Kasparov-Deep Thought 1997 (second game)

Discussion of anything and everything relating to chess playing software and machines.

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
rabbits23
Posts: 144
Joined: Fri Sep 12, 2014 2:57 am
Location: Randwick Australia

Re: Kasparov-Deep Thought 1997 (second game)

Post by rabbits23 » Mon Jul 24, 2017 8:45 am

Possibly Rodolfo; but I didn't see it that way.
rabbits

Rodolfo Leoni
Posts: 544
Joined: Tue Jun 06, 2017 2:49 pm
Location: Italy

Re: Kasparov-Deep Thought 1997 (second game)

Post by Rodolfo Leoni » Mon Jul 24, 2017 9:01 am

rabbits23 wrote:Possibly Rodolfo; but I didn't see it that way.
rabbits
I apologize, then.

BTW, I let read those two phrases to 2 friends of mine, women. They had a lot of fun and they asked me: "Are you saying to be not complicated too?" And all of us laugh a lot about it...

BTW (again..) Some years ago, when playing a game, a friend of mine (man) said me: "I play chess because it's the search for God." His sense of humour was clear and I think it's not an offense to religion. Maybe not really the search for God, but I think a "search for truth" is a good definition. At least, for most players.
F.S.I. Chess Teacher

rabbits23
Posts: 144
Joined: Fri Sep 12, 2014 2:57 am
Location: Randwick Australia

Re: Kasparov-Deep Thought 1997 (second game)

Post by rabbits23 » Mon Jul 24, 2017 10:18 am

Well it's good to learn that you aren't sexist Rodolfo and don't take things too seriously. The world needs more people like you.
rabbits :D

Rodolfo Leoni
Posts: 544
Joined: Tue Jun 06, 2017 2:49 pm
Location: Italy

Re: Kasparov-Deep Thought 1997 (second game)

Post by Rodolfo Leoni » Mon Jul 24, 2017 10:32 am

rabbits23 wrote:Well it's good to learn that you aren't sexist Rodolfo and don't take things too seriously. The world needs more people like you.
rabbits :D
Thanks. :)

Experience should teach everybody that any discrimination, any fascism, any racism or any other way to build differences are offences to nature.
F.S.I. Chess Teacher

Lyudmil Tsvetkov
Posts: 6052
Joined: Tue Jun 12, 2012 10:41 am

Re: Kasparov-Deep Thought 1997 (second game)

Post by Lyudmil Tsvetkov » Mon Jul 24, 2017 10:42 am

MikeGL wrote:
Lyudmil Tsvetkov wrote:
MikeGL wrote:
Leto wrote:So in conclusion Kasparov is wrong in his new book, black is not lost after 45.Ra6 and therefore Deep Blue 2.0 almost allowed Kasparov to draw with the awful 45.Ra6.
Agreed. Almost all strong branches including their leaf were analyzed to death. No
win for white after Ra6. SF8 + powerful hardware by posters on this thread
was thrown at this position and nothing comes up. Besides it was proven in
previous pages that SF8 was stronger than DB2, if even SF8 can't find
a win, how much more for the weaker Deep Blue II?

sorry, I forgot the engine of Lyudmil got +5.00 score against Leoni and +10.00 against Ignacio but those games were a clear draw, and even Lyudmil didn't apologize. Very long analysis gives wrong score, how much more garbage do we get for those 1 minute shootout?
the only garbage are your tongue-in-cheek comments.

1 minute games are statistically significant, do you understand that, when a larger number is played?

what draw are you talking man, what draw?
take only my Qg4+ line;

after Qg4, hg4 wins for white in all lines
if Kf7 instead, Rc7 Rc7 dc7 wins for white
black probably holds only after Qg4 hg4 Ke6 Rd7 Kd7, by a thread and a miracle, but, take only even that very simple pawn endgame, take the different possible lines, and only 1 out of 5 or 6 draws for black, the other are all lost.

is that what you call an easy draw?
really?

for me, this position assessment is big white winning chances, or +-

take even Rodolfo line, Kd6 loses
Rd6 instead continues the game, but, again,
white wins 4/5 of all available lines there too.

is that what you call a drawn game?
really?

so that, when white wins in 90% of all cases, you can not possibly speak of an easy draw.

you might be certain Kasparov would have lost that game one way or another, as he even does not see the very engine-like Re8 reply early into the line.

so, your claim would be Kasparov, who analysed this position very thouroughly for his recent book, and played the game too, has an inferior assessment to Nunn, who did not play the game, analysed it some 10 years ago with much inferior software, and is a much weaker player than Kasparov?

really?
That's why I don't reply to most of your posts because
those were nonsense.
after Qg4, hg4 wins for white in all lines
Maybe you need to check at page 6 where Ignacio Santos played out your beloved Qg4+ line which just ends up in draw, you even claimed it was a +10.00 for white, when the draw was clarified by Ignacio, and your +10.00 just made out of thin air with no proof whatsoever.
black probably holds only after Qg4 hg4 Ke6 Rd7 Kd7, by a thread and a miracle, but, take only even that very simple pawn endgame, take the different possible lines, and only 1 out of 5 or 6 draws for black, the other are all lost.
Sorry but did the rules of chess change recently? where a multi-pv is required for a proof of a win. If the draw is a single line compared to 6 other lines which loses, that doesn't mean the other side lost. Take the below position as a sample. In your definition of a chess win, White is lost and black wins because the pv shows a single line only which draws. The draw is a single thread of a miracle while all the other line wins for black, so it is clearly winning for black? What kind of rubbish is this you are claiming?

[d]3qrrk1/pbpppp2/1p6/7Q/8/8/5nn1/K7 w - - 0 1

Code: Select all

0.00	1.Qg5+ Kh8 2.Qh6+ Kg8 3.Qg5+				
-27.81	1.Qh2 f5 2.Qh5 Rf6 3.Kb1 e5 4.Qe2 Ng4 5.Qa2+ Kh7 6.Qb2 N2e3 7.Qc1 Be4+ 8.Kb2				
-28.7	1.Qe5 Ne4 2.Qh2 Ng5 3.Qe5 f5 4.Qb2 e6 5.Qc1 Qf6+ 6.Kb1 Be4+ 7.Ka2 Bd5+ 8.Kb1 f4 9.Qxc7 Ne3 10.Qxa7				
-61.43	1.Qf5 e5 2.Qxf2 Qh4 3.Qc2 Qd4+ 4.Qb2 Qxb2+ 5.Kxb2 Kg7 6.Kc3 e4 7.Kd2 Kf6				
-61.43	1.Qe2 e5 2.Qxf2 Qh4 3.Qc2 Qd4+ 4.Qb2 Qxb2+ 5.Kxb2 Kg7 6.Kc3 e4 7.Kd2 Kf6				
-63.56	1.Qb5 e5 2.Ka2 Qh4 3.Qxd7 Qc4+ 4.Ka3 Qc5+ 5.Kb2 Qd4+ 6.Qxd4 exd4 7.Ka3 Ne3 8.Ka4				
-63.56	1.Ka2 e5 2.Qf5 Qh4 3.Qxd7 Qc4+ 4.Ka3 Qc5+ 5.Ka2 Qd5+ 6.Qxd5 Bxd5+ 7.Ka3 Nd3 8.Ka4 Bc6+ 9.Ka3				
-63.64	1.Qh6 e5 2.Qd2 Qh4 3.Qxd7 Qd4+ 4.Qxd4 exd4 5.Kb2 Kg7 6.Kb1 Kf6 7.Kc1				
-64.44	1.Kb1 e5 2.Qe2 Qh4 3.Qb5 Qe4+ 4.Kc1 Qe1+ 5.Kb2 Qd2+ 6.Kb3 Qd5+ 7.Qxd5 Bxd5+ 8.Kc3 f5 9.Kd2				
-64.44	1.Kb2 e5 2.Qf5 Ne3 3.Qh5 Qe7 4.Kb3 Bd5+ 5.Kc3 Qa3+ 6.Kd2 Qb4+ 7.Ke2 Qg4+ 8.Qxg4+ Nexg4 9.Kd2				
This position is a black win according to Lyudmil's chess knowledge because only by a thread can white hold a draw, and all other remaining lines would win for black. This is why i didn't reply on this nonsense post of yours at page 6, now we are at page 14.
so, your claim would be Kasparov, who analysed this position very thouroughly for his recent book, and played the game too, has an inferior assessment to Nunn
I didn't claim this, all I know for sure is that the people posting on this forum have more modern, more powerful hardware than Kasparov or his publisher or publishers clerks. If Kasparov can confirm he has more than 7 people with powerful hardware who analyzed this position thoroughly then I might change my opinion.

regards
you are funny.

Kasparov is a millionaire, of course he has access to all the most powerful hardware in the world there is.

forum members here have just some random 16 cores, couple of 32 or 64 core machines.
so, nothing special.

the line you post is quite simple and meaningless.

I was referring to winning chances, which are exhibited of course only by multi-pv choices at different plies.

that is the stabdard definition for winning chances, and white has over 90% winning chances on the main diagram after Ra6.

we still have an awful lot of moves missed/not analysed, everyone, engines and top humans skip some moves at different plies, too many moves to consider otherwise, among those pruned there might be some sensible moves that actually change the theoretical game outcome, but that we all have skipped/pruned, we are and our engines are simply too weak to know that.

btw., any hardware does not help a lot in an ocean of lines, 4 or 64 cores pale down before the 2^16 or so moves we have even in the average middlegame position to a modest depth, and in such a case I prefer to stick to refined eval which considerably favours white.

Lyudmil Tsvetkov
Posts: 6052
Joined: Tue Jun 12, 2012 10:41 am

Re: Kasparov-Deep Thought 1997 (second game)

Post by Lyudmil Tsvetkov » Mon Jul 24, 2017 10:49 am

Guenther wrote:
Very long analysis gives wrong score, how much more garbage do we get for those 1 minute shootout?
Lyudmil Tsvetkov wrote:
1 minute games are statistically significant, do you understand that, when a larger number is played?
You really did understand nothing and even perverted it to the worst.

You mix up statistical significance for producing engine ratings with analysing positions?? (and not the first time)
That is the biggest joke ever made here. A position can only have 3 results and no shootout
even at 1 hour games will slightly resemble an analysis.
It seems you simply don't know what an analysis is and how it can be acchieved.

This makes all threads about analysis with your participation an unpleasant tragicomedy.

EOF forever...
as we don't have the option to conduct a full perfect analysis(100 billion positions or so at least have to be investigated for that), we are doing what we can, assess chances by approximation, which might not be perfect, but still mostly works.

all chess engines do that while testing.

which language, C or C++ used eof() instead of EOF?

Lyudmil Tsvetkov
Posts: 6052
Joined: Tue Jun 12, 2012 10:41 am

Re: Kasparov-Deep Thought 1997 (second game)

Post by Lyudmil Tsvetkov » Mon Jul 24, 2017 10:54 am

Rodolfo Leoni wrote:
Lyudmil Tsvetkov wrote: take even Rodolfo line, Kd6 loses
Rd6 instead continues the game, but, again,
white wins 4/5 of all available lines there too.

is that what you call a drawn game?
really?
Here, only first line is true. It's not nice to see own words perverted in order to be right. Better if you use your own arguments. A war just ended and I'm not willing to start another for some silly reasons.
what do you want me to do now: not to mention your name?

Rodolfo Leoni
Posts: 544
Joined: Tue Jun 06, 2017 2:49 pm
Location: Italy

Re: Kasparov-Deep Thought 1997 (second game)

Post by Rodolfo Leoni » Mon Jul 24, 2017 10:57 am

Lyudmil Tsvetkov wrote:
Rodolfo Leoni wrote:
Lyudmil Tsvetkov wrote: take even Rodolfo line, Kd6 loses
Rd6 instead continues the game, but, again,
white wins 4/5 of all available lines there too.

is that what you call a drawn game?
really?
Here, only first line is true. It's not nice to see own words perverted in order to be right. Better if you use your own arguments. A war just ended and I'm not willing to start another for some silly reasons.
what do you want me to do now: not to mention your name?
You can do what you want, except than altering my words. Agreed?
F.S.I. Chess Teacher

User avatar
Leto
Posts: 2028
Joined: Thu May 04, 2006 1:40 am
Location: Dune

Re: Kasparov-Deep Thought 1997 (second game)

Post by Leto » Mon Jul 24, 2017 1:51 pm

Lyudmil Tsvetkov wrote:
MikeGL wrote:
Leto wrote:So in conclusion Kasparov is wrong in his new book, black is not lost after 45.Ra6 and therefore Deep Blue 2.0 almost allowed Kasparov to draw with the awful 45.Ra6.
Agreed. Almost all strong branches including their leaf were analyzed to death. No
win for white after Ra6. SF8 + powerful hardware by posters on this thread
was thrown at this position and nothing comes up. Besides it was proven in
previous pages that SF8 was stronger than DB2, if even SF8 can't find
a win, how much more for the weaker Deep Blue II?

sorry, I forgot the engine of Lyudmil got +5.00 score against Leoni and +10.00 against Ignacio but those games were a clear draw, and even Lyudmil didn't apologize. Very long analysis gives wrong score, how much more garbage do we get for those 1 minute shootout?
the only garbage are your tongue-in-cheek comments.

1 minute games are statistically significant, do you understand that, when a larger number is played?

what draw are you talking man, what draw?
take only my Qg4+ line;

after Qg4, hg4 wins for white in all lines
if Kf7 instead, Rc7 Rc7 dc7 wins for white
black probably holds only after Qg4 hg4 Ke6 Rd7 Kd7, by a thread and a miracle, but, take only even that very simple pawn endgame, take the different possible lines, and only 1 out of 5 or 6 draws for black, the other are all lost.

is that what you call an easy draw?
really?

for me, this position assessment is big white winning chances, or +-

take even Rodolfo line, Kd6 loses
Rd6 instead continues the game, but, again,
white wins 4/5 of all available lines there too.

is that what you call a drawn game?
really?

so that, when white wins in 90% of all cases, you can not possibly speak of an easy draw.

you might be certain Kasparov would have lost that game one way or another, as he even does not see the very engine-like Re8 reply early into the line.

so, your claim would be Kasparov, who analysed this position very thouroughly for his recent book, and played the game too, has an inferior assessment to Nunn, who did not play the game, analysed it some 10 years ago with much inferior software, and is a much weaker player than Kasparov?

really?
Can you post Kasparov's analysis of this position (45.Ra6) from his new book?

Lyudmil Tsvetkov
Posts: 6052
Joined: Tue Jun 12, 2012 10:41 am

Re: Kasparov-Deep Thought 1997 (second game)

Post by Lyudmil Tsvetkov » Mon Jul 24, 2017 2:50 pm

Leto wrote:
Lyudmil Tsvetkov wrote:
MikeGL wrote:
Leto wrote:So in conclusion Kasparov is wrong in his new book, black is not lost after 45.Ra6 and therefore Deep Blue 2.0 almost allowed Kasparov to draw with the awful 45.Ra6.
Agreed. Almost all strong branches including their leaf were analyzed to death. No
win for white after Ra6. SF8 + powerful hardware by posters on this thread
was thrown at this position and nothing comes up. Besides it was proven in
previous pages that SF8 was stronger than DB2, if even SF8 can't find
a win, how much more for the weaker Deep Blue II?

sorry, I forgot the engine of Lyudmil got +5.00 score against Leoni and +10.00 against Ignacio but those games were a clear draw, and even Lyudmil didn't apologize. Very long analysis gives wrong score, how much more garbage do we get for those 1 minute shootout?
the only garbage are your tongue-in-cheek comments.

1 minute games are statistically significant, do you understand that, when a larger number is played?

what draw are you talking man, what draw?
take only my Qg4+ line;

after Qg4, hg4 wins for white in all lines
if Kf7 instead, Rc7 Rc7 dc7 wins for white
black probably holds only after Qg4 hg4 Ke6 Rd7 Kd7, by a thread and a miracle, but, take only even that very simple pawn endgame, take the different possible lines, and only 1 out of 5 or 6 draws for black, the other are all lost.

is that what you call an easy draw?
really?

for me, this position assessment is big white winning chances, or +-

take even Rodolfo line, Kd6 loses
Rd6 instead continues the game, but, again,
white wins 4/5 of all available lines there too.

is that what you call a drawn game?
really?

so that, when white wins in 90% of all cases, you can not possibly speak of an easy draw.

you might be certain Kasparov would have lost that game one way or another, as he even does not see the very engine-like Re8 reply early into the line.

so, your claim would be Kasparov, who analysed this position very thouroughly for his recent book, and played the game too, has an inferior assessment to Nunn, who did not play the game, analysed it some 10 years ago with much inferior software, and is a much weaker player than Kasparov?

really?
Can you post Kasparov's analysis of this position (45.Ra6) from his new book?
after I buy it, definitely, but still have not done so.

maybe someone knows more.

Post Reply