| |
Main
Date: 07 Nov 2007 01:27:21
From: help bot
Subject: GetClub vs. Jester games faked?
|
After weeks, probably months, of reading about how GetClub's
success against the Jester program, I finally went to the Web
site again (I did not like the graphic chess board before) and
played a couple of games.
This site (ludochess.com) claims that Jester is the world's
strongest Jave chess program, so it seems ludicrous that the
games Sanny has presented here are a fair representation of
its chess play. Sure enough, Jester not only moved very
quickly (unlike anything available at GetClub), but in the first
game it did not make even a single "questionable" move. In
fact, it was much like playing Fritz (which is not a Java applet).
I noticed that even when I set the program for a somewhat
slower pace, it still moved rapidly, and thus, so did I. Just
like five minute blitz chess, this program was determined not
to get into a bind on the clock -- come what may. I lost two
out of two, and I still don't know if it would have claimed a
win on time, had I played slower and exceeded my allotment.
IMO, the only way for a decent game to occur between the
Jester I played and GetClub, would be to set Jester to move
in one second, versus GetClub's Master or Advance level --
taking a half-hour or much more per move, on average. My
question to Sanny is this: why is it that his results show
Jester playing poorly, missing obvious tactics for example,
when every time I have tried it the program plays both
quickly and strongly?
My view is that this program is most likely not much worse
than commercial programs like Fritz or Crafty, in spite of the
Java platform. The games we have seen posted here by
Sanny make no sense, unless he has crippled the Jester
program severely. If so, please inform me how this can be
done, so I don't get clobbered by Jester ever again. : >D
-- help bot
|
|
| |
Date: 10 Nov 2007 00:47:45
From: Sanny
Subject: Re: GetClub vs. Jester games faked?
|
On Nov 9, 5:12 pm, "Thomas T. Veldhouse" <[email protected] > wrote:
> In rec.games.chess.computer Sanny <[email protected]> wrote:
>
> > Here is what I think.
>
> > Beginner: 1700
> > Easy: 1800
> > Normal: 1900
> > Master: 2000
> > Advance: 2100
>
> > Recently the game was improved now Higher Levels will play much
> > stronger Games.
>
> So Sanny ... you didn't address the subject of the thread. Did you fake the
> match?
>
> --
> Thomas T. Veldhouse
>
> When the ax entered the forest, the trees said, "The handle is one of us!"
> -- Turkish proverb
There is no fun in showing Fake Games. I always try to give correct
information. I only show those matches where GetClub Wins the Match So
if 1/3 Matches the GetClub Easy level Wins I only show the game in
which getclub wins. And Other games where it lost I do not show them.
Play Chess at: http://www.GetClub.com/Chess.html
Playing Chess is more fun when playing with your mind instead of help
from other Computer. I use other programs to see how GetClub performs
with them. So that if any mistake is found that is removed.
Yesterday The game was playing the worst in last 2 months due to a
bug. That bug has been removed so I think now it will play descent
game.
Bye
Sanny
Play Chess at: http://www.GetClub.com/Chess.html
|
| |
Date: 09 Nov 2007 22:48:33
From: Sanny
Subject: Re: GetClub vs. Jester games faked?
|
On Nov 9, 5:18 pm, help bot <[email protected] > wrote:
> On Nov 9, 12:47 am, Sanny <[email protected]> wrote:
>
> > I played the games between Jester & Easy Level. Play a few games with
> > Easy Level they will play as good as Jester. I played with default
> > settings in Jester Vs Easy Level.
>
> For the record, at default levels Jester plays a move
> in just three seconds.
>
> > I am seeing that Zebediah yesterday played 2-3 games with Beginner
> > Level. He managed to beat Beginner level in just 13 moves. While
> > others take 50-60 Moves to beat the Beginner Level.
>
> I just replayed that game. While it is possible that this
> was a well-known book trap, it looked a lot like a decentchessprogram which got lucky whenGetClubrefused
> to defend its King. Generally speaking, all otherchess
> programs value the King, and they use tactical search
> extensions whenever there is a check or capture to
> avoid this sort of fiasco. BTW, if it was a well-known
> book trap, your team should be congratulated for
> getting the program to find all the book moves up to
> where it went astray -- that's quite an improvement.
>
> Also note that Zeb played exactly three games vs.
> the Beginner level; probably this was because your
> site claims you get one-third of a point for each win.
> Beating the Beginner level three times takes far less
> time than a single game against the Master level, or
> even the Normal level since it plays much tougherchess.
>
> > The reasion you loose with jester is that it has very small board to
> > analyze whileGetClubBoard is quite Big to analyze.
>
> While it is true I don't like the graphicchessboard
> and that the dark pieces on a dark backgrround are
> difficult to see, I believe I lost because Jester played
> very much like Fritz -- good, solid moves. How it
> manages to play poorly in your tests escapes me,
> but I could simply conduct my own tests and report
> the results here. Want to bet what the outcome
> will be? :>D
>
> > FurtherGetClub
> > shows what move it is thinking So you get an idea what it is going to
> > play.
>
> GetClubdisplays "thinking on Queen" or something
> to that effect; that is not the same as revealing the
> move it intends to play! There could be ten legal
> moves for that piece, or it could simply not be up to
> date.
>
> Generally speaking, except against the Beginner
> level, I switch to another "tab" instead of dying of
> boredom watching thechessscreen while the
> program thinks.
>
> > Thinking on Pawn, Thinking on Queen ... And you know what is going in
> >GetClubMind. So you get more time to descide your move.
>
> I will take that as indicating that what is on the
> display is the best move found so far; I was not
> aware of that. Manychessprograms will display
> the move they are *currently analyzing*, even if it
> is believed to be a titanic blunder.
>
> > Now Easy Level atGetClubplays as good as Jester. And in 4 Matches I
> > feel It will give 2-2 Score.
>
> Few will believe your test results, since you were
> reporting dramatic improvements which no one else
> could seem to "see", seemingly for decades before
> the noticeable improvements came along. Now many
> of the meanies or critics have disappeared -- just when
> your cries of "wolf!" are becoming genuine. I can feel
> the strength improving, but the entire village has turned
> a deaf ear.
>
> > My Computer is P-IV 2.5 Ghz 512 MB RAM. So it is Quite fast.
>
> How do you know whether the limited resources
> are fairly apportioned between the two Java applets?
> It seems to me that your reported games show
> Jester playing much weakerchessthan what I
> experienced, though I gave it a bit more time per
> move. Perhaps you need to utilize two computers,
> or perhaps this is not even possible without some
> tweaking by an expert.
>
> -- help bot
Yesterday there was a bug which was making it mplay badly. Today that
bug was spotted and removed.
Bye
Sanny
Play Chess at: http://www.GetClub.com/Chess.html
|
| |
Date: 09 Nov 2007 04:18:41
From: help bot
Subject: Re: GetClub vs. Jester games faked?
|
On Nov 9, 12:47 am, Sanny <[email protected] > wrote:
> I played the games between Jester & Easy Level. Play a few games with
> Easy Level they will play as good as Jester. I played with default
> settings in Jester Vs Easy Level.
For the record, at default levels Jester plays a move
in just three seconds.
> I am seeing that Zebediah yesterday played 2-3 games with Beginner
> Level. He managed to beat Beginner level in just 13 moves. While
> others take 50-60 Moves to beat the Beginner Level.
I just replayed that game. While it is possible that this
was a well-known book trap, it looked a lot like a decent
chess program which got lucky when GetClub refused
to defend its King. Generally speaking, all other chess
programs value the King, and they use tactical search
extensions whenever there is a check or capture to
avoid this sort of fiasco. BTW, if it was a well-known
book trap, your team should be congratulated for
getting the program to find all the book moves up to
where it went astray -- that's quite an improvement.
Also note that Zeb played exactly three games vs.
the Beginner level; probably this was because your
site claims you get one-third of a point for each win.
Beating the Beginner level three times takes far less
time than a single game against the Master level, or
even the Normal level since it plays much tougher
chess.
> The reasion you loose with jester is that it has very small board to
> analyze while GetClub Board is quite Big to analyze.
While it is true I don't like the graphic chess board
and that the dark pieces on a dark backgrround are
difficult to see, I believe I lost because Jester played
very much like Fritz -- good, solid moves. How it
manages to play poorly in your tests escapes me,
but I could simply conduct my own tests and report
the results here. Want to bet what the outcome
will be? : >D
> Further GetClub
> shows what move it is thinking So you get an idea what it is going to
> play.
GetClub displays "thinking on Queen" or something
to that effect; that is not the same as revealing the
move it intends to play! There could be ten legal
moves for that piece, or it could simply not be up to
date.
Generally speaking, except against the Beginner
level, I switch to another "tab" instead of dying of
boredom watching the chess screen while the
program thinks.
> Thinking on Pawn, Thinking on Queen ... And you know what is going in
> GetClub Mind. So you get more time to descide your move.
I will take that as indicating that what is on the
display is the best move found so far; I was not
aware of that. Many chess programs will display
the move they are *currently analyzing*, even if it
is believed to be a titanic blunder.
> Now Easy Level at GetClub plays as good as Jester. And in 4 Matches I
> feel It will give 2-2 Score.
Few will believe your test results, since you were
reporting dramatic improvements which no one else
could seem to "see", seemingly for decades before
the noticeable improvements came along. Now many
of the meanies or critics have disappeared -- just when
your cries of "wolf!" are becoming genuine. I can feel
the strength improving, but the entire village has turned
a deaf ear.
> My Computer is P-IV 2.5 Ghz 512 MB RAM. So it is Quite fast.
How do you know whether the limited resources
are fairly apportioned between the two Java applets?
It seems to me that your reported games show
Jester playing much weaker chess than what I
experienced, though I gave it a bit more time per
move. Perhaps you need to utilize two computers,
or perhaps this is not even possible without some
tweaking by an expert.
-- help bot
|
| |
Date: 08 Nov 2007 21:49:40
From: Sanny
Subject: Re: GetClub vs. Jester games faked?
|
> Let's have another go at guesstimating ratings:
>
> Beginner: 1600
>
> Easy: 1700
>
> Normal: 1800
>
> Master: 1900
>
> Advance: 2000
Here is what I think.
Beginner: 1700
Easy: 1800
Normal: 1900
Master: 2000
Advance: 2100
Recently the game was improved now Higher Levels will play much
stronger Games.
Bye
Sanny
Play Chess at: http://www.GetClub.com/Chess.html
|
| | |
Date: 09 Nov 2007 12:12:40
From: Thomas T. Veldhouse
Subject: Re: GetClub vs. Jester games faked?
|
In rec.games.chess.computer Sanny <[email protected] > wrote:
> Here is what I think.
>
> Beginner: 1700
> Easy: 1800
> Normal: 1900
> Master: 2000
> Advance: 2100
>
> Recently the game was improved now Higher Levels will play much
> stronger Games.
>
So Sanny ... you didn't address the subject of the thread. Did you fake the
match?
--
Thomas T. Veldhouse
When the ax entered the forest, the trees said, "The handle is one of us!"
-- Turkish proverb
|
| |
Date: 08 Nov 2007 21:47:09
From: Sanny
Subject: Re: GetClub vs. Jester games faked?
|
I played the games between Jester & Easy Level. Play a few games with
Easy Level they will play as good as Jester. I played with default
settings in Jester Vs Easy Level.
Play Chess at: http://www.GetClub.com/Chess.html
I am seeing that Zebediah yesterday played 2-3 games with Beginner
Level. He managed to beat Beginner level in just 13 moves. While
others take 50-60 Moves to beat the Beginner Level.
The reasion you loose with jester is that it has very small board to
analyze while GetClub Board is quite Big to analyze. Further GetClub
shows what move it is thinking So you get an idea what it is going to
play.
Thinking on Pawn, Thinking on Queen ... And you know what is going in
GetClub Mind. So you get more time to descide your move.
Now Easy Level at GetClub plays as good as Jester. And in 4 Matches I
feel It will give 2-2 Score.
My Computer is P-IV 2.5 Ghz 512 MB RAM. So it is Quite fast.
Bye
Sanny
Play Chess at: http://www.GetClub.com/Chess.html
|
| | |
Date: 09 Nov 2007 12:14:31
From: Thomas T. Veldhouse
Subject: Re: GetClub vs. Jester games faked?
|
In rec.games.chess.computer Sanny <[email protected] > wrote:
>
> My Computer is P-IV 2.5 Ghz 512 MB RAM. So it is Quite fast.
>
2.4GHz ... and old ... you need more memory to be running Java chess :-)
--
Thomas T. Veldhouse
When the ax entered the forest, the trees said, "The handle is one of us!"
-- Turkish proverb
|
| |
Date: 08 Nov 2007 01:57:07
From: help bot
Subject: Re: GetClub vs. Jester games faked?
|
On Nov 7, 12:17 pm, SBD <[email protected] > wrote:
> Well, incompetence on Sanny's end (re: the allocation of resources) is
> a very likely explanation, to be sure. Anyone who, for a year, works
> on a chess program that can't play as well as one I programmed in the
> 1980s in BASIC is surely not too good at programming.
>
> But the combination of incompetence + incompetent fakery seems just as
> likely.
Surely you are not suggesting that your program, running
on a computer from the 1980s, would defeat Sanny's program,
running on a modern computer? This seems impossible,
even with the knowledge that his Java is oh-so-slow.
Maybe you want a fair match, where both programs run
on modern computers (GetClub's program would cause
older computers to explode in disgust!)? As I said, except
against Beginner level, the GetClub program has improved
considerably. Note well that former critics are no longer
testing their mettle, so what you have is a few recent
victories (or losses) which I believe represent the Beginner
level, which tries to move "quickly". The higher levels are
much tougher.
Let's have another go at guesstimating ratings:
Beginner: 1600
Easy: 1700
Normal: 1800
Master: 1900
Advance: 2000
Granted, these higher levels are dead slow, but I
doubt I have overestimated. Only at the very end,
when the horrific problem of King inactivity surfaces,
does the program's strength fall off a cliff. I'm pretty
good at surviving tactics, and getting to that point
so I can rack up yet another "easy" win, but many
players are not. : >D
-- help bot
|
| |
Date: 07 Nov 2007 17:17:39
From: SBD
Subject: Re: GetClub vs. Jester games faked?
|
Well, incompetence on Sanny's end (re: the allocation of resources) is
a very likely explanation, to be sure. Anyone who, for a year, works
on a chess program that can't play as well as one I programmed in the
1980s in BASIC is surely not too good at programming.
But the combination of incompetence + incompetent fakery seems just as
likely.
|
| |
Date: 07 Nov 2007 09:06:34
From: help bot
Subject: Re: GetClub vs. Jester games faked?
|
On Nov 7, 10:57 am, SBD <[email protected] > wrote:
> It took you that long to figure out?
Well, if you check the archives, I made a post very similar
to that one /a long time ago/.
It is possible that Sanny is trying to run both Java applets
on a single computer, and somehow or other it has allocated
most of the resources to the GetClub program, or for whatever
reason it is not running normally on his end.
My experience has been that Jester runs quick and strong,
but that does not mean it works that way for everyone. For
instance, I seem to be the only one complaining about the
*frequent* disconnects at GetClub, despite numerous other
posters here who will complain about everything imaginable.
Obviously, if Sanny never responds to my questions then
people will assume the worst (like they haven't already!).
But this at least gives him another opportunity to address
the issue (or stop posting those "fake" games).
I have been having a bit more trouble lately with the Normal
level at GetClub, and I believe this is because it is playing
stronger than before. Against Beginner level however, I still
win quite easily. It might be interesting to compare the
actual results of Sanny's chess program to what naysayers
predicted a year or so ago; I believe the program has either
been switched for a /real/ chess program, or else it has
improved dramatically. (Yet if there had been a switch, how
do you explain my beating it on any level?)
-- help bot
|
| |
Date: 07 Nov 2007 15:57:28
From: SBD
Subject: Re: GetClub vs. Jester games faked?
|
It took you that long to figure out?
|
|
