:christian.ant...@gmail.com]
Sent: 20 August 2009 07:28
To: Ian Shaw
Cc: bug-gnubg@gnu.org
Subject: Re: [Bug-gnubg] Feature Request: Gnu to use rollouts
for playing
Most of the code is already in place. Things to consider
Can you check the speed of the threaded/non-threaded rollouts (say
0ply eval 6ply truncation).
Christian.
On Fri, Aug 21, 2009 at 6:54 PM, Ian Shawian.s...@riverauto.co.uk wrote:
-Original Message-
From: Christian Anthon [mailto:christian.ant...@gmail.com]
Sent: 20 August 2009 07:28
On 18/06/09 9:01 PM, Michael Petch mpe...@capp-sysware.com wrote:
Click new, click pencil/paper icon (new position), set up the position.
Since it is a cube analysis you are interested in don't set the dice to
anything. Now, Click Edit to stop editing.
Click on the Hint icon. The cube analysis
On 19/06/09 3:45 AM, Alain Redlinger redlin...@polytech.univ-montp2.fr
wrote:
When setting up some variation starting from an existing position where the
dice have already been set, the dice selection dialog
box does not offer means to nullify this dice selection (only to modify it for
some
On Fri, Jun 19, 2009 at 6:40 PM, Michael Petchmpe...@capp-sysware.com wrote:
On 19/06/09 3:45 AM, Alain Redlinger redlin...@polytech.univ-montp2.fr
wrote:
When setting up some variation starting from an existing position where the
dice have already been set, the dice selection dialog
box
On 19/06/09 5:06 PM, motiv4u moti...@gmail.com wrote:
You can click on the small checker icon near the names of the players
to set turn ( and thus clear the dice)
Either in Edit mode or 'normal' play mode (easier when both players
Human, otherwise for the top player you must Game / Computer
On Wed, 17 Jun 2009, Michael Petch wrote:
To get that data, I do cube rollouts by clicking on the position then go
to the Analyze menu button, select Rollout/Cube. That analysis
window should have an ok button as you describe. Click on Ok. Now
the window is gone. Now make sure on the View
pane, gnubg just rolls the dice for me, and it says
No analysis available.
Generally I do rollouts from positions that were in an entire match. My
method works when you are using an existing move in a match (Not editing a
position). I did as you suggested and I must admit - it doesn't work the way
On 18/06/09 9:01 PM, Michael Petch mpe...@capp-sysware.com wrote:
On a side note. If you want to enter a position directly, you don't need to
start a new money session or match. Click new, and then click on the left
most icon (pencil and paper). That will drop you right into edit position
that information again. How do I find and save that information?
This seems like a very basic question but the documentation on rollouts
is very unhelpful in this regard.
Tim
___
Bug-gnubg mailing list
Bug-gnubg@gnu.org
http://lists.gnu.org/mailman
the window simply disappears and I can't even figure out how to
find that information again. How do I find and save that information?
This seems like a very basic question but the documentation on rollouts
is very unhelpful in this regard.
To get that data, I do cube rollouts by clicking
On 17/06/09 8:34 PM, Michael Petch mpe...@capp-sysware.com wrote:
Cubeless analysis
Should have been
Cube Analysis
___
Bug-gnubg mailing list
Bug-gnubg@gnu.org
http://lists.gnu.org/mailman/listinfo/bug-gnubg
On Wed, Jun 10, 2009 at 10:45 AM, Jim Segravej...@j-e-s.net wrote:
I haven't checked, but the rollout logs (which I put in to debug the
changes to rollouts I was making) were completely non-thread
safe. There was a single file handle to which all moves were being
output. I tried using it once
-ohio dca...@daychem.com
Date: Sunday, 14 June 2009, at 8:58 p.m.
auto save rollout after every (x)# of trials. When running really
reay long rollouts, it would be nice to not lose a ton of data due
to a power outage or similar problem
On Tue 10 Feb 2009 (10:01 +0100), Christian Anthon wrote:
Hi Bob,
I wont have time to look into it any time soon, so here is just a
couple of suggestions.
a) truncation of the rollouts may help you track down what is going on
b) turn on rollout logs by issuing the commands
set rollout
Hi Bob,
I wont have time to look into it any time soon, so here is just a
couple of suggestions.
a) truncation of the rollouts may help you track down what is going on
b) turn on rollout logs by issuing the commands
set rollout log on
set rollout logfile bob_is_an_intermediate_player
which
One of the available criteria to end a rollout is when standard deviations
are small enough. For instance, it could be :
Stop rollout when STD's are small enough
After 1296 games, rollouts will stop if the STDs are small enough.
Rollouts can stop when the ratio |STD/value| is less than 0.0040
On Wed, Jan 28, 2009 at 12:08 AM, Matt Cohn-Geier mmcohnge...@gmail.com wrote:
Thanks, this seems to work great so far. When you get a chance could you
also add something for cube rollouts? Not as high a priority for me though.
That is the plan.
I do notice that this crashes when you try
Sorry for seeming stupid about this, but having now run a fair number of
cube decision rollouts, what I find is that in a position where a rollout
will show the double as clear and the take decision close, such as GNUBG: ID
pp3BgAmYb4MHAA:cAmgACAAGAAA ticking Stop rollout when one move appears
If it did work that way it was broken. The two rollouts are expected
to have the same number of trials.
Christian.
On Wed, Jan 28, 2009 at 5:51 PM, Neil Robins neil.robi...@ntlworld.com wrote:
Sorry for seeming stupid about this, but having now run a fair number of
cube decision rollouts, what
In every rollout of a cube decision I have tried (single thread), a 1 and 2
Rank/no. never appears and both the Centred 1-cube and Player x owns 2-cube
have JSD figures against them, suggesting some mythical third choice must be
best.___
Bug-gnubg
Reading the Changelog is not a bad idea when you are using devel
software :) The first JSD is that of the double/nodouble decision and
the second is that of the take/pass decision. If things work as they
should, one should now be able to use the JSD criteria with cube
rollouts.
Christian.
On Tue
Hi all,
I was looking for ways to test the rollout code and came up with the following:
a) Setup evaluation on ply 1, and turn off the pruning net.
b) Setup rollouts to truncate after 1ply, 36 trials, quasi-random dice
on, varredn off, 0ply cubeful play and cube(I believe that pruning
isn't
the CL equities for matches.
Christian.
On Sun, Dec 14, 2008 at 7:59 PM, stieb...@onlinehome.de wrote:
I have a question concerning the cubeless equity derived from cubeful
rollouts
Date: Dezember 14, 2008
Move number 3: X to play 32
GNU Backgammon Position ID: mGdrBgDLzInBAA
Hi list,I am trying to run rollouts on Debian. I don't seem to be able to do
rollouts via the CLI and save them as .sgfs. The examples I have looked at
have rollouts being done by way of hint which is not what I want, as this
rollout data can't be saved or extended later. If I step through
I have a question concerning the cubeless equity derived from cubeful rollouts
Date: Dezember 14, 2008
Move number 3: X to play 32
GNU Backgammon Position ID: mGdrBgDLzInBAA
Match ID : cImpACAA
+24-23-22-21-20-19--18-17-16-15-14-13-+ O: somebody
I tried using the GNU Backgammon rollout feature yesterday for the first time
but had a problem. I was setting up a position from a problems book I own to
confirm its answer but the position requires having a man on the bar.
Everything else went fine but for the life of me I couldn't figure a
***
Here is a summary of some of the weird behavior I?ve noticed when
doing rollouts involving noise. I suspect it may just be an issue of
something being done in a wrong order. I used a 1 point match, zero
skill position so the weakening of either player
Hi folks,
I'm forwarding an e-mail by Bob Koca (he also tried to show me this
behavior of gnubg whwn we met in Monte Calro):
***
Here is a summary of some of the weird behavior I?ve noticed when
doing rollouts involving noise. I suspect
You need to navigate to the right position first. If you are still
having problems, send us a file and tell us exactly what you are
trying to do.
Christian.
On Sat, May 24, 2008 at 10:44 AM, [EMAIL PROTECTED] wrote:
So, is it possible to do rollouts of a checker play and/or cube decision
On Sat, May 24, 2008 at 10:44 AM, [EMAIL PROTECTED] wrote:
So, is it possible to do rollouts of a checker play and/or cube
decision and save the results without doing anything in
the GUI? I'm
confused, I know I have never succeeded in doing it, but I
also know
that my
Am 27.05.2008 um 16:22 schrieb Ian Shaw:
Try export GAME or export MATCH rather than export POSITION.
Nonetheless export position worked here with rollout results.
Ciao
Achim
___
Bug-gnubg mailing list
Bug-gnubg@gnu.org
Am 24.05.2008 um 10:44 schrieb [EMAIL PROTECTED] [EMAIL PROTECTED]
:
So, is it possible to do rollouts of a checker play and/or cube
decision and save the results without doing anything in the GUI?
I'm confused, I know I have never succeeded in doing it, but I also
know that my
So, is it possible to do rollouts of a checker play and/or cube decision and
save the results without doing anything innbsp; the GUI? I#39;m confused, I
know I have never succeeded in doing it, but I also know that my question is
hardly original, and advice can be found on the web (e.g.
http
Sounds theoretically true. I tried a bit to setup such a
situation but I didn't manage to get one. Anybody has one
example of an eval leading to a too good/take ?
Øystein O. Johansen [EMAIL PROTECTED] wrote on 15/04/2008
14:53:35:
GNU Backgammon Position ID: tm0DAHB32w4AAA
Jonathan Kinsey [EMAIL PROTECTED] wrote on 14/04/2008 15:18:00:
From: [EMAIL PROTECTED]
- how can a position be too good for player A and be a take for player B
?
I thought that, by definition, a position is too good when my
(normalized)
equity is grater then +1.000. In this case, if
Sounds theoretically true. I tried a bit to setup such a situation but I
didn't manage to get one. Anybody has one example of an eval leading to a
too good/take ? MaX.
Yes, this one:
Position ID: bdsBKgC27QwGAA
Match ID: UQkgARAAKAAA
/Svempa
Not sure anyone has been following rec.games.backgammon, but I am (And so
are others) wondering how this position can be Too good to redouble, take
I can understand that its possible to be Too good to redouble, take if the
opponent can be gammoned, I'm just not sure how this can be possible
This the odd/even ply problem at its worst. First of all the cube
action is strange because gnubg grossly mis-evaluates the position for
one or both sides. Look at Max' examples and I'm sure you can come up
with a reasonable explanation.
So, is it a bug in the code or in the net(s)? This is
On Mon, Apr 14, 2008 at 2:19 PM, Massimiliano Maini
[EMAIL PROTECTED] wrote:
- how can a position be too good for a player that can not score a gammon ?
Well, he could get backgammons, but it doesn't seem to be the case above.
The most likely cause is that O makes some bad takes on the next
for Windows. Will this compile
for Linux? And will doing so make the process more efficient? Am I
completely misunderstanding all of this?
When I run two instances of gnubg doing rollouts the cores max out and
I notice that the estimated time to completion stays the same for both
rollouts. That seems
-thread
support but that the binaries are only for Windows. Will this compile
for Linux? And will doing so make the process more efficient? Am I
completely misunderstanding all of this?
When I run two instances of gnubg doing rollouts the cores max out and
I notice that the estimated time
Michael Petch [EMAIL PROTECTED] writes:
According to Debian Testing they are using a snapshot of GnuBG source
from November 2006. The multithreaded support was released around the
end of January 2007.
Yeah, Debian testing is using the latest stable snapshot (rel_0_15). I
keep hoping y'all
to the
evaluation of game over positions that is if 1.00 is sometimes (for example)
0.91.
Personally I think that adding noise is a bit of an abomination, which has
only small justification in human vs. computer play, and no justification
whatsoever when doing rollouts. The data you obtain might be right
but this is not the case. Note that
if cubeful is on the results are all over the place. Also if cubeful is not
used but the bearoff database is not used.For all of these rollouts I used
variance reduction, quasi random dice, and mersenne with seed 1000.
If the table below is not formatted nicely see
Hi,
I just built gnubg on my new quadcore pc with threading enabled (on Ubuntu
7.04).
When I run calibrate, indeed it shows on my system monitor that all 4 cores
are used.
When I run a rollout and I have specified that I want to use only 1 core,
there is no problem.
But, when I specify that I
1 - 100 of 117 matches
Mail list logo