🚀 go-pugleaf

Version 2.0, no idea if it's better than v1.0. It's voice enabled (may
be the only difference) it sure plays a mean game of 'gammon though !!
John

Lex Weaver wrote:
> 
> Can someone please tell me the version number of the TD-GAMMON package
> distributed in the Family Fun Pack as part of OS/2 Warp 4?
> 
> I have OS/2 v2.1 with TD-GAMMON v1.0, and am wondering if the latest
> release is any newer or better.
> 
> cheers,
> Lex
> 
> 
> 
> ----------------------------------------------------------------------------
> Lex Weaver
> lex@cs.anu.edu.au               http://cs.anu.edu.au/~Lex.Weaver
> Ph: +61 6 279 8180 (w)           Add: c/- Dept of Computer Science
>                                       Australian National University
>                                       Australia 0200
> ============================================================================
> 
>         "Often the difference between a successful man and a failure
>         is not one's better abilities or ideas, but the courage that
>         one has to bet on his ideas, to take a calculated risk - and
>         to act."
>                         Maxwell Maltz

John Littler wrote:
> 
> Version 2.0, no idea if it's better than v1.0. It's voice enabled (may
> be the only difference) it sure plays a mean game of 'gammon though !!
> John
> 
> Lex Weaver wrote:
> >
> > Can someone please tell me the version number of the TD-GAMMON package
> > distributed in the Family Fun Pack as part of OS/2 Warp 4?
> >
> > I have OS/2 v2.1 with TD-GAMMON v1.0, and am wondering if the latest
> > release is any newer or better.


Okay, I'll bite.  Where do we find version 2.0?

Peter Pollack

--

It's on the application sampler CD that comes with Warp 4

Ciao
  Johannes

In article <01bc182f$5ab0d880$132ee926@dlane>, DLane@ContactPt.com says...
>
>Is a version that will work with Warp 3 available?
>
>Anybody know?

The version that one could get for fee from IBM's ftp server works fine
under Warp 3 FP17. Actually, that's no quite true as the colour of the
pieces got wierd with the ATI MACH64 drivers I was using.

Oh and there was alot of net speculation that the damn thing cheats! I'm
not a backgammon slouch but it could trash me consistently. And it always
seemed to get the die rolls it needed to cover an otherwise vulnerable
position...

>David.

DDA

-- 
David D'Antonio, CNE  (dantonio@process.com)
 Some they do and some they don't and some ya just can't tell
  Some they will and some they won't and some it's just as well.
                                        -SuperTramp

On 11 Feb 97 18:44:51 -0500, dantonio@process.com (David D'Antonio)
wrote:

>In article <01bc182f$5ab0d880$132ee926@dlane>, DLane@ContactPt.com says...
>>
>>Is a version that will work with Warp 3 available?
>>
>>Anybody know?
>
>The version that one could get for fee from IBM's ftp server works fine
>under Warp 3 FP17. Actually, that's no quite true as the colour of the
>pieces got wierd with the ATI MACH64 drivers I was using.
>
>Oh and there was alot of net speculation that the damn thing cheats! I'm
>not a backgammon slouch but it could trash me consistently. And it always
>seemed to get the die rolls it needed to cover an otherwise vulnerable
>position...

I'm glad that someone else thinks the die rolls were too good to be
true ! I thought I was going paranoid. :-)

________________________________________
David Zuidema                
http://www.physics.monash.edu.au/~davidz

David Zuidema wrote:
> 
> >Oh and there was alot of net speculation that the damn thing cheats! I'm
> >not a backgammon slouch but it could trash me consistently. And it always
> >seemed to get the die rolls it needed to cover an otherwise vulnerable
> >position...
> 
> I'm glad that someone else thinks the die rolls were too good to be
> true ! I thought I was going paranoid. :-)
> 

Good players get more good rolls than bad players. What's new?

-- 
Stephen Turner     sret1@cam.ac.uk      http://www.statslab.cam.ac.uk/~sret1/
  Stochastic Networks Group, Statistical Laboratory,
  16 Mill Lane, Cambridge, CB2 1SB, England    Tel.: +44 1223 337955
  "Collection of rent is subject to Compulsive Competitive Tendering" Cam. City

Actually, I have a theory about this.  I'm not "in the know" about this - I
don't know the people who put TD-GAMMON together - but I have noticed the
rather blatant appearance of cheating (program has three unprotected pieces,
only way I *won't* hit it at least once is if I roll a 1 2, and guess what I
roll).  My theory is that this is a game that was trained via massive
reiteration of playing against itself.  During the course of that learning,
it also "learned" the pattern to the pseudo-random numbers used to generate
the die rolls.  I understand that they probably change the seed from game to
game, but depending upon the quality of the random-number generator, there can
still be a repeating pattern, and that pattern might be built into the neural
net (or whatever other algorithm is being used) that drives the computer play.

In article <33014410.20749346@NewsServer.cc.monash.edu.au>,
David Zuidema <David.Zuidema@sci.monash.edu.au> wrote:
>On 11 Feb 97 18:44:51 -0500, dantonio@process.com (David D'Antonio)
>wrote:
>
>>In article <01bc182f$5ab0d880$132ee926@dlane>, DLane@ContactPt.com says...
>>>
>>>Is a version that will work with Warp 3 available?
>>>
>>>Anybody know?
>>
>>The version that one could get for fee from IBM's ftp server works fine
>>under Warp 3 FP17. Actually, that's no quite true as the colour of the
>>pieces got wierd with the ATI MACH64 drivers I was using.
>>
>>Oh and there was alot of net speculation that the damn thing cheats! I'm
>>not a backgammon slouch but it could trash me consistently. And it always
>>seemed to get the die rolls it needed to cover an otherwise vulnerable
>>position...
>
>I'm glad that someone else thinks the die rolls were too good to be
>true ! I thought I was going paranoid. :-)
>
>________________________________________
>David Zuidema                
>http://www.physics.monash.edu.au/~davidz
>


-- 
Jim Babka         babka@austin.ibm.com (Internet), BABKA at AUSTIN (VM)
AIX Kernel Service and Support                     USIB6UKS at IBMMAIL
What my employer thinks about what I say is best left unprinted...
  If we could just get everyone to close their eyes and visualize

In article <5dsm20$s4s@ausnews.austin.ibm.com>,
Jim Babka <babka@austin.ibm.com> wrote:
>Actually, I have a theory about this.  I'm not "in the know" about this - I
>don't know the people who put TD-GAMMON together - but I have noticed the
>rather blatant appearance of cheating (program has three unprotected pieces,
>only way I *won't* hit it at least once is if I roll a 1 2, and guess what I
>roll).  My theory is that this is a game that was trained via massive
>reiteration of playing against itself.  During the course of that learning,
>it also "learned" the pattern to the pseudo-random numbers used to generate
>the die rolls.  I understand that they probably change the seed from game to
>game, but depending upon the quality of the random-number generator, there can
>still be a repeating pattern, and that pattern might be built into the neural
>net (or whatever other algorithm is being used) that drives the computer play.
>

a) you're right about the technique used for building TD-gammon (last time I 
   heard, it was a self-trained neural network.

b) you're almost certainly wrong in your suggestion that TD-gammon acts based 
   upon learned knowledge about the way the dice perform.

  What you're proposing (as I understand) it is that TD-gammon gives high 
equity ratings to positions which, though bad, are made good by a persistent 
bias of the dice, and thus are 'safe'.  For this sort of persistent bias to 
exist, it would appear to be necessary for the dice to be linked in some way to
the positions being seen, a near-impossibility.  Given that TD's evaluation of 
positions is based upon a 'generalization', if you will, of a plethora of 
somewhat similar situations, I rather suspect that you could get away with 
using a decidedly _non_ random die sequence for training, and that you would 
still end up with a fairly valid evaluation engine, i.e. one which is not 
biased towards certain risky situations.

  As further evidence that no such bias exists, witness the fine performance 
of neural net-based 'bots like td-one, jellyfish, and mloner on FIBS, where the
die-rolling algorithm is almost certainly not the one that these 'bots were 
trained with.

  How about this: I'll give you ten to one that you're mistaken.

john

David Zuidema wrote:
 
> I'm glad that someone else thinks the die rolls were too good to be
> true ! I thought I was going paranoid. :-)
> 
> ________________________________________
> David Zuidema
> http://www.physics.monash.edu.au/~davidz
Fight back!  Put it on Manual Dice Entry whenever the computer cheats!

If it's programmed to cheat, maybe it's programmed to know when it's
been caught!
-- 
Nick Cassimatis
My opinions are mine alone.

Jim Babka wrote:

> Actually, I have a theory about this.  I'm not "in the know" about this - I
> don't know the people who put TD-GAMMON together - but I have noticed the
> rather blatant appearance of cheating (program has three unprotected pieces,
> only way I *won't* hit it at least once is if I roll a 1 2, and guess what I
> roll).  My theory is that this is a game that was trained via massive
> reiteration of playing against itself.  During the course of that learning,
> it also "learned" the pattern to the pseudo-random numbers used to generate
> the die rolls.  I understand that they probably change the seed from game to
> game, but depending upon the quality of the random-number generator, there
> can still be a repeating pattern, and that pattern might be built into the
> neural net (or whatever other algorithm is being used) that drives the
> computer play.

There are only three big problems with this theory.

1. The "patterns" in any halfway-decent pseudo-random number generator
   are *way* too subtle to be spotted by a neural network. They just
   aren't the sort of thing you can detect that way.

2. TD-Gammon wasn't trained on a PC running OS-2. It was trained on an
   RS/6000 running (I think) AIX. It's unlikely (though possible) that
   these two systems use the same random number generator.

3. What a neural network like TD-Gammon learns is an estimate of the
   value of the position (I am over-simplifying here). It doesn't get
   told (e.g.) the move number, or the current state of the RNG, or
   anything like that.

In other words: your theory is interesting, but it is a million miles
away from being possible.

-- 
Gareth McCaughan       Dept. of Pure Mathematics & Mathematical Statistics,
gjm11@dpmms.cam.ac.uk  Cambridge University, England.

In <5dsm20$s4s@ausnews.austin.ibm.com>, babka@austin.ibm.com (Jim Babka) writes:
>Actually, I have a theory about this.  I'm not "in the know" about this - I
>don't know the people who put TD-GAMMON together - but I have noticed the
>rather blatant appearance of cheating (program has three unprotected pieces,
>only way I *won't* hit it at least once is if I roll a 1 2, and guess what I
>roll).  My theory is that this is a game that was trained via massive
>reiteration of playing against itself.  During the course of that learning,
>it also "learned" the pattern to the pseudo-random numbers used to generate
>the die rolls.  I understand that they probably change the seed from game to
>game, but depending upon the quality of the random-number generator, there can
>still be a repeating pattern, and that pattern might be built into the neural
>net (or whatever other algorithm is being used) that drives the computer play.
>
>In article <33014410.20749346@NewsServer.cc.monash.edu.au>,
>David Zuidema <David.Zuidema@sci.monash.edu.au> wrote:

I think you may be on to something there.  I have noticed that the computer's luck
with the die seems to be directlly related to the skill setting.  Possibly the higher
skill levels used more of the learning you describe.

Possibly the skill levels relate to ones ability to prevail over an increasingly
dis-honest opponent?   


>>On 11 Feb 97 18:44:51 -0500, dantonio@process.com (David D'Antonio)
>>wrote:
>>
>>>In article <01bc182f$5ab0d880$132ee926@dlane>, DLane@ContactPt.com says...
>>>>
>>>>Is a version that will work with Warp 3 available?
>>>>
>>>>Anybody know?
>>>
>>>The version that one could get for fee from IBM's ftp server works fine
>>>under Warp 3 FP17. Actually, that's no quite true as the colour of the
>>>pieces got wierd with the ATI MACH64 drivers I was using.
>>>
>>>Oh and there was alot of net speculation that the damn thing cheats! I'm
>>>not a backgammon slouch but it could trash me consistently. And it always
>>>seemed to get the die rolls it needed to cover an otherwise vulnerable
>>>position...
>>
>>I'm glad that someone else thinks the die rolls were too good to be
>>true ! I thought I was going paranoid. :-)
>>
>>________________________________________
>>David Zuidema                
>>http://www.physics.monash.edu.au/~davidz
>>
>
>
>-- 
>Jim Babka         babka@austin.ibm.com (Internet), BABKA at AUSTIN (VM)
>AIX Kernel Service and Support                     USIB6UKS at IBMMAIL
>What my employer thinks about what I say is best left unprinted...
>  If we could just get everyone to close their eyes and visualize


//=======================
// Chuck Cook
// cookc@clark.net
// www.clark.net/pub/cookc 
//=======================

In message <33019D58.1CB9@cam.ac.uk> - Stephen Turner <sret1@cam.ac.uk>
writes:
:>
:>David Zuidema wrote:
:>> 
:>> >Oh and there was alot of net speculation that the damn thing cheats! I'm
:>> >not a backgammon slouch but it could trash me consistently. And it always
:>> >seemed to get the die rolls it needed to cover an otherwise vulnerable
:>> >position...
:>> 
:>> I'm glad that someone else thinks the die rolls were too good to be
:>> true ! I thought I was going paranoid. :-)
:>> 

I have been suspecting that it cheats very badly at times but not always.
Maybe it looks at the date or some other thing to decide when it will
cheat.  Some days I can beat it every time, but on some days it will roll
exactly what it needs again and again and it is just about unbeatable in
this mode.
 
:>
:>Good players get more good rolls than bad players. What's new?
:>
:>-- 
:>Stephen Turner     sret1@cam.ac.uk      http://www.statslab.cam.ac.uk/~sret1/
:>  Stochastic Networks Group, Statistical Laboratory,
:>  16 Mill Lane, Cambridge, CB2 1SB, England    Tel.: +44 1223 337955
:>  "Collection of rent is subject to Compulsive Competitive Tendering" Cam. City

Bob Selfinger
******************************************************
Views expressed here may not actually be views.
******************************************************

John Clements wrote:
> 
> 
>   How about this: I'll give you ten to one that you're mistaken.
> 

Only ten? I'll give you 100.

-- 
Stephen Turner     sret1@cam.ac.uk      http://www.statslab.cam.ac.uk/~sret1/
  Stochastic Networks Group, Statistical Laboratory,
  16 Mill Lane, Cambridge, CB2 1SB, England    Tel.: +44 1223 337955
  "Collection of rent is subject to Compulsive Competitive Tendering" Cam. City

In article <5dsvjv$umu@rtpnews.raleigh.ibm.com>,
Nick Cassimatis  <NickC@!ibm.net> wrote:
>
>Fight back!  Put it on Manual Dice Entry whenever the computer cheats!
>

Hold on... if you get to re-roll the dice whenever the computer gets a really
really good roll, shouldn't the _computer_ get to re-roll the dice whenever
_you_ get a really good roll?

(BTW, think motive... what possible motive could there be for a programmer
to write a backgammon program that cheats?)

john

In <5dtqgv$3l2@news1.ee.net>, selfinger@ee.net (Bob Selfinger) writes:
>In message <33019D58.1CB9@cam.ac.uk> - Stephen Turner <sret1@cam.ac.uk>
>writes:
>:>
>:>David Zuidema wrote:
>:>> 
>:>> >Oh and there was alot of net speculation that the damn thing cheats! I'm
>:>> >not a backgammon slouch but it could trash me consistently. And it always
>:>> >seemed to get the die rolls it needed to cover an otherwise vulnerable
>:>> >position...
>:>> 
>:>> I'm glad that someone else thinks the die rolls were too good to be
>:>> true ! I thought I was going paranoid. :-)
>:>> 
>
>I have been suspecting that it cheats very badly at times but not always.
>Maybe it looks at the date or some other thing to decide when it will
>cheat.  Some days I can beat it every time, but on some days it will roll
>exactly what it needs again and again and it is just about unbeatable in
>this mode.
> 
>:>
>:>Good players get more good rolls than bad players. What's new?
>:>
>:>-- 
>:>Stephen Turner     sret1@cam.ac.uk      http://www.statslab.cam.ac.uk/~sret1/
>:>  Stochastic Networks Group, Statistical Laboratory,
>:>  16 Mill Lane, Cambridge, CB2 1SB, England    Tel.: +44 1223 337955
>:>  "Collection of rent is subject to Compulsive Competitive Tendering" Cam. City
>
>Bob Selfinger
>******************************************************
>Views expressed here may not actually be views.
>******************************************************
>

In <01bc182f$5ab0d880$132ee926@dlane>, "David Lane" <DLane@ContactPt.com> writes:
>Is a version that will work with Warp 3 available?
>
>Anybody know?
>
>David.
>
>
Yes,  It is avaliable as a free download from 
http://www.software.ibm.com

Warning: It is a serious player!

//=======================
// Chuck Cook
// cookc@clark.net
// www.clark.net/pub/cookc 
//=======================

In <5e4mi8$ij6@netnews.hinet.net>, geo1716@ms5.hinet.net writes:
>In <5dtqgv$3l2@news1.ee.net>, selfinger@ee.net (Bob Selfinger) writes:
>>In message <33019D58.1CB9@cam.ac.uk> - Stephen Turner <sret1@cam.ac.uk>
>>writes:
>>:>
>>:>David Zuidema wrote:
>>:>> 
>>:>> >Oh and there was alot of net speculation that the damn thing cheats! I'm
>>:>> >not a backgammon slouch but it could trash me consistently. And it always
>>:>> >seemed to get the die rolls it needed to cover an otherwise vulnerable
>>:>> >position...
>>:>> 
>>:>> I'm glad that someone else thinks the die rolls were too good to be
>>:>> true ! I thought I was going paranoid. :-)
>>:>> 
>>
>>I have been suspecting that it cheats very badly at times but not always.
>>Maybe it looks at the date or some other thing to decide when it will
>>cheat.  Some days I can beat it every time, but on some days it will roll
>>exactly what it needs again and again and it is just about unbeatable in
>>this mode.
>> 
>>:>
>>:>Good players get more good rolls than bad players. What's new?
>>:>
>>:>-- 
>>:>Stephen Turner     sret1@cam.ac.uk      http://www.statslab.cam.ac.uk/~sret1/
>>:>  Stochastic Networks Group, Statistical Laboratory,
>>:>  16 Mill Lane, Cambridge, CB2 1SB, England    Tel.: +44 1223 337955
>>:>  "Collection of rent is subject to Compulsive Competitive Tendering" Cam. City
>>
>>Bob Selfinger
>>******************************************************
>>Views expressed here may not actually be views.
>>******************************************************
>>
>
You're right the damn thing cheats!  I'm a fairly competent backgammon player and the system has the most unbelievable luck!
So I cheat right back using manual dice!

Rob Hutchens

gowest@ibm.net writes:

>You're right the damn thing cheats!  I'm a fairly competent backgammon
>player and the system has the most unbelievable luck! So I cheat right
>back using manual dice! 

TD-Gammon's creator, Gerry Tesauro, has affirmed publicly, and more than
once, that the TD-Gammon program does not cheat. That's good enough for
me.

Considering that Gerry's program plays backgammon on an expert level, it's
not surprising that a merely "fairly competent" player would have to cheat
to beat it.

No one I've ever known has been able to *examine* his or her game and
become a good player without first learning to blame poor play, not bad
luck, for their losses.  Some never learn this.  That's ok. I happen to
like fish. 

-- 
_______________________
Daniel Murphy
San Francisco
raccoon@cityraccoon.com

In article <raccoon.856599064@shellx>, raccoon@best.com says...
>
>gowest@ibm.net writes:
>
>>You're right the damn thing cheats!  I'm a fairly competent backgammon
>>player and the system has the most unbelievable luck! So I cheat right
>>back using manual dice! 
>
>TD-Gammon's creator, Gerry Tesauro, has affirmed publicly, and more than
>once, that the TD-Gammon program does not cheat. That's good enough for
>me.

I'm sure it doesn't "cheat" (or at least I'm sure Gerry didn't write it
to cheat), but it gets the most amazing luck to pull it out of weak positions.
And I never seem to be able to get that level of luck going for me.

>Considering that Gerry's program plays backgammon on an expert level, it's
>not surprising that a merely "fairly competent" player would have to cheat
>to beat it.

No, it doesn't ALWAYS play at expert level (at least the version I have).
But it still gets lucky. :-)

>No one I've ever known has been able to *examine* his or her game and
>become a good player without first learning to blame poor play, not bad
>luck, for their losses.  Some never learn this.  That's ok. I happen to
>like fish. 

A good backgammon player can still be defeated by the dice. It just seems
that TD-GAMMON always gets the dice on its side. And a good backgammon
player will NEVER lose if they have the die on their side!

>Daniel Murphy

DDA

-- 
David D'Antonio, CNE  (dantonio@process.com)
 Some they do and some they don't and some ya just can't tell
  Some they will and some they won't and some it's just as well.
                                        -SuperTramp

David D'Antonio wrote:
> 
> In article <raccoon.856599064@shellx>, raccoon@best.com says...
> >
> >gowest@ibm.net writes:
> >
> >>You're right the damn thing cheats!  I'm a fairly competent backgammon
> >>player and the system has the most unbelievable luck! So I cheat right
> >>back using manual dice!
> >
> >TD-Gammon's creator, Gerry Tesauro, has affirmed publicly, and more than
> >once, that the TD-Gammon program does not cheat. That's good enough for
> >me.
> 
> I'm sure it doesn't "cheat" (or at least I'm sure Gerry didn't write it
> to cheat), but it gets the most amazing luck to pull it out of weak positions.
> And I never seem to be able to get that level of luck going for me.

I bought a real (actual, physical) backgammon board to see if I'd have
better luck against a real (actual, physical) person.  The actual person
cheats just as much as TD-GAMMON does!  There's no way anyone can throw
that many double sixes!

-- 
John Abraham
Urban Modelling and Transportation Planning
j!abraham@acs.ucalgary.ca

(the "!" isn't part of my email address - be sure to remove it before
replying.
it's just there to keep the spammers at bay.)