Computerchess Misc (3)
Rolf Tueschen
Kc5 Rg4 Bh3 Pb2
Ka4 Qe2 Ra3 Pa5 b3 c4 d4 e4 f4 h4 h5
W to move and win.
Sure it looks artificial but it's still good for analysis.
Different to my Misc1 I don't want to make a secret with the task and the
key move.
The original solution wins with 1.Rg1.
Garry Kasparov allegedly proposed 1.Rg5 winning.
My questions:
1. Can you find the winning line after 1.Rg1?
2. Can you find the winning line after 1.Rg5?
I crosspost into 'analysis' too because perhaps someone without a
computerchess program might find interesting cookings ...
See you later!
Rolf Tueschen
Tim Eberly
Rolf Tueschen wrote in message <6tk0ms$4pi$1...@news02.btx.dtag.de>...
>This is a beautiful position:
>
>Kc5 Rg4 Bh3 Pb2
>Ka4 Qe2 Ra3 Pa5 b3 c4 d4 e4 f4 h4 h5
>
>W to move and win.
>
>Sure it looks artificial but it's still good for analysis.
>
>Different to my Misc1 I don't want to make a secret with the task and the
>key move.
>
>The original solution wins with 1.Rg1.
>
>Garry Kasparov allegedly proposed 1.Rg5 winning.
>
>
>
>My questions:
>
>1. Can you find the winning line after 1.Rg1?
>2. Can you find the winning line after 1.Rg5?
How about 1.Rg5 Qg4 2.Rxg4 hxg4 3.Bxg4 c3 Bd7#
and
1.Rg1 Qg4 2.Rxg4 hxg4 3.Bxg4 c3 Bd7#
??
Tim
Rolf Tueschen
>How about 1.Rg5 Qg4 2.Rxg4 hxg4 3.Bxg4 c3 Bd7#
>and
>1.Rg1 Qg4 2.Rxg4 hxg4 3.Bxg4 c3 Bd7#
Exactly. You reveiled the basic idea of the position. mate on d7.
What do you play if I don't take your Rook on g4 in both variations? The
idea for Black is making a path in direction of freedom for the caught
King...
Only idea I see is beaming the Pb2 into pieces. But I think you're very
close to the solution. Let the position work in your mind for some hours.
Then you will enjoy the big bang suddenly. And I promise you will enjoy it.
Thanks for your invoice so far.
Have fun!
ilias kastanas 08-14-90
Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote:
@"Tim Eberly" <chek...@email.msn.com> wrote:
@
@
@>How about 1.Rg5 Qg4 2.Rxg4 hxg4 3.Bxg4 c3 Bd7#
@>and
@>1.Rg1 Qg4 2.Rxg4 hxg4 3.Bxg4 c3 Bd7#
@
@Exactly. You reveiled the basic idea of the position. mate on d7.
@
@What do you play if I don't take your Rook on g4 in both variations? The
@idea for Black is making a path in direction of freedom for the caught
@King...
Right; 2... Ra2 and 3... R:b2 refutes 2. R:g4.
Necessarily so, too, if this is a composition; there cannot be
a double solution!
SPOILER:
1. Rg1, Qg4 2. B:g4, h:g4 3. Rc1, c3 4. Rd1, d3 5. Re1, e3
6. Rf1, f3 7. Rg1, g3 8. Rh1, h3 9. R:h3, ~ 10. Rh4 #
Ilias
Rolf Tueschen
>In article <6tkjhq$bdr$1...@news00.btx.dtag.de>,
>Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote:
>@"Tim Eberly" <chek...@email.msn.com> wrote:
>@
>@
>@>How about 1.Rg5 Qg4 2.Rxg4 hxg4 3.Bxg4 c3 Bd7#
>@>and
>@>1.Rg1 Qg4 2.Rxg4 hxg4 3.Bxg4 c3 Bd7#
>@
>@Exactly. You reveiled the basic idea of the position. mate on d7.
>@
>@What do you play if I don't take your Rook on g4 in both variations? The
>@idea for Black is making a path in direction of freedom for the caught
>@King...
> Right; 2... Ra2 and 3... R:b2 refutes 2. R:g4.
> Necessarily so, too, if this is a composition; there cannot be
>a double solution!
What does it mean this "cannot"?
Also this "never".
Many compositions had nice cookings in the past.
> SPOILER:
> 1. Rg1, Qg4 2. B:g4, h:g4 3. Rc1, c3 4. Rd1, d3 5. Re1, e3
>6. Rf1, f3 7. Rg1, g3 8. Rh1, h3 9. R:h3, ~ 10. Rh4 #
Ah, Ilias you ruined my plans. Of course this is the original solution. But
I had wished more answers. You know, it's not that your contribution is
bad. I like it and would invite you to look for all the planned Misc. *But*
try to be more like Socrates. Let's ask questions without giving the final
answers. It's better if others found them.
I already told you that we are very lazy in general. And I could give the
example of your profound analysis in the Short-Adams game number two was
it? Butwhen I saw the many lines I also thought, oh boy, first I had to
enter all that, then analyse it ... and if the guy was right overall? What
else could I do with my time? - - - You see, it's always the wrapping.
Ok, but what you found here is NOT what I had in mind. Simply because I
published it for the computerchess freaks in mind.
Let me ask the question:
Could a computer or even *you* prove that also 1.Rg5 wins? :)
As I told you it was Kasparov who pointed out that move.
So my question remains. Is it a position with two solutions??
>
> Ilias
ilias kastanas 08-14-90
Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote:
>ika...@sol.uucp (ilias kastanas 08-14-90) wrote:
>
>>In article <6tkjhq$bdr$1...@news00.btx.dtag.de>,
>>Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote:
>>@"Tim Eberly" <chek...@email.msn.com> wrote:
>>@
>>@>How about 1.Rg5 Qg4 2.Rxg4 hxg4 3.Bxg4 c3 Bd7#
>>@>and
>>@>1.Rg1 Qg4 2.Rxg4 hxg4 3.Bxg4 c3 Bd7#
>>@
>>@Exactly. You reveiled the basic idea of the position. mate on d7.
>>@
>>@What do you play if I don't take your Rook on g4 in both variations? The
>>@idea for Black is making a path in direction of freedom for the caught
>>@King...
>> Right; 2... Ra2 and 3... R:b2 refutes 2. R:g4.
>
>> Necessarily so, too, if this is a composition; there cannot be
>>a double solution!
>What does it mean this "cannot"?
>Also this "never".
>Many compositions had nice cookings in the past.
Sure; and then they are not compositions any more!...
>> SPOILER:
>
>> 1. Rg1, Qg4 2. B:g4, h:g4 3. Rc1, c3 4. Rd1, d3 5. Re1, e3
>>6. Rf1, f3 7. Rg1, g3 8. Rh1, h3 9. R:h3, ~ 10. Rh4 #
>
>Ah, Ilias you ruined my plans. Of course this is the original solution. But
>I had wished more answers. You know, it's not that your contribution is
>bad. I like it and would invite you to look for all the planned Misc. *But*
>try to be more like Socrates. Let's ask questions without giving the final
>answers. It's better if others found them.
Sorry... But I did say "spoiler"!
It's no simple thing to be like Socrates in patience and persistense...
If we believe Alcibiades (well, quoted by Aristodemus, quoted by Phoenix,
quoted by Apollodorus in Plato's "Symposium"), one morning during the Potidaea
expedition Socrates stopped dead in his tracks, to think about something...
He stood there, motionless, throughout the day, through the night, until next
morning...
OK, "others"... "-- Oh great guru, what are we here for?" "-- To
love the others." "-- Uh... what are the _others_ here for?!"
>I already told you that we are very lazy in general. And I could give the
>example of your profound analysis in the Short-Adams game number two was
>it? Butwhen I saw the many lines I also thought, oh boy, first I had to
>enter all that, then analyse it ... and if the guy was right overall? What
>else could I do with my time? - - - You see, it's always the wrapping.
>
>Ok, but what you found here is NOT what I had in mind. Simply because I
>published it for the computerchess freaks in mind.
>
>Let me ask the question:
>
>Could a computer or even *you* prove that also 1.Rg5 wins? :)
Hey, I'm lazy myself.
>As I told you it was Kasparov who pointed out that move.
Did you say "Kasparian"?!... Oh, "Kasparov"...
>So my question remains. Is it a position with two solutions??
After a brief look, I don't quite fathom what lies behind 1. Rg5,
Qg4. If 2. B:g4, h:g4 there is no mating mechanism in Kb6 or K:c4;
or any other trick. If 2. R:g4, Ra2 3. Rg5, R:b2 4. Bd7+, or 4. K~
don't work. 2. Kb6, Q:h3 3. R:a5+, Kb4 4. b:a3+ "just" fails...
Hmm, was it GK quoted by X quoted by Y ...?
Ilias
zene...@wwa.com
On 21 Sep 1998, ilias kastanas 08-14-90 wrote:
> After a brief look, I don't quite fathom what lies behind 1. Rg5,
> Qg4. If 2. B:g4, h:g4 there is no mating mechanism in Kb6 or K:c4;
> or any other trick. If 2. R:g4, Ra2 3. Rg5, R:b2 4. Bd7+, or 4. K~
> don't work. 2. Kb6, Q:h3 3. R:a5+, Kb4 4. b:a3+ "just" fails...
> Hmm, was it GK quoted by X quoted by Y ...?
Aha, I think I see it!
1.Rg5 Qg4 2.Bxg4 hxg4 3.Kb6! Now White threatens Rc5 and Rc4#. Black can
only stop it by 3...Ra2 and 4...Rxb2. So 3...Ra2 is forced. But now:
4.Rxa5+ Kb4 5.Rb5+! Ka4 6.Kc5!! Now White threatens 7.Rb8 and 8.Ra8# and
Black is absolutely powerless to stop it!! 6...Rxb2 7.Rb8 Ka3 8.Ra8# and
the rook on b2 blocks the king's escape!
Wow, that's beautiful.
Jeff C.
ilias kastanas 08-14-90
<zene...@wwa.com> wrote:
@
@On 21 Sep 1998, ilias kastanas 08-14-90 wrote:
@
@> After a brief look, I don't quite fathom what lies behind 1. Rg5,
@> Qg4. If 2. B:g4, h:g4 there is no mating mechanism in Kb6 or K:c4;
@> or any other trick. If 2. R:g4, Ra2 3. Rg5, R:b2 4. Bd7+, or 4. K~
@> don't work. 2. Kb6, Q:h3 3. R:a5+, Kb4 4. b:a3+ "just" fails...
@> Hmm, was it GK quoted by X quoted by Y ...?
@
@Aha, I think I see it!
@
@1.Rg5 Qg4 2.Bxg4 hxg4 3.Kb6! Now White threatens Rc5 and Rc4#. Black can
@only stop it by 3...Ra2 and 4...Rxb2. So 3...Ra2 is forced. But now:
@
@4.Rxa5+ Kb4 5.Rb5+! Ka4 6.Kc5!! Now White threatens 7.Rb8 and 8.Ra8# and
@Black is absolutely powerless to stop it!! 6...Rxb2 7.Rb8 Ka3 8.Ra8# and
@the rook on b2 blocks the king's escape!
@
@Wow, that's beautiful.
Very nice idea! Actually, there is no mate on c4; Rc5, Kb4...
The threat is the one you gave: eliminate Pa5 and mate on the "a" file.
So 3... Ra2 isn't forced; Black can defend against R:a5+...Rb5... by
3... c3. But then there _is_ mate... on d4: 3. Rd5! +-
Do we have a cook, then? No!... Black has a more subtle defense:
3... e3! The idea is 4. R:a5+, Kb4 5. Rb5+, Ka4 6. Kc5, e2 7. Rb8?,
e1=Q 8. Ra8+, Qa5 White can inject the finesse 7. Rb4+, Ka5 8. Rb8,
but then 8... Ka6 9. Kc6, Ka7 10. Rb7+, Ka8
White might also try 6. Re5 (and then Kc5); however, 6... Kb4!
foils him.
The augustly recommended 1. Rg5 hasn't proven itself yet. Wh's
Rook is on the wrong side! Yes, Rooks belong behind Passed Pawns... but
_hey_ ...
Ilias
Rolf Tueschen
>On 21 Sep 1998, ilias kastanas 08-14-90 wrote:
>> After a brief look, I don't quite fathom what lies behind 1. Rg5,
>> Qg4. If 2. B:g4, h:g4 there is no mating mechanism in Kb6 or K:c4;
>> or any other trick. If 2. R:g4, Ra2 3. Rg5, R:b2 4. Bd7+, or 4. K~
>> don't work. 2. Kb6, Q:h3 3. R:a5+, Kb4 4. b:a3+ "just" fails...
>> Hmm, was it GK quoted by X quoted by Y ...?
>Aha, I think I see it!
>1.Rg5 Qg4 2.Bxg4 hxg4 3.Kb6! Now White threatens Rc5 and Rc4#. Black can
>only stop it by 3...Ra2 and 4...Rxb2. So 3...Ra2 is forced. But now:
>4.Rxa5+ Kb4 5.Rb5+! Ka4 6.Kc5!! Now White threatens 7.Rb8 and 8.Ra8# and
>Black is absolutely powerless to stop it!! 6...Rxb2 7.Rb8 Ka3 8.Ra8# and
>the rook on b2 blocks the king's escape!
>Wow, that's beautiful.
Jeff, congratulations. You finally found another mating idea.
But to be absolutely sure we have to look if Black cannot choose 'another'
defense. Let me end with the rather cryptic remark. Perhaps it gives you
the next bright flash of an idea ...
Tim had the first, Jeff had the second.
Ilias is the BIG BRAIN who always found the defenses up to now. Ilias
please let's enjoy it for a moment longer ...
Isn't it fun, such a wonderful position? Of course it got first prize and
is world-wide known among the experts of endgame studies. But then Kasparov
came (allegedly to a famous German computerchess newspaper) and showed the
1.Rg5. Question remains if he was absolutely right. As it seems at least
after the line Jeff gave us: "YES".
So the question is, who's right. A genius on endgame studies or a genius of
OTB tournament chess?
I thought it worth for a "short" look.
>Jeff C.
ilias kastanas 08-14-90
Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote:
@<zene...@wwa.com> wrote:
@
@>On 21 Sep 1998, ilias kastanas 08-14-90 wrote:
@
@>> After a brief look, I don't quite fathom what lies behind 1. Rg5,
@>> Qg4. If 2. B:g4, h:g4 there is no mating mechanism in Kb6 or K:c4;
@>> or any other trick. If 2. R:g4, Ra2 3. Rg5, R:b2 4. Bd7+, or 4. K~
@>> don't work. 2. Kb6, Q:h3 3. R:a5+, Kb4 4. b:a3+ "just" fails...
@>> Hmm, was it GK quoted by X quoted by Y ...?
@
@>Aha, I think I see it!
@
@>1.Rg5 Qg4 2.Bxg4 hxg4 3.Kb6! Now White threatens Rc5 and Rc4#. Black can
@>only stop it by 3...Ra2 and 4...Rxb2. So 3...Ra2 is forced. But now:
@
@>4.Rxa5+ Kb4 5.Rb5+! Ka4 6.Kc5!! Now White threatens 7.Rb8 and 8.Ra8# and
@>Black is absolutely powerless to stop it!! 6...Rxb2 7.Rb8 Ka3 8.Ra8# and
@>the rook on b2 blocks the king's escape!
@
@>Wow, that's beautiful.
@
@Jeff, congratulations. You finally found another mating idea.
@
@But to be absolutely sure we have to look if Black cannot choose 'another'
@defense. Let me end with the rather cryptic remark. Perhaps it gives you
@the next bright flash of an idea ...
@
@Tim had the first, Jeff had the second.
@
@Ilias is the BIG BRAIN who always found the defenses up to now. Ilias
@please let's enjoy it for a moment longer ...
I don't know... big, or with a bunch of little squares in it?
@Isn't it fun, such a wonderful position? Of course it got first prize and
@is world-wide known among the experts of endgame studies. But then Kasparov
@came (allegedly to a famous German computerchess newspaper) and showed the
@1.Rg5. Question remains if he was absolutely right. As it seems at least
@after the line Jeff gave us: "YES".
@
@So the question is, who's right. A genius on endgame studies or a genius of
@OTB tournament chess?
@
@I thought it worth for a "short" look.
Ah... so that's what it is. See what happens with secrecy?!...
I gave it a look shorter than it deserved. If you had said something like
"Kubbel", or "Bron", or "Korolkov, Shakhmaty v SSSR, 1940", or "Prokes", or
other random examples, I would have known better. After all, it could
have been just a position from a _game_, right?!
Did GK claim a cook and give analysis for 1. Rg5, or did he just
say it wins, or probably wins? (See my follow-up to Jeff's post).
OK, I hear you... I'll sit back and wait for further developments!
Ilias
zene...@wwa.com
On 22 Sep 1998, ilias kastanas 08-14-90 wrote:
> After all, it could have been just a position from a _game_, right?!
Is this a joke, Ilias? An actual game position where White has sacked a
queen and *seven* pawns for a bishop and a mating net? I can almost
believe the queen, but not the seven pawns.
ilias kastanas 08-14-90
@
@> After all, it could have been just a position from a _game_, right?!
@
@Is this a joke, Ilias? An actual game position where White has sacked a
@queen and *seven* pawns for a bishop and a mating net? I can almost
@believe the queen, but not the seven pawns.
And the Pawns lined up?!...
Only if you enjoy playing "prove-it"... A Queen is worth 10 Pawns?
Prove it. Remove Qd1 and give Black 10 extra Pawns at b6, c6, ... g6, c5,
d5, e5, f5. (Which side would you take?)
Ilias
ilias kastanas 08-14-90
ilias kastanas 08-14-90 <ika...@sol.uucp> wrote:
@In article <Pine.GSO.3.96.98092...@tekka.wwa.com>,
@ <zene...@wwa.com> wrote:
@@
@@On 22 Sep 1998, ilias kastanas 08-14-90 wrote:
@@
@@
@@Is this a joke, Ilias? An actual game position where White has sacked a
@@queen and *seven* pawns for a bishop and a mating net? I can almost
@@believe the queen, but not the seven pawns.
@
@
@
@ And the Pawns lined up?!...
@
@ Only if you enjoy playing "prove-it"... A Queen is worth 10 Pawns?
@Prove it. Remove Qd1 and give Black 10 extra Pawns at b6, c6, ... g6, c5,
^^^
That's "Qd8", of course...
@d5, e5, f5. (Which side would you take?)
@ Ilias
Rolf Tueschen
Worse is the more irregular formations. Or *holes* ...
>@Isn't it fun, such a wonderful position? Of course it got first prize and
>@is world-wide known among the experts of endgame studies. But then Kasparov
>@came (allegedly to a famous German computerchess newspaper) and showed the
>@1.Rg5. Question remains if he was absolutely right. As it seems at least
>@after the line Jeff gave us: "YES".
>@
>@So the question is, who's right. A genius on endgame studies or a genius of
>@OTB tournament chess?
>@
>@I thought it worth for a "short" look.
> Ah... so that's what it is. See what happens with secrecy?!...
>I gave it a look shorter than it deserved. If you had said something like
>"Kubbel", or "Bron", or "Korolkov, Shakhmaty v SSSR, 1940",
special <g>
> or "Prokes", or
>other random examples, I would have known better.
Yes, exactly. That's why I thought it worthwhile to do it this way!
You know what researchers did with the monkeys. No, wrong usage of English,
what science did with human beings when they made experiments the way they
studied the thought process of monkeys.
You know ther's the banana, and there's a stick. Now what will the ape do
with that stick??
I always liked the human experiments where you had a scissors, a chewing
gum and a cord. Ah, I forgot the candal and the matches. Now the task. How
do you get the key of that door to get out of that room ...?
Ilias, it's so damn boring if we always meet the same old stuff. Book
thirteen, chapter 45. Bingo. And next one.
It gives us much more fun to learn the same old stuff again in another
wrapping. Because what do we do without that moment when we could say
"bingo"? Or as a famous US author has said on the question what all that
chess for? Could we do *without* that fresh air? It was Irving Chernev. The
man with the five notebooks in his pockets.
Sure. Shahmaty 1940. Bingo. End of the story. Automatic reaction.
Dubdidubdidub. Drawer closed again.
Kasparov *did* recommand Rg5 because I have it written inthe position from
his buddy's German CSS. *Funny* Kasparov moved the h-pawn. You the e-pawn.
*Close*, I would say in full impostordom. My Crafty played the d-pawn. Now,
what is that? Confusion all over?
[Talking to myself-mode ON]
If I could only offer the deal of a new car or a 655MHz processor from
SONY. We had more reactions from the people. So sad.
Giving such perls to Ilias is as if I wanted to explain to a 12 y. old how
to w a l k. He would be bored too. My dic says something like "not to
stand for" but what does that mean? Not to stand for that offer? :)
Did I therefore forget about the L. Evans position in (1)? :))
d6 draw and Qe7 winning!?
[Talking to myself-mode OFF]
Let's continue to enjoy and have fun.
Rolf:)
After all, it could
>have been just a position from a _game_, right?!
zene...@wwa.com
On 23 Sep 1998, ilias kastanas 08-14-90 wrote:
> Only if you enjoy playing "prove-it"... A Queen is worth 10 Pawns?
> Prove it. Remove Qd8 and give Black 10 extra Pawns at b6, c6, ... g6,
> c5, d5, e5, f5. (Which side would you take?)
Ouch, certainly the pawns in this case. The variant I've heard of is to
remove the White queen and give White 8 extra pawns, all on the third
rank. Apparently with best play the queen should win but among club
players it's easier to play and win with the pawns.
Probably the best test is somewhere in between your variant and this one.
Maybe remove White's queen and give White 8 pawns at b3, c3 ... g3, d4,
e4?
Jeff C.
Noam D. Elkies
and never considered 1 Rg5. Now that I do consider it, though,
it seems that Black can answer 1 Rg5 Qg4 2 B:g4 h:g4 3 Kb6 with
3...e3! and after 4 R:a5+ (or 4 Rc5) Kb4 5 Rb5+ Ka4 White must
take the draw by repetition because after 6 Kc5? e2 and 7...e1Q
the new Queen interposes at a5, leaving Black with more than
enough pawns to finish White off.
--Noam D. Elkies (USCF 2263)
[remove plural from e-address to reply]
ilias kastanas 08-14-90
Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote:
@ika...@sol.uucp (ilias kastanas 08-14-90) wrote:
@
@>In article <6u832o$7he$1...@news00.btx.dtag.de>,
@>Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote:
@>@Tim had the first, Jeff had the second.
@>@
@>@Ilias is the BIG BRAIN who always found the defenses up to now. Ilias
@>@please let's enjoy it for a moment longer ...
@
@
@> I don't know... big, or with a bunch of little squares in it?
@Worse is the more irregular formations. Or *holes* ...
Hmm... or unauthorized possession of a pointed head...
@>@Isn't it fun, such a wonderful position? Of course it got first prize and
@>@is world-wide known among the experts of endgame studies. But then Kasparov
@>@came (allegedly to a famous German computerchess newspaper) and showed the
@>@1.Rg5. Question remains if he was absolutely right. As it seems at least
@>@after the line Jeff gave us: "YES".
@>@
@>@So the question is, who's right. A genius on endgame studies or a genius of
@>@OTB tournament chess?
@>@
@>@I thought it worth for a "short" look.
@
@
@> Ah... so that's what it is. See what happens with secrecy?!...
@>I gave it a look shorter than it deserved. If you had said something like
@>"Kubbel", or "Bron", or "Korolkov, Shakhmaty v SSSR, 1940",
@
@special <g>
@
@> or "Prokes", or other random examples, I would have known better.
@Yes, exactly. That's why I thought it worthwhile to do it this way!
@
@You know what researchers did with the monkeys. No, wrong usage of English,
@what science did with human beings when they made experiments the way they
@studied the thought process of monkeys.
Then they found out a monkey was thinking: "I have conditioned Prof.
Skinner... Every time I press this lever he gives me a banana". That caused
a substantial revision.
@You know ther's the banana, and there's a stick. Now what will the ape do
@with that stick??
@
@I always liked the human experiments where you had a scissors, a chewing
@gum and a cord. Ah, I forgot the candal and the matches. Now the task. How
@do you get the key of that door to get out of that room ...?
You make a ball of candle wax and cord, stick the scissors into it
and chant words of voodoo until an Evil Spirit appears and says "Push the key
off the keyhole with the scissors, onto the chewing gum, and pull it in under
the door with the cord. Now leave me alone, okay?"
@Ilias, it's so damn boring if we always meet the same old stuff. Book
@thirteen, chapter 45. Bingo. And next one.
@
@It gives us much more fun to learn the same old stuff again in another
@wrapping. Because what do we do without that moment when we could say
@"bingo"? Or as a famous US author has said on the question what all that
@chess for? Could we do *without* that fresh air? It was Irving Chernev. The
@man with the five notebooks in his pockets.
@
@Sure. Shahmaty 1940. Bingo. End of the story. Automatic reaction.
@Dubdidubdidub. Drawer closed again.
No, "Korolkov" and "shakhmaty" don't make it a foregone conclusion;
but they do give a measure of what to expect. (Are you using Socratic
irony to draw me out?!).
I saw once in a weekly chess column (next to crosswords etc) the
study Ke6, Be2, Ne7; Ke8, Ne5 (Adamson, 1924), Wh to move and win. The
newspaper's solution was the line
1. Nd5, Nd7 2. Kd6!, Kd8 3. Bd3, Ke8 4. Nc7+, Kd8 5. Ne6+, Kc8 6. Ba6+
-- which is fine for the paper's purposes. Still, there is more to the
study. The symmetric "solution", 1. Nf5, Nf7 2. Kf6... 4. Ng7+... does fail
since Wh cannot play 6. Bi6+ (or else Bl too would reply 6... Kg9 !). Un-
fortunately, 4. Bc6+! (instead of Ng7+) leads to a Wh win as well; 1. Nf5
is a cook! The paper wouldn't care; a chess magazine would.
@Kasparov *did* recommand Rg5 because I have it written inthe position from
@his buddy's German CSS. *Funny* Kasparov moved the h-pawn. You the e-pawn.
@*Close*, I would say in full impostordom. My Crafty played the d-pawn. Now,
@what is that? Confusion all over?
Haven't I told him a dozen times?... But he still confuses "h"
and "e"...
Eh, when GK moves the h Pawn, it's not in response to 3. Kb6.
But you have forbidden me to talk about the subject...
@[Talking to myself-mode ON]
@
@If I could only offer the deal of a new car or a 655MHz processor from
@SONY. We had more reactions from the people. So sad.
Van der Wiel has toyed with related ideas. "I said: "-- $50,000, a
Mercedes, two apartments with women, or do you just want a draw?..." Torre
just wanted a draw."
@Giving such perls to Ilias is as if I wanted to explain to a 12 y. old how
@to w a l k. He would be bored too. My dic says something like "not to
@stand for" but what does that mean? Not to stand for that offer? :)
@Did I therefore forget about the L. Evans position in (1)? :))
@d6 draw and Qe7 winning!?
Yeah, so I didn't offer any prize, nag, nag, nag... All right; as I
don't know your taste in, uh, cars etc, the prize is: no kid-stuff 655MHz
processor, but a Gigabyte of dual-ported, write-only, 0.25 picosecond memory
-- fastest on the planet. Yes, that's "pico".
Right, I remember it's a game by Evans; which I don't have... nor do
we need it! In [Kh2, Qe4, Pd5, g2, h7; Kb7, Qc3, Pa7, b2] he played
1. d6+?, Ka6 2. d7 -- and 2... b1=Q+ 3. Q:b1, Qe5+* was perpetual check!
Its entire analysis is just "conjugate squares", e.g. (K)h2 - (Q)e5, marked on
the board, s-S, t-T, u-U... (s = h2, S = e5); chosen to be 'connected'
and to avoid mishaps like ...Qc7+?. Kh1, or the WQ interposing. (In particular
the WK must not reach a1 !). One possible choice is
h2-e5 h1-h5 g1-c5 f1- (c4 or f8) f2- (c5 or f8) g3-d6 h3-e6
e1- (b4 or e7) e2-e7 d1-d6 d2-b4 c1-f4 c2- (e4 or c5) b2-d4
a3-a4 b3-d5 f3- (c3 or f8) g4-b4 h5-c5 ... and a few more.
(it's crucial that stalemate makes ...Qb4+ and ...Qa4+ available!).
Draw a picture; visually the whole thing is rather clear and motivated.
Building such a list doesn't take too much effort, and provides
a complete blueprint for play; what move to make becomes table lookup.
On the other hand, imagine a usual search algorithm trying to do as much!...
The WK can roam over 25 squares or so. The combinatorial explosion is
absolutely prohibitive.
@[Talking to myself-mode OFF]
@
@Let's continue to enjoy and have fun.
Ilias
Rolf Tueschen
>This is a beautiful position:
>Kc5 Rg4 Bh3 Pb2
>Ka4 Qe2 Ra3 Pa5 b3 c4 d4 e4 f4 h4 h5
>W to move and win.
>Sure it looks artificial but it's still good for analysis.
>Different to my Misc1 I don't want to make a secret with the task and the
>key move.
>The original solution wins with 1.Rg1.
>Garry Kasparov allegedly proposed 1.Rg5 winning.
>My questions:
>1. Can you find the winning line after 1.Rg1?
>2. Can you find the winning line after 1.Rg5?
>I crosspost into 'analysis' too because perhaps someone without a
>computerchess program might find interesting cookings ...
>See you later!
>Rolf Tueschen
Here's the SOLUTION
======================
Primarily I posted the problem into rgcc, our computerchess group. I did so
because I found a little discovery with my favorite computer program FRITZ
(vers. 5 actually).
The whole story is this: I met the position reading a wonderful piece of
chess lit from Arnold Denker. Who was talking about Chernev. *We* found
that point (see below) and I went into my database where I then found the
participation of Garry Kasparov. Exactly in the line that proved Kasparov
*wrong* *we* (that's FRITZ and yours truly!) had "found" a new refutation
of Korolkov's solution ...
But read for yourself:
Korolkov,V - [+3410.18c5a4]
2310.18.c5a4 ECE Chess in USSR, 1940
[Rolf W. Tüschen]
{NB that I produced my own abrevation key for position after I'd read a
book from a British specialist. Later when ChessBase adopted the other key
I had done my work for "nothing".}
1.Rg1
[1.Rg5 G. Kasparov
{That's from the German CSS 1993 where surely F. Friedel presented it;
perhaps someone has the original?}
1...Qg4 2.Bxg4 hxg4 3.Kb6
A) 3...e3 4.Rxa5+ Kb4 5.Rb5+ Ka4 wfr5 {that's my FRITZ engine in CB7} sees
= {a draw} until deep in 16 HZ {that's ply} please note that for a computer
program a position is (can be) completely different if it's *before* the
candidate move made or *after*. For me as a computerchess newbie such facts
are miraculous. I mean chess for human beings is always about a move you
judged as one of the best in a position. And it's then of no importance if
the move is only *seen* or already been made on the board only in our minds
of course. It's of philosophical relevance perhaps if a human player sees a
move preferably if it's not yet been made or if it's already virtually
played.
This position is for FRITZ of a completely different kind than after Kc5 is
outperformed. *Then* FRITZ seems to have forgotten all his crap (see below)
and he sees the Blacks winning.
6.Kc5 after that has been played however, wfr5 sees immediately -+
(6.Ra5+ crafty makes a correct direct draw without even spend precious time
on Kc5.)
6...e2
A1) 7.Rb4+ Ka5 8.Rb8 Ka6 9.Re8 Kb7-+ {of course Kb7 wins but wfr5 saw a
draw before Kc5 because he "saw" only 9...Ra4 with 10.Kc6!=;
9...Ra5+ That's crafty's choice. Somewhat perverse because he thinks that
the position is simpler *without* his Rook. Possibly a new Law of Clearness
by Prof. Hyatt. 10.Kc6 Rc5+ 11.Kxc5 Kb7 12.Re7+ Kc8-+) ;
A2) 7.Rb8 7...e1Q-+;
B) 3...h3? {without knowing the exact text of CSS I must admit that G.
Kasparov surely gave that line only as sort of example for a funny idea.
There's no doubt that he saw all the details good enough to NOT even think
longer than a second that 3-h3 really could solve the position. I think
it's only fair to state that ... More probable is however that "Charly",
the ape or was it a monkey of F. Friedel, played that line and that it was
then presented in a new contest for the readers where they could win a new
chess program. Generally *not* FRITZ.}
4.Rxa5+ Kb4 5.Rb5+ Ka4 6.Kc5 h2 7.Rb8 h1Q 8.Ra8# anyway, that's a beautiful
line, no?]
1...Qg4 2.Bxg4 hxg4 3.Rc1 c3 4.Rd1 d3 5.Re1 e3 6.Rf1 f3 7.Rg1 g3 8.Rh1 h3
9.Rxh3 +-
Well, the last line was already given by Ilias and I read it in the Denker
text for the first time. Also Noam Elkies joined the party after two weeks
of heavy night&day analysis OTB.
I want to thank all for their participation. While I want to explain to the
astonished crowd who omitted to get in touch with such artificial stuff,
that it was only presented to enjoy our hobby for some minutes of fresh
air. In a world full of smog. No more no less.
But such a little chessic position also has a general morals.
We all have our ideas. Some go deeper. But often the momentary deepest idea
could still be corrected by a new one. But this is not the whole truth.
Even if we are personally not capable to go too deep the important step is
to try to find a little depth on your own. Only then you will be able to
enjoy the presented depth of the experts. And you will become part of the
community. Not enough. You get in touch with history.
More than nothing, no?
Rolf Tueschen
P.S.
Ilias or others, how about joining with your own Misc (nm)?! We have plenty
of numbers left ... Ilias, each of your little examples had earned a
seperated number I think. A position and then a little story behind it, ok?
Rolf Tueschen
[I snipped a lot to keep it short.]
[about experiments with man and monkeys.]
> Then they found out a monkey was thinking: "I have conditioned Prof.
>Skinner... Every time I press this lever he gives me a banana". That caused
>a substantial revision.
Depends on what you define as thinking. But I would agree with that. It's a
deep concept. But one of a clear distinction of master and slave. With all
implications. The moment the slave is no longer of interest? Reminds me on
Willie, the whale.
>@I always liked the human experiments where you had a scissors, a chewing
>@gum and a cord. Ah, I forgot the candal and the matches. Now the task. How
>@do you get the key of that door to get out of that room ...?
> You make a ball of candle wax and cord, stick the scissors into it
>and chant words of voodoo until an Evil Spirit appears and says "Push the key
>off the keyhole with the scissors, onto the chewing gum, and pull it in under
>the door with the cord. Now leave me alone, okay?"
I screwed up the whole setting. Because how do you put the key into the
chewing gum? No, the original solution was to create a pendulum with the
scissors as weight ... but I forgot the original experiment. Well, it's
been a long time. The point with such stuff was the breaking well learnt
patterns in your mind before you could find the "solution". I liked that in
psychology and in chess too. Unfortunately in OTB chess you have the
"normal" stuff before you meet the extraordinary. That's why I like endgame
studies and a little less the mate problems. There you have the exception
at the touch of a button.
Since I have my FRITZ I'm back in the arena however. :)
> No, "Korolkov" and "shakhmaty" don't make it a foregone conclusion;
>but they do give a measure of what to expect. (Are you using Socratic
>irony to draw me out?!).
Please give us the original content.
> I saw once in a weekly chess column (next to crosswords etc) the
>study Ke6, Be2, Ne7; Ke8, Ne5 (Adamson, 1924), Wh to move and win. The
>newspaper's solution was the line
>1. Nd5, Nd7 2. Kd6!, Kd8 3. Bd3, Ke8 4. Nc7+, Kd8 5. Ne6+, Kc8 6. Ba6+
>-- which is fine for the paper's purposes. Still, there is more to the
>study. The symmetric "solution", 1. Nf5, Nf7 2. Kf6... 4. Ng7+... does fail
>since Wh cannot play 6. Bi6+ (or else Bl too would reply 6... Kg9 !). Un-
>fortunately, 4. Bc6+! (instead of Ng7+) leads to a Wh win as well; 1. Nf5
>is a cook! The paper wouldn't care; a chess magazine would.
Wait a minute. I suffer from an overflooding of information.
>@Kasparov *did* recommand Rg5 because I have it written inthe position from
>@his buddy's German CSS. *Funny* Kasparov moved the h-pawn. You the e-pawn.
>@*Close*, I would say in full impostordom. My Crafty played the d-pawn. Now,
>@what is that? Confusion all over?
> Haven't I told him a dozen times?... But he still confuses "h"
>and "e"...
> Eh, when GK moves the h Pawn, it's not in response to 3. Kb6.
>But you have forbidden me to talk about the subject...
Please end the torture and tell us the whole story.
>@[Talking to myself-mode ON]
>@
>@If I could only offer the deal of a new car or a 655MHz processor from
>@SONY. We had more reactions from the people. So sad.
> Van der Wiel has toyed with related ideas. "I said: "-- $50,000, a
>Mercedes, two apartments with women, or do you just want a draw?..." Torre
>just wanted a draw."
>@Giving such perls to Ilias is as if I wanted to explain to a 12 y. old how
>@to w a l k. He would be bored too. My dic says something like "not to
>@stand for" but what does that mean? Not to stand for that offer? :)
>@Did I therefore forget about the L. Evans position in (1)? :))
>@d6 draw and Qe7 winning!?
> Yeah, so I didn't offer any prize, nag, nag, nag... All right; as I
>don't know your taste in, uh, cars etc, the prize is: no kid-stuff 655MHz
>processor, but a Gigabyte of dual-ported, write-only, 0.25 picosecond memory
> -- fastest on the planet. Yes, that's "pico".
Pico-bello. As we say it in German. :)
> Right, I remember it's a game by Evans; which I don't have... nor do
>we need it! In [Kh2, Qe4, Pd5, g2, h7; Kb7, Qc3, Pa7, b2] he played
>1. d6+?, Ka6 2. d7 -- and 2... b1=Q+ 3. Q:b1, Qe5+* was perpetual check!
>Its entire analysis is just "conjugate squares", e.g. (K)h2 - (Q)e5, marked on
>the board, s-S, t-T, u-U... (s = h2, S = e5); chosen to be 'connected'
>and to avoid mishaps like ...Qc7+?. Kh1, or the WQ interposing. (In particular
>the WK must not reach a1 !). One possible choice is
>h2-e5 h1-h5 g1-c5 f1- (c4 or f8) f2- (c5 or f8) g3-d6 h3-e6
>e1- (b4 or e7) e2-e7 d1-d6 d2-b4 c1-f4 c2- (e4 or c5) b2-d4
>a3-a4 b3-d5 f3- (c3 or f8) g4-b4 h5-c5 ... and a few more.
Ok, you're my next editor in chief. Let's build up a whole series of MISC
(nm). But with a certain relation to dumb machines too. Hope you're also
familiar with PC machines ... :)
>
>(it's crucial that stalemate makes ...Qb4+ and ...Qa4+ available!).
>Draw a picture; visually the whole thing is rather clear and motivated.
> Building such a list doesn't take too much effort, and provides
>a complete blueprint for play; what move to make becomes table lookup.
>On the other hand, imagine a usual search algorithm trying to do as much!...
>The WK can roam over 25 squares or so. The combinatorial explosion is
>absolutely prohibitive.
>
I will try to get into contact with Hsu and his team to explain that.
Perhaps he will then come back to normal with all the DEEP BLUE hype. He
had lost all respect for the human champ. All IMHO of course.
Noam D. Elkies
Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote:
>TUESCHEN.MEDIZ...@t-online.de (Rolf Tueschen) wrote:
[snipped for brevity --NDE]
>>Kc5 Rg4 Bh3 Pb2
>>Ka4 Qe2 Ra3 Pa5 b3 c4 d4 e4 f4 h4 h5
>>W to move and win.
>>The original solution wins with 1.Rg1.
>>Garry Kasparov allegedly proposed 1.Rg5 winning.
>>My questions:
>>1. Can you find the winning line after 1.Rg1?
>>2. Can you find the winning line after 1.Rg5?
>[1 Rg5?! (Kasparov, allegedly) Qg4 2 B:g4 h:g4 3 Kb6]
>B) 3...h3? {without knowing the exact text of CSS I must admit that G.
>Kasparov surely gave that line only as sort of example for a funny idea.
>[...] 4.Rxa5+ Kb4 5.Rb5+ Ka4 6.Kc5 h2 7.Rb8 h1Q 8.Ra8# anyway,
>that's a beautiful line, no?]
It would be (the Black Pe4 which should have moved on Black's 3rd
obstructs the h1-a8 line) if forced, but White can also play 7 Rb7
and 8 Ra7# when the Qh1 is completely irrelevant.
>Well, the last line was already given by Ilias and I read it in the Denker
>text for the first time. Also Noam Elkies joined the party after two weeks
>of heavy night&day analysis OTB.
:-) Not -- only a few days ago did I notice the post claiming a
beautiful win after 1 Rg5 and 3 Kb6, and having missed the original
post I didn't even know what position was being discussed until I
found that post in the archives. Once I did, the position was not
sufficiently cluttered that a literal analysis board was needed,
and it took less than 20 minutes to find the key idea of blocking a5
with a new Queen.
So, did Kasparov actually claim 1 Rg5 as a win!?
>P.S.
>
>Ilias or others, how about joining with your own Misc (nm)?! We have plenty
>of numbers left ... Ilias, each of your little examples had earned a
>seperated number I think. A position and then a little story behind it, ok?
Okay, how about this one with only seven men on the board:
{.......q}
{........}
{Q..R....}
{..r.....}
{...p.k..}
{........}
{........}
{......K.}
{WTM}
White Kg1,Qa6,Rd6; Black Kf4,Qh8,Rc5,d4; White to play -- what result?
--Noam D. Elkies (USCF 2263)
[remove extra plural from e-address to reply]
Hans Havermann
> ...I published it for the computerchess freaks in mind.
8/8/8/p1K4p/k1ppppRp/rp5B/1P2q3/8 w - -
09/01|00:00:01| 127372| -136| Rxf4 Qg4 Rxg4 Ra2 Rxe4 Rxb2 Rxd4 Rc2
Bf5 Rc1 Rxc4+ Rxc4+ Kxc4
09/04|00:00:01| 149714| +0| Rg5 Qg4 Bxg4 hxg4 Kb6 Ra2 Rxa5+ Kb4
Rb5+ Ka4 Ra5+
10/01|00:00:01| 343667| +0| Rg5 Qg4 Bxg4 hxg4 Kb6 Ra2 Rxa5+ Kb4
Rb5+ Ka4 Ra5+
11/01|00:00:03| 1008443| +0| Rg5 Qg4 Bxg4 hxg4 Kb6 e3 Rxa5+ Kb4
Rb5+ Ka4 Ra5+
12/01|00:00:06| 2323417| +0| Rg5 Qg4 Bxg4 hxg4 Kb6 e3 Rxa5+ Kb4
Rb5+ Ka4 Ra5+
13/01|00:00:14| 5914960| +0| Rg5 Qg4 Bxg4 hxg4 Kb6 e3 Rxa5+ Kb4
Rb5+ Ka4 Ra5+
14/01|00:00:36| 15708908| +0| Rg5 Qg4 Bxg4 hxg4 Kb6 e3 Rxa5+ Kb4
Rb5+ Ka4 Ra5+
15/01|00:01:47| 42574246| +0| Rg5 Qg4 Bxg4 hxg4 Kb6 e3 Rxa5+ Kb4
Rb5+ Ka4 Ra5+
16/01|00:04:53| 119328180| +0| Rg5 Qg4 Bxg4 hxg4 Kb6 e3 Rxa5+ Kb4
Rb5+ Ka4 Ra5+
17/01|00:14:24| 360355092| +0| Rg5 Qg4 Bxg4 hxg4 Kb6 e3 Rxa5+ Kb4
Rb5+ Ka4 Ra5+
18/01|00:53:21| 1225106815| +0| Rg5 Qg4 Bxg4 hxg4 Kb6 e3 Rxa5+ Kb4
Rb5+ Ka4 Ra5+
19/01|03:05:17| 4344679908| +0| Rg5 Qg4 Bxg4 hxg4 Kb6 e3 Rxa5+ Kb4
Rb5+ Ka4 Ra5+
19/12|08:55:35| 12946815555| +M10| Rg1 Qg4 Bxg4 hxg4 Rc1 c3 Rd1 d3
Re1 e3 Rf1 f3 Rg1 g3 Rh1 h3 Rxh3 cxb2
Rh4+
19/17|09:01:02| 13079804906|
search time : 09:01:02
total nodes : 13079804906
nodes per second : 402926
ab-nodes : 11116878592
ab-nodes per second : 342458
evaluations : 8786449192
evaluations per second : 270668
move-generations : 3687438036
ab-move-generations : 2296554028
ttable-hits : 2123485625
ttable-cutoffs : 1968815216
ab-branching-factor : 4.841
MacChess (5.0b03) on a 300 MHz G3.
> Could a computer or even *you* prove that also 1.Rg5 wins? :)
1....Qg4 is forced. White now has 18 possible moves. If, for each one,
we can show that black wins or draws, then Rg5 does *not* win. I think
this is still too many plies distant for most computers. Humans approach
it thus: *If* the black King is allowed to escape the confines of its
a-file prison, *then* the overwhelming pawn threat will take care of the
rest.
Let's assume this to be true. Here's the groundwork (of a pseudo-proof):
*2.bxa3
*2....Kxa3. The King escapes.
*2.R? or *2.Kd5
*2....Ra2 and *3....Rxb2. The King escapes.
*2.Kc6 or *2.Kd6 or *2.Kxc4
*2....Ra2. If *3.Rxa5, then *3....Kxa5, else *3....Rxb2. Either way, the
King escapes.
*2.Kxd4
*2....Ra2. If *3.Rxa5, then *3....Kxa5 and the King escapes. If *3.Kc3
[exercise A], else *3....Rxb2. Pending outcome of A, the King escapes.
*2.Kb6 [exercise B]
2.Bxg4
Forces 2....hxg4.
*3.bxa3
*3....Kxa3. The King escapes.
*3.R? or *3.Kd5
*3....Ra2 and *3....Rxb2. With some complications, the King escapes.
*3.Kc6 or *3.Kd6 or *3.Kxc4
*3....Ra2. If *4.Rxa5, then *4....Kxa5, else *4....Rxb2. Either way,
with some complications, the King still escapes.
*3.Kxd4
*3....Ra2. If *4.Rxa5, then *4....Kxa5 and the King escapes. If *4.Kc3
[exercise C], else *4....Rxb2. Pending outcome of C, with some
complications, the King still escapes.
3.Kb6
3....e3! White draws. [See Elkies' reply.]
This is hard work! How'm I doin'?
Rolf Tueschen
>In article <6ulbmm$eet$1...@news02.btx.dtag.de>,
>Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote:
>>P.S.
>>
>>Ilias or others, how about joining with your own Misc (nm)?! We have plenty
>>of numbers left ... Ilias, each of your little examples had earned a
>>seperated number I think. A position and then a little story behind it, ok?
>Okay, how about this one with only seven men on the board:
>{.......q}
>{........}
>{Q..R....}
>{..r.....}
>{...p.k..}
>{........}
>{........}
>{......K.}
>{WTM}
>White Kg1,Qa6,Rd6; Black Kf4,Qh8,Rc5,d4; White to play -- what result?
>--Noam D. Elkies (USCF 2263)
> [remove extra plural from e-address to reply]
>
Thanks, Noam. I think you will agree that you position deserves a new Misc
<5>. Please let's use the same letters to have a nice serie of MISC in the
end. Also the brackets should be the same. I have here adopted your <>
while I had the (). But basically that's not important.
Here is my answer as a first challenge although my task as "professional"
doesn't allow that I spend too much time on such puzzles. :)
Here's my story:
It looks as if Black (B) has a material advantage of 1 P. But White (W) is
on the move.
So, W can even try to win. And W will win if B blunders. *If* however B
understands that he propably can't win then he has a clear draw after a
possible sac of his Rook.
A little comment on the Pd4. In one line where W can win this Pd4 is like a
shield for his King, so that W can't be disturbed by checks of the black
Queen.
My verdict is "DRAW".
Two variations.
I. Qf1 is easier but W can't go for a win.
II. Rf6+ is the try for a winner. Kg5 Rg6 Kh5 Qe6 and now:
IF B goes for Rg5+ he will lose!
If however he will play Rc1+ he will draw.
I also analysed what my little hardware could provide for the micro
computer programs. Like DEEP BLUE in 1997 they can't see the dozens of
checks and they almost completely misjudge the positions. If you want to
cooperate with their undoubtable strength you'll have to combine human
thought process as the leading force and the machines as slaves so to
speak. So finally it's still the human expert with the deeper calculation
and experience who will profit the most of the machine help.
Is that ok, Noam, or do you want to send me back in school to do the
basics?
C'mon folks, who could prove the Pope Rolf of being badly wrong?
P.S.
Noam, I hope you will give us your "story" behind that problem?
Noam D. Elkies
Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote:
>[I wrote:]
>
>>White Kg1,Qa6,Rd6; Black Kf4,Qh8,Rc5,d4; White to play -- what result?
>Thanks, Noam. I think you will agree that you position deserves a new Misc <5>.
I agree, assuming that a Misc 4 has been posted (I haven't seen it yet).
>[...] I also analysed what my little hardware could provide for the micro
>computer programs. Like DEEP BLUE in 1997 they can't see the dozens of
>checks and they almost completely misjudge the positions.
Interesting you should bring this up, since the answer to your question
>P.S. Noam, I hope you will give us your "story" behind that problem?
is that it's a study I published in 1990 that was cooked by Deep Thought!
I gave the full story and almost full analysis in my article "Chess Art
in the Computer Age" in the _American Chess Journal_ II (1995). If you
have figured out the entire analysis independently in one day then it is
most impressive. Your verdicts and description of Pd4's role are correct.
One interesting thing is that in all three variations (1 Qf1+, 1 Rf6+ ...
3 Qe6! Rg5+?, and 3...Rc1+!) it is not enough to calculate "dozens of checks"
since several quiet moves are critical to the outcome. Since others may
not analyze as quickly (and may not know my ACJ article, or for that matter
the stem study by Rinck) I won't elaborate further for the time being to
see what other computers, silicon as well as carbon based, come up with.
Rolf Tueschen
>My verdict is "DRAW".
>Two variations.
>I. Qf1 is easier but W can't go for a win.
>II. Rf6+ is the try for a winner. Kg5 Rg6 Kh5 Qe6 and now:
>IF B goes for Rg5+ he will lose!
This Pope (!!!) must be crazy. I can't see a forced winning line. Of course
-- with the help of your opponent you still can win even lost positions ...
I apologize to such subtle authors as Noam and Ilias. Draw!
>If however he will play Rc1+ he will draw.
>I also analysed what my little hardware could provide for the micro
>computer programs. Like DEEP BLUE in 1997 they can't see the dozens of
>checks and they almost completely misjudge the positions. If you want to
>cooperate with their undoubtable strength you'll have to combine human
>thought process as the leading force and the machines as slaves so to
>speak. So finally it's still the human expert with the deeper calculation
>and experience who will profit the most of the machine help.
>Is that ok, Noam, or do you want to send me back in school to do the
>basics?
>C'mon folks, who could prove the Pope Rolf of being badly wrong?
Hans Havermann
> I won't elaborate further for the time being to
> see what other computers, silicon as well as carbon based, come up with.
Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote:
> IF B goes for Rg5+ he will lose!
>
> If however he will play Rc1+ he will draw.
>
> I also analysed what my little hardware could provide for the micro
> computer programs. Like DEEP BLUE in 1997 they can't see the dozens of
> checks and they almost completely misjudge the positions. If you want to
> cooperate with their undoubtable strength you'll have to combine human
> thought process as the leading force and the machines as slaves so to
> speak. So finally it's still the human expert with the deeper calculation
> and experience who will profit the most of the machine help.
OK. I'm intrigued. MacChess thinks it sees a draw at 9-ply:
Rf6+ Kg5 Rg6+ Kh5 Qe6 *Rc1+* Kf2 Qf8+ Rf6 Rc2+ Kf1 Rc1+ Kf2 [37 seconds]
and then drops it at 10-ply in favour of this:
Rf6+ Kg5 Rg6+ Kh5 Qe6 *Rg5+* Kf2 Qh7 Rxg5+ Kxg5 Qe5+ Qf5+ Qxf5+ Kxf5 Ke2
Kf4 Kd3 Ke5 [68 seconds]
This is wrong?
Incidentally, I pushed this to 15-ply in hopes of something better.
Nada!
ilias kastanas 08-14-90
@Okay, how about this one with only seven men on the board:
@
@{.......q}
@{........}
@{Q..R....}
@{..r.....}
@{...p.k..}
@{........}
@{........}
@{......K.}
@{WTM}
@
@White Kg1,Qa6,Rd6; Black Kf4,Qh8,Rc5,d4; White to play -- what result?
After the "obvious" 1. Rf6+ most replies allow short, simple
wins; I was wondering... knowing that Noam posted it! Then, in 1... Kg5,
the Troitzky-Rinck thing showed up: 2. Rg6+, Kh5 3. Qe6, Rg5+ 4. Kf2!
Black hangs on with 4... Qh7 -- only to fall into Zugzwang: any piece move
brings mate or material loss. So the WK strolls over and blocks Pd3!
And yet, the real story is elsewhere:
1. Rf6+, Q:f6! 2. Q:f6+, Ke4 =
Black will advance the P to the 6th with the R behind it, and keep
K and R together. It's a theoretical draw... and offhand I don't see how
White could prevent this. (I'll look more carefully when I get home. Also,
a possible winning try is 1. Qf1+, Ke5 2. Rg6 ).
By an uncanny coincidence, similar endings-theory trouble surfaced
in a Rinck study with the exact same material as this one! (That's another
post).
Ilias
Rolf Tueschen
>OK. I'm intrigued. MacChess thinks it sees a draw at 9-ply:
>Rf6+ Kg5 Rg6+ Kh5 Qe6 *Rc1+* Kf2 Qf8+ Rf6 Rc2+ Kf1 Rc1+ Kf2 [37 seconds]
>and then drops it at 10-ply in favour of this:
>Rf6+ Kg5 Rg6+ Kh5 Qe6 *Rg5+* Kf2 Qh7 Rxg5+ Kxg5 Qe5+ Qf5+ Qxf5+ Kxf5 Ke2
>Kf4 Kd3 Ke5 [68 seconds]
>This is wrong?
No, thT's a draw. But I think W had wished to go for a win? Is there
something in the air?
Rolf Tueschen
> After the "obvious" 1. Rf6+ most replies allow short, simple
>wins; I was wondering... knowing that Noam posted it! Then, in 1... Kg5,
>the Troitzky-Rinck thing showed up: 2. Rg6+, Kh5 3. Qe6, Rg5+ 4. Kf2!
>Black hangs on with 4... Qh7 -- only to fall into Zugzwang: any piece move
>brings mate or material loss. So the WK strolls over and blocks Pd3!
Do you want to play Ke1? Well, I play Re5!!. It's a DRAW.
> And yet, the real story is elsewhere:
> 1. Rf6+, Q:f6! 2. Q:f6+, Ke4 =
Damned. That might be true.
Noam D. Elkies
ilias kastanas 08-14-90 <ika...@sol.uucp> wrote:
>@White Kg1,Qa6,Rd6; Black Kf4,Qh8,Rc5,d4; White to play -- what result?
> After the "obvious" 1. Rf6+ most replies allow short, simple
>wins; I was wondering... knowing that Noam posted it! Then, in 1... Kg5,
>the Troitzky-Rinck thing showed up: 2. Rg6+, Kh5 3. Qe6, Rg5+ 4. Kf2!
>Black hangs on with 4... Qh7 -- only to fall into Zugzwang: any piece move
>brings mate or material loss. So the WK strolls over and blocks Pd3!
The "Troitzky-Rinck thing" is certainly involved (but how does Troitzky
get into this? I thought Rinck was the discoverer), but your continuation
has already been refuted by Rolf Tueschen (...Re5+!). That's why Rinck's
original position (which, rotated to get bKh5, is Kf1,Qe6,Rg6/Kh5,Qh7,Rg5)
is *mutual* Zugzwang -- even the distant White King has no neutral moves.
> And yet, the real story is elsewhere:
>
> 1. Rf6+, Q:f6! 2. Q:f6+, Ke4 =
>
> Black will advance the P to the 6th with the R behind it, and keep
>K and R together. It's a theoretical draw... and offhand I don't see how
>White could prevent this.
That Q vs. R+P ending is theoretical indeed, but is it a draw? With only
5 men on the board it can be looked up in the Thompson databases. At any
rate "P to the 6th with the R behind it, and keep K and R together" is not
by itself enough to draw; consider the position Kd2,Qf6/Ke4,Rd5,d3 BTM.
--Noam D. Elkies (USCF 2263)
[delete superfluous plural from e-address to reply]
ilias kastanas 08-14-90
Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote:
@ika...@sol.uucp (ilias kastanas 08-14-90) wrote:
@
@chewing gum? No, the original solution was to create a pendulum with the
Flatten the chewing gum and slide it under the door?
...
@> I saw once in a weekly chess column (next to crosswords etc) the
@>study Ke6, Be2, Ne7; Ke8, Ne5 (Adamson, 1924), Wh to move and win. The
@>newspaper's solution was the line
@
@>1. Nd5, Nd7 2. Kd6!, Kd8 3. Bd3, Ke8 4. Nc7+, Kd8 5. Ne6+, Kc8 6. Ba6+
@
@>-- which is fine for the paper's purposes. Still, there is more to the
@>study. The symmetric "solution", 1. Nf5, Nf7 2. Kf6... 4. Ng7+... does fail
@>since Wh cannot play 6. Bi6+ (or else Bl too would reply 6... Kg9 !). Un-
@>fortunately, 4. Bc6+! (instead of Ng7+) leads to a Wh win as well; 1. Nf5
@>is a cook! The paper wouldn't care; a chess magazine would.
@
@Wait a minute. I suffer from an overflooding of information.
Sorry, that was too fast. The position is symmetric (in fact all
pieces are on the axis of symmetry, the "e" file). So the mirror-image of
the solution "ought" to be a solution too; it's interesting to see why not
(4 files to the left of "e" and only 3 to the right). It's a standard
point re symmetric problems.
But it turned out that there _was_ a solution starting 1. Nf5;
a very different one. (The Black N ends up trapped at a7... after 20
moves or so). Sort of unexpected, with such limited material.
There are various symmetry-breaking factors. A tidbit by Dawson:
Ke8, Ne6, Pe7; Ke2, Pd3, f3 W to move. Obviously, either 1. Nd4+
or 1. Nf4+. One wins, the other doesn't.
@>@Kasparov *did* recommand Rg5 because I have it written inthe position from
@>@his buddy's German CSS. *Funny* Kasparov moved the h-pawn. You the e-pawn.
@>@*Close*, I would say in full impostordom. My Crafty played the d-pawn. Now,
@>@what is that? Confusion all over?
@
@> Haven't I told him a dozen times?... But he still confuses "h"
@>and "e"...
@
@> Eh, when GK moves the h Pawn, it's not in response to 3. Kb6.
@>But you have forbidden me to talk about the subject...
@
@Please end the torture and tell us the whole story.
Me?!... I have no easy access to German CSS's!... Eh, I meant it's
highly unlikely GK would miss the correct defense 3... e3 to 3. Kb6.
Maybe he was describing a search for a cook, not claiming one.
A fun offshoot: suppose, after Rg5, 3. K:d4, Kb4? (sure,
3... R:a2; but never mind) 4. b:a3+, K:a3. It's an ending in itself;
Rook against a sea of Pawns! A couple of lines yield a draw, but it's the
R that has winning chances.
Cooks do happen, even to the very best. Consider: Kc2, Qh4, Rh2;
Kb7, Qf7, Rf1, Pc4 Wh to win (Rinck, 1926). It's a beautifully simple
illustration of the cross-pin: 1. Qe4+, Ka6 2. Rh6+, Rf6 3. Qf5! The
catch, however, is: 3... R:h6 4. Q:f7, Rc6 = !
Pawns on their 5th lose this ending... except for the Bishop P, which
draws (as discovered by Cheron). Roughly, the best Wh can achieve is:
Pc4 blocked by WKc3, and BRc8, BKb8 attacked by 1. Qb6+, Ka8. Wh trian-
gulates to lose a tempo: 2. Kd2, c3+ 3. Kc2. But 3... Rc7! saves Black.
Cheron revised Rinck's study: Kc3, Qe1, Rh3; Ka8, Qf8, Rf2, Pc5.
@> Yeah, so I didn't offer any prize, nag, nag, nag... All right; as I
@>don't know your taste in, uh, cars etc, the prize is: no kid-stuff 655MHz
@>processor, but a Gigabyte of dual-ported, write-only, 0.25 picosecond memory
@> -- fastest on the planet. Yes, that's "pico".
@
@Pico-bello. As we say it in German. :)
Okay... "bello" is German; "schoen" must be Italian.
(footnote: is it OK to write Umlauts as e's?)
@> Right, I remember it's a game by Evans; which I don't have... nor do
@>we need it! In [Kh2, Qe4, Pd5, g2, h7; Kb7, Qc3, Pa7, b2] he played
@>1. d6+?, Ka6 2. d7 -- and 2... b1=Q+ 3. Q:b1, Qe5+* was perpetual check!
@>Its entire analysis is just "conjugate squares", e.g. (K)h2 - (Q)e5, marked on
@>the board, s-S, t-T, u-U... (s = h2, S = e5); chosen to be 'connected'
@>and to avoid mishaps like ...Qc7+?. Kh1, or the WQ interposing. (In particular
@>the WK must not reach a1 !). One possible choice is
@
@
@>h2-e5 h1-h5 g1-c5 f1- (c4 or f8) f2- (c5 or f8) g3-d6 h3-e6
@>e1- (b4 or e7) e2-e7 d1-d6 d2-b4 c1-f4 c2- (e4 or c5) b2-d4
@>a3-a4 b3-d5 f3- (c3 or f8) g4-b4 h5-c5 ... and a few more.
@
@Ok, you're my next editor in chief. Let's build up a whole series of MISC
@(nm). But with a certain relation to dumb machines too. Hope you're also
@familiar with PC machines ... :)
I had better be; else the joke is on all those I've been doing
software engineering work for!
Of course, we use _real_ computers... but anyway...
So then, perp + didn't prove very popular; still, it's a very
good illustration of strengths/weaknesses, C- vs Si- based... Maybe
it's time for some endings now... regular ones instead of "endings" like, say
Kb1, Ng1, Ng6, Pa2, f2, h2, h6; Kf7, Ba3, Bd1, Pa5, c7, e6, g5 (W to move)
(White's new Q will get roughed up; the BB's fly back and open fire on WK
and Q alike, two-ways. What is the outcome here?). It's pretty, but not
exactly everyday fare! (Incidentally, Si should do fine on this one).
@>(it's crucial that stalemate makes ...Qb4+ and ...Qa4+ available!).
@>Draw a picture; visually the whole thing is rather clear and motivated.
@
@
@> Building such a list doesn't take too much effort, and provides
@>a complete blueprint for play; what move to make becomes table lookup.
@
@>On the other hand, imagine a usual search algorithm trying to do as much!...
@>The WK can roam over 25 squares or so. The combinatorial explosion is
@>absolutely prohibitive.
@I will try to get into contact with Hsu and his team to explain that.
@Perhaps he will then come back to normal with all the DEEP BLUE hype. He
@had lost all respect for the human champ. All IMHO of course.
Just a stab in the dark, but I'm beginning to think you don't
fully respect the Champion of the Planet, nor its entourage...
The man-years, silicon-years and money spent do seem to indicate a
fair amount of respect for GK, operationally speaking, uh?! Anyway, DB could
actually afford a separate module for perp +... possibly using some form
of conjugate squares algorithm. (Whether it has one, I don't know).
The perp + in GK-DB 2nd wasn't as long as the one above. (It did have
some "quiet" moves, but this mostly impacts humans). Apparently it was too
much; OTB at least, DB didn't "see" it. (Nor did GK -- or anybody else,
in real-time).
Ilias
ilias kastanas 08-14-90
@ilias kastanas 08-14-90 <ika...@sol.uucp> wrote:
@>In article <6uljlm$8kd$1...@news.fas.harvard.edu>, [I wrote:]
@
@>@White Kg1,Qa6,Rd6; Black Kf4,Qh8,Rc5,d4; White to play -- what result?
@
@> After the "obvious" 1. Rf6+ most replies allow short, simple
@>wins; I was wondering... knowing that Noam posted it! Then, in 1... Kg5,
@>the Troitzky-Rinck thing showed up: 2. Rg6+, Kh5 3. Qe6, Rg5+ 4. Kf2!
@>Black hangs on with 4... Qh7 -- only to fall into Zugzwang: any piece move
@>brings mate or material loss. So the WK strolls over and blocks Pd3!
@
@The "Troitzky-Rinck thing" is certainly involved (but how does Troitzky
@get into this? I thought Rinck was the discoverer), but your continuation
@has already been refuted by Rolf Tueschen (...Re5+!). That's why Rinck's
@original position (which, rotated to get bKh5, is Kf1,Qe6,Rg6/Kh5,Qh7,Rg5)
@is *mutual* Zugzwang -- even the distant White King has no neutral moves.
Right; I was referring not to this specific mechanism but to
earlier studies, like Kb4, Qf1, Re2; Kc6, Qg7, Rf6 (Troitzky 1898),
or Ke2, Qh2, Re5; Kd4, Qb7, Rc7, Pc5 (Rinck 1904).
Heh, heh... I was careful to block rather than capture the "d"
Pawn, to avoid ...Rd5+; but on the way there is also ...Re5+! I saw it
when I got home -- too late.
@> And yet, the real story is elsewhere:
@>
@> 1. Rf6+, Q:f6! 2. Q:f6+, Ke4 =
@>
@> Black will advance the P to the 6th with the R behind it, and keep
@>K and R together. It's a theoretical draw... and offhand I don't see how
@>White could prevent this.
@
@That Q vs. R+P ending is theoretical indeed, but is it a draw? With only
@5 men on the board it can be looked up in the Thompson databases. At any
@rate "P to the 6th with the R behind it, and keep K and R together" is not
@by itself enough to draw; consider the position Kd2,Qf6/Ke4,Rd5,d3 BTM.
It was a thumbnail sketch I gave!... Your position is in fact a
Zugzwang to watch out for -- after all, K and R are about to be forcibly
separated! It occurs, translated, in Kf1, Qe6; Kd4, Rf4, Pf2 (Prokop 1925)
... with Black on move; and Black duly loses. But if White is on move,
Black draws. It's a matter of minding Zugzwangs and tempi; the K and R
have to stay together properly!
In our case the WK is not yet blocking the P. It looks like a
pretty certain draw; Black should have no trouble entering the right posi-
tions. To be sure, the tablebases have the last word.
Ilias
Noam D. Elkies
ilias kastanas 08-14-90 <ika...@sol.uucp> wrote:
> Right; I was referring not to this specific mechanism but to
>earlier studies, like Kb4, Qf1, Re2; Kc6, Qg7, Rf6 (Troitzky 1898),
>or Ke2, Qh2, Re5; Kd4, Qb7, Rc7, Pc5 (Rinck 1904).
Sacrificing a Rook in the midst of a checking attack is still not as
striking as a passive R-sac contained in a Zugzwang...
I wrote about the ending Kg1,Qf6+/Kf4,Re5,d4:
>@ That Q vs. R+P ending is theoretical indeed, but is it a draw? With only
>@ 5 men on the board it can be looked up in the Thompson databases. At any
>@ rate "P to the 6th with the R behind it, and keep K and R together" is not
>@ by itself enough to draw; consider the position Kd2,Qf6/Ke4,Rd5,d3 BTM.
> It was a thumbnail sketch I gave!... Your position is in fact a
>Zugzwang to watch out for -- after all, K and R are about to be forcibly
>separated! It occurs, translated, in Kf1, Qe6; Kd4, Rf4, Pf2 (Prokop 1925)
>... with Black on move; and Black duly loses. But if White is on move,
>Black draws. It's a matter of minding Zugzwangs and tempi; the K and R
>have to stay together properly!
I think you overestimate Black's chances here. In the position I gave
Black loses whether or not B is on move: WTM passes the move to Black
with 1 Qe6+ Kd4 2 Qe3+ Kc4 3 Qb6. The Prokop position you cite is a
*mutual* Zugzwang only because Black's pawn is already on the 7th; move
it up a rank, and WTM wins with 1 Qd7+ Ke5 (Ke4 2 Qd6!) 2 Ke3! since now
2...f2 allows 3 Qe7+, 4 Qe4+, 5 Q:f5.
>In our case the WK is not yet blocking the P. It looks like a pretty
>certain draw; Black should have no trouble entering the right positions.
BK's position is not ideal either; one line that comes to mind is 3 Qh4+
Ke5 4 Kf2 Rd5 5 Qe7+ Kf4 6 Qf6+ Ke4 7 Ke2 d3+ (Re5?) 8 Kd2 reducing to
a previously solved problem.
>To be sure, the tablebases have the last word.
Indeed; will anybody look them up and post the answer (and winning line
if it is a White win)?
NDE
ilias kastanas 08-14-90
@ilias kastanas 08-14-90 <ika...@sol.uucp> wrote:
@
@> Right; I was referring not to this specific mechanism but to
@>earlier studies, like Kb4, Qf1, Re2; Kc6, Qg7, Rf6 (Troitzky 1898),
@>or Ke2, Qh2, Re5; Kd4, Qb7, Rc7, Pc5 (Rinck 1904).
@
@Sacrificing a Rook in the midst of a checking attack is still not as
@striking as a passive R-sac contained in a Zugzwang...
These two studies introduced me to endings of this type. I use
them to epitomize... and certainly not to insinuate their being as striking
as 3. Kh6!
@I wrote about the ending Kg1,Qf6+/Kf4,Re5,d4:
@
@>@ That Q vs. R+P ending is theoretical indeed, but is it a draw? With only
@>@ 5 men on the board it can be looked up in the Thompson databases. At any
@>@ rate "P to the 6th with the R behind it, and keep K and R together" is not
@>@ by itself enough to draw; consider the position Kd2,Qf6/Ke4,Rd5,d3 BTM.
@
@> It was a thumbnail sketch I gave!... Your position is in fact a
@>Zugzwang to watch out for -- after all, K and R are about to be forcibly
@>separated! It occurs, translated, in Kf1, Qe6; Kd4, Rf4, Pf2 (Prokop 1925)
@>... with Black on move; and Black duly loses. But if White is on move,
@>Black draws. It's a matter of minding Zugzwangs and tempi; the K and R
@>have to stay together properly!
@
@I think you overestimate Black's chances here. In the position I gave
You might well be right!
@Black loses whether or not B is on move: WTM passes the move to Black
@with 1 Qe6+ Kd4 2 Qe3+ Kc4 3 Qb6. The Prokop position you cite is a
@*mutual* Zugzwang only because Black's pawn is already on the 7th; move
@it up a rank, and WTM wins with 1 Qd7+ Ke5 (Ke4 2 Qd6!) 2 Ke3! since now
@2...f2 allows 3 Qe7+, 4 Qe4+, 5 Q:f5.
And with Pf3, Rf6, Kf7 it's drawn with or without the move.
Usually if Wh wins with the move, he also wins without it, yes; but not
always, e.g. Pf3, Kf2, Rg2 facing WKh4, WQe4 or WKh1, WQe5. Anyway,
all I meant was: hereabouts, most wins [draws] are by reaching [avoiding]
certain key Zugzwang positions.
@>In our case the WK is not yet blocking the P. It looks like a pretty
@>certain draw; Black should have no trouble entering the right positions.
@
@BK's position is not ideal either; one line that comes to mind is 3 Qh4+
@Ke5 4 Kf2 Rd5 5 Qe7+ Kf4 6 Qf6+ Ke4 7 Ke2 d3+ (Re5?) 8 Kd2 reducing to
@a previously solved problem.
3. Qh4+ is a very good move! Your line demonstrates the hazards
of K & R being too closely behind the P. Black may have to resort to
3... Ke5 4. Kf2, Rc2+ (based on 5. Ke1, d3 = , a draw of the P-supports-R
type, for a change!).
If Wh doesn't relish ...d4-d3 he might insert 3... Ke5 4. Qg5+,
Kd6 5. Qd8+, Ke5 or so before 6. Kf2. It could be that Bl should
consider 3...Kd5!? .
@>To be sure, the tablebases have the last word.
@
@Indeed; will anybody look them up and post the answer (and winning line
@if it is a White win)?
Yes... please!
Ilias
michael adams
Micky.
Rolf Tueschen
>In article <6uq8j3$ikb$1...@hades.csu.net>,
>ilias kastanas 08-14-90 <ika...@sol.uucp> wrote:
>>In article <6uljlm$8kd$1...@news.fas.harvard.edu>, [I wrote:]
>>@White Kg1,Qa6,Rd6; Black Kf4,Qh8,Rc5,d4; White to play -- what result?
>> After the "obvious" 1. Rf6+ most replies allow short, simple
>>wins; I was wondering... knowing that Noam posted it! Then, in 1... Kg5,
>>the Troitzky-Rinck thing showed up: 2. Rg6+, Kh5 3. Qe6, Rg5+ 4. Kf2!
>>Black hangs on with 4... Qh7 -- only to fall into Zugzwang: any piece move
>>brings mate or material loss. So the WK strolls over and blocks Pd3!
>The "Troitzky-Rinck thing" is certainly involved (but how does Troitzky
>get into this? I thought Rinck was the discoverer), but your continuation
>has already been refuted by Rolf Tueschen (...Re5+!). That's why Rinck's
>original position (which, rotated to get bKh5, is Kf1,Qe6,Rg6/Kh5,Qh7,Rg5)
>is *mutual* Zugzwang -- even the distant White King has no neutral moves.
>> And yet, the real story is elsewhere:
>>
>> 1. Rf6+, Q:f6! 2. Q:f6+, Ke4 =
>>
>> Black will advance the P to the 6th with the R behind it, and keep
>>K and R together. It's a theoretical draw... and offhand I don't see how
>>White could prevent this.
>That Q vs. R+P ending is theoretical indeed, but is it a draw? With only
>5 men on the board it can be looked up in the Thompson databases.
I have to correct myself another time. Of course this is not a draw as the
CDR shows. It's a win for the Queen.
Just a little aside. I think it's now worthwhile to enter the detailed
story of your study, Noam. I never saw that one nor read the articles you
mentioned. Also no more interested will show up, I fear.
For what it's worth. What was the orifinal idea you had. A win? And now it
was cooked to a draw? What did Deep Thought exactly find outr?
It's not necessary to quote the *complete* analysis, if it takes several
pages. Just tell us your story, please.
I think experts like you are much too seldom appearing here in the chess
groupes. Also on the chess sites studies seem to have sort of alibi
function. I wished we had as much material on the servers as there's the
actual chessgames.
Here in rgcc I almost never read a post about the mate and self-mate
features of FRITZ. A sprecial study-mode seems to be lacking in all
programs ...
There you have the relevance of that field in the eyes of the technocrats.
Studies however are the real pearls of chess, no?!
> At any
>rate "P to the 6th with the R behind it, and keep K and R together" is not
>by itself enough to draw; consider the position Kd2,Qf6/Ke4,Rd5,d3 BTM.
>--Noam D. Elkies (USCF 2263)
Carl Tillotson
> Why i hate the French,e4-e6-etc.It seems to me they 'av no roast-bif.
>
> Micky.
And I hate thick pillocks like yourself who add a mere one line to a
post after quoting the entire post.
--
Adios Amigo
Carl Tillotson
Lancashire Chess Association
homepage: http://www.lancashirechess.demon.co.uk/
Virtual Access 4.02 build 244 (32-bit)
Using Win95
ilias kastanas 08-14-90
Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote:
[Kg1, Qf6; Ke4, Rc5, Pd4 w]
>I have to correct myself another time. Of course this is not a draw as the
>CDR shows. It's a win for the Queen.
Well, if the tablebase says so, that's it. Do you have the
optimal line?
Ilias
Noam D. Elkies
Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote:
>I have to correct myself another time. Of course this is not a draw as the
>CDR shows. It's a win for the Queen.
>
>Just a little aside. I think it's now worthwhile to enter the detailed
>story of your study, Noam. I never saw that one nor read the articles you
>mentioned. Also no more interested will show up, I fear.
>For what it's worth. What was the original idea you had. A win? And now it
>was cooked to a draw? What did Deep Thought exactly find out?
Yes, it was suppose to be a win after
1 Rf6+ Kg5 2 Rg6+ Kh5 3 Qe6! Rg5+ 4 Kf2!! Qh7!
This quite move is Black's only serious defense (short of conceding
a lost Q vs. R+P ending) against 5 Qh3+; the check 4...Qf8+ is worse:
5 Rf6 Qa3 (White threatened 6 Qh3#) 6 Rh6#! Also 4...Kf4 5 Qe4+ Kh5
6 Qh1+ and wins.
5 Kf1!!
Mutual Zugzwang, and the only reason White played 4 Kf2 instead of 4 Kf1.
Without Pd4 this is the Rinck Zugzwang rotated 90 degrees; White to move
only draws, e.g. Ke1(2) Re5+ or Rf6 Qd3(b1)+. Black to move loses without
the d-pawn: Qh8 6 Qh3+, Q-else 6 Rh6#, Kh4 6 Qe4+, R:g6 6 Qh3+, Q:g6 6 Qh3#!
But what about that extra pawn?...
5 ... d3 6 Rf6!
My discovery: the pawn blocks Qd3(b1)+ so White can safely make this
quite move which threatens 7 Qh3#; Black has no defense, e.g. Rg4
7 Qe5+ or Kh4 7 Rf4+ winning quickly.
But what about 3...Rc1+, you ask. Black denudes his own King but goes
all out for perpetual check. Well, I had a long analysis to refute that:
3...Rc1+ 4 Kf2! Qf8+!
Not Rc2+ 5 Ke1 Rc1+ 6 Kd2 +-.
5 Rf6 Rc2+ 6 Kf1 Rc1+ 7.Ke2 d3+! 8.Kd2
This wins the rook (Rc2+? 9 Kd1 and if for instance 9...Rd2+
10 K:d2 Qb4+ 11 Ke3! and White soon escapes), but Black's Queen
can now try for perpetual by herself:
8...Qb4+! 9 K:c1 Qc3+ 10 Kb1 Qb4(c2)+ 11 Ka2(a1) Qa4+ 12 Kb2 Qb4+
13 Qb3 Qd2+ 14 Kb1!
Avoiding 14.Ka3 Qa5+ 15.Qa4 Qc3+! and the Rf6 goes lost.
14...Qe1+ 15 Ka2! Qd2+ 16 Qb2 Qa5+ 17 Qa3 Qd2+ 18 Kb3 Qc2(d1)+ 19 Kb4
Phew. Now White Queen can cover both the King and Rook after
Qd2(e1)+ 20 Qc3 (or Qb1+ 20 Qb3 Qe1+ 21 Qc3) and White's trek
is over, leaving him safely a Rook up.
But alas there is a fly in this ointment. Having seen this much,
can you or your computer figure out what I missed in this analysis?
>I think experts like you are much too seldom appearing here
There are several strong players who do post here regularly;
my US rating is only equivalent to about 2150 ELO, and I haven't
competed in tournament chess for about ten years, so I'm surely
far from the best chessplayer on this group. I don't post here
more often for the same reason I dropped out of tournament chess:
shortage of disposable time.
Rolf Tueschen
> [Kg1, Qf6; Ke4, Rc5, Pd4 w]
>Do you have the
>optimal line?
Yes, but as you know "the" optimal line is not there. Because often you
have several possibilities that lead to the same goal with exactly the same
number of moves. You might also know that in human chess it's not
unimportant to exploit cases where you a) have only single moves and b)
where the choice between right or wrong is most difficult, e.g. if you only
have two alternatives and both basically look the "same".
The situation we have here is that White gets the black Pawn in 19 moves
optimally. Thereafter the Q will win against the R.
Here's one optimal line with some single movers.
(Elkies,N 1990) (NB that the position above is after already 2 moves been
made in the original study; it's the 1.Rf6+ Qxf6 line)
3.Kg2 Rc3 4.Qd6 Ke3 5.Qd5 d3 6.Qa5 Rc2+ 7.Kg3™ d2 8.Qa3+™ Ke4 9.Qf3+™ Kd4
10.Kf4™ Rb2 11.Qa3 Kc4 12.Qd6 Rc2 13.Ke3 Kb3 14.Qd4 Ka3 15.Qg4 Kb3 16.Kd3
Rb2 17.Qd1+ Kb4 18.Kd4 Ka3 19.Kc3 Ka2 20.Qf1 Line
Always where you see the strange symbol it's the INFORMATOR symbol for a
single move, all others don't win. Especially in the case of 8.Qa3+ it's
very difficult to detect. However if you think of the Pd2 that wants to
queen, the idea with the following Qf3 is not too mystic. Then 10.Kf4
already is something obvious. But that's kind of nonsense. Because why then
not 8.Qe5+? Oh, well.
Rolf Tueschen
>In article <6v51vq$c90$1...@news00.btx.dtag.de>,
>Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote:
>>I have to correct myself another time. Of course this is not a draw as the
>>CDR shows. It's a win for the Queen.
>>
>>Just a little aside. I think it's now worthwhile to enter the detailed
>>story of your study, Noam. I never saw that one nor read the articles you
>>mentioned. Also no more interested will show up, I fear.
>>For what it's worth. What was the original idea you had. A win? And now it
>>was cooked to a draw? What did Deep Thought exactly find out?
>Yes, it was suppose to be a win after
> 1 Rf6+ Kg5 2 Rg6+ Kh5 3 Qe6! Rg5+ 4 Kf2!! Qh7!
> This quite move is Black's only serious defense (short of conceding
> a lost Q vs. R+P ending) against 5 Qh3+; the check 4...Qf8+ is worse:
> 5 Rf6 Qa3 (White threatened 6 Qh3#) 6 Rh6#! Also 4...Kf4 5 Qe4+ Kh5
> 6 Qh1+ and wins.
>5 Kf1!!
Like in all other fields it seems to be at least NOT contra-indicated to be
informed about the general knowledge of that particular field's history. I
must admit that I didn't see this. Neither my program.
> Mutual Zugzwang, and the only reason White played 4 Kf2 instead of 4 Kf1.
> Without Pd4 this is the Rinck Zugzwang rotated 90 degrees; White to move
> only draws, e.g. Ke1(2) Re5+ or Rf6 Qd3(b1)+. Black to move loses without
> the d-pawn: Qh8 6 Qh3+, Q-else 6 Rh6#, Kh4 6 Qe4+, R:g6 6 Qh3+, Q:g6 6 Qh3#!
> But what about that extra pawn?...
>5 ... d3 6 Rf6!
> My discovery: the pawn blocks Qd3(b1)+ so White can safely make this
> quite move which threatens 7 Qh3#; Black has no defense, e.g. Rg4
> 7 Qe5+ or Kh4 7 Rf4+ winning quickly.
>But what about 3...Rc1+, you ask. Black denudes his own King but goes
>all out for perpetual check. Well, I had a long analysis to refute that:
>
> 3...Rc1+ 4 Kf2! Qf8+!
>
> Not Rc2+ 5 Ke1 Rc1+ 6 Kd2 +-.
>
> 5 Rf6 Rc2+ 6 Kf1 Rc1+ 7.Ke2 d3+! 8.Kd2
>
> This wins the rook (Rc2+? 9 Kd1 and if for instance 9...Rd2+
> 10 K:d2 Qb4+ 11 Ke3! and White soon escapes), but Black's Queen
> can now try for perpetual by herself:
Exactly, I played that.
> 8...Qb4+! 9 K:c1 Qc3+ 10 Kb1 Qb4(c2)+ 11 Ka2(a1) Qa4+ 12 Kb2 Qb4+
> 13 Qb3 Qd2+
I stopped here with a draw. Premature juvenile behaviour ...
>14 Kb1!
>
> Avoiding 14.Ka3 Qa5+ 15.Qa4 Qc3+! and the Rf6 goes lost.
> 14...Qe1+ 15 Ka2! Qd2+
I think you overlooked the cook right here, yes? I found a tricky line.
15... Qe2+!!. Excuse my naivety. Of course this move is obvious. Because
after 16.Qb2 it allows 16... d3-d2!!! That's a good one, no? That's a Bobby
Fischer move. Alas, I would be a GM too if I could foresee such moves OTB.
>16 Qb2 Qa5+ 17 Qa3 Qd2+ 18 Kb3 Qc2(d1)+ 19 Kb4
> Phew. Now White Queen can cover both the King and Rook after
> Qd2(e1)+ 20 Qc3 (or Qb1+ 20 Qb3 Qe1+ 21 Qc3) and White's trek
> is over, leaving him safely a Rook up.
At first I saw a cook here. With 19... Qg4 and 20... Qc8 but then after
21.Kb3 I had to play 21... Qc2 and then after 22.Kb4 I had no longer Qg4. I
had to enter your two main lines, and then it's over. Alright so far?
>But alas there is a fly in this ointment. Having seen this much,
>can you or your computer figure out what I missed in this analysis?
>>I think experts like you are much too seldom appearing here
>There are several strong players who do post here regularly;
>my US rating is only equivalent to about 2150 ELO, and I haven't
>competed in tournament chess for about ten years, so I'm surely
>far from the best chessplayer on this group. I don't post here
>more often for the same reason I dropped out of tournament chess:
>shortage of disposable time.
I see. But you seem too shy to accept a simple fact. It may be true that
we've a lot of good *players* here around. But of people like you or T.
Krabbé only too few. Creative (almost) artists. With your endgame you
really succeeded to involve me without second-thoughts into stuff like
that. Perhaps I hadn't tried it if you had directly told your story.
A little indiscretion. Did you therefore overlook the second line because
you lost control because of dozens of sheets of paper or was it only in
your memory and you missed the d3-d2? Somewhat strange because before you
were smart enough to find subtleties like Kf1!! where also the black Pd4
had his important move. I asked for papers because I really can't
understand how you did all the work without the help of a computer. And
even more I can't understand how the artists decades ago could figure out
and calculate all that ...
What is the most important talent in this case? Memory for the most little
details or a good functioning bookkeeping?
And now the unbeliavable itself.
How could you succeed to put that pawn on d4 and **when**? In the creative
period you had to *see* something, no? Perhaps already Kf1. But the
connection to the later highly complicated and endless checks and
repetitions is imposssible to foresee, no???
I would like to hear from the experts how one could help you technically in
that field of endgame studies. Perhaps K. Thompson should give you access
to his (existing?) 6(7?)-pieces tablebases ...
Computerchess experts should invent a little engine that should be able to
search for possible connections to the exact bases. A priori. So that you
could get a "feeling" for the possibilities of certain positions and
material. Could you tell us if and where you could use that in the whole
process?
You see I already feared that you might run away. So I simply asked a few
questions. :)
Noam D. Elkies wrote:
Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote:
>[I wrote:]
>>Yes, it was supposed to be a win after
>> 1 Rf6+ Kg5 2 Rg6+ Kh5 3 Qe6! Rg5+ 4 Kf2!! Qh7! [...] 5 Kf1 !!
>Like in all other fields it seems to be at least NOT contra-indicated to be
>informed about the general knowledge of that particular field's history. I
>must admit that I didn't see this. Neither my program.
>>5 ... d3 6 Rf6! [1-0]
>>But what about 3...Rc1+, you ask. Black denudes his own King but goes
>>all out for perpetual check. Well, I had a long analysis to refute that:
>>
>> 3...Rc1+ 4 Kf2! Qf8+! 5 Rf6 Rc2+ 6 Kf1 Rc1+ 7.Ke2 d3+! 8.Kd2
>> This wins the rook (Rc2+? 9 Kd1 and if for instance 9...Rd2+
>> 10 K:d2 Qb4+ 11 Ke3! and White soon escapes), but Black's Queen
>> can now try for perpetual by herself:
>Exactly, I played that.
(Though 11 K:d3 seems sufficient too)
>> 8...Qb4+! 9 K:c1 Qc3+ 10 Kb1 Qb4(c2)+ 11 Ka2(a1) Qa4+ 12 Kb2 Qb4+
>> 13 Qb3 Qd2+
>I stopped here with a draw. Premature juvenile behaviour ...
>>14 Kb1! [...] Qe1+ 15 Ka2! Qd2+
>I think you overlooked the cook right here, yes? I found a tricky line.
>15... Qe2+!!. Excuse my naivety. Of course this move is obvious. Because
>after 16.Qb2 it allows 16... d3-d2!!! That's a good one, no? That's a Bobby
>Fischer move. Alas, I would be a GM too if I could foresee such moves OTB.
Well it's obvious once one is told there's a draw -- the point being
that Qe2 controls b5 and e5 so White has no checks and can do no better
than meekly trade his Rook for the advanced pawn (17 Rd6). But seeing
such a move in the midst of a 20-move analysis is a different matter...
>A little indiscretion. Did you therefore overlook the second line because
>you lost control because of dozens of sheets of paper or was it only in
>your memory and you missed the d3-d2? Somewhat strange because before you
>were smart enough to find subtleties like Kf1!! where also the black Pd4
>had his important move. I asked for papers because I really can't
>understand how you did all the work without the help of a computer. [...]
Yes, I did it without computer aid, but it was not a matter of getting
lost in reams of notes. For an analyst, whether human or computer, to
go out to 30 ply and beyond, he/she/it must disregard most possibilities.
In an ending like this, with Queens and Rooks roaming the board and hardly
any pawn cover, the usual heuristic is that one keeps checking as long
as one can; effective quiet moves are rare. Sequences like 4 Kf2! Qh7!
5 Kf1! d3 6 Rf6! are most exceptional, which is precisely what makes
this study interesting. (And it's not a matter of being "smart enough
to find" these moves, but of *constructing* a position which requires
them.) One doesn't expect to run into quiet moves like d3-d2 in a
"random" position in the middle of a subsidiary line where White
is trying to escape perpetual check. So here is a case where a
computer's less refined heuristics for candidate moves actually
proved to the computer's advantage...
>It may be true that we've a lot of good *players* here around.
>But of people like you or T. Krabbe' only too few. Creative
>(almost) artists.
Why "almost"? The chess study (and the chess problem) is an art form;
a minor art, surely, with little market value and no pretensions to
Plumb the Depths of the Human Condition, but an art nevertheless
[NB Troitzky got Russia's "Honored Art Worker" title in 1928],
boasting dozens of periodicals, some 1000 years of history, more
than a thousand practitioners, and millions of at least occasional
appreciators.
>How could you succeed to put that pawn on d4 and **when**? In the
>creative period you had to *see* something, no? Perhaps already Kf1.
>But the connection to the later highly complicated and endless checks
>and repetitions is impossible to foresee, no???
The Rinck study does not exploit the fact that the key Zugzwang is mutual.
I set out to do that by providing a "thematic try" (4 Kf1!?) that fails
only because it ends with that position with White to move. The extra
pawn was the first of various attempts that worked -- or at least seemed
to work.
>I would like to hear from the experts how one could help you technically in
>that field of endgame studies. Perhaps K. Thompson should give you access
>to his (existing?) 6(7?)-pieces tablebases ...
Not 7, certainly, and even 6 will take a while to make accessible.
These tables are huge, even when compressed. The annoying thing is
that the *5*-man databases, which used to be available on the Web at
http://www3.traveller.com/scripts/chess_kt_endings/ , have been removed
(there is still an interface but it is not attached to the data) -- any
idea why this was done, and if an alternative site is available?
A different, and potentially much more useful, kind of assistance would
be a database of the already existing studies which can be searched
according to material, key position, theme, etc., allowing for variations
that do not affect the substance of the idea. Harold van der Heijden
has made a good start towards that goal, but it's still very hard to
fully automate and you can imagine how much time it takes to enter
by hand 100,000+ studies with full source, analysis, and thematic
information. What is needed here is a computer program not only
strong enough not only to solve the study but also capable of
understanding and detecting the artistic intent.
(The same can of course be said for other genres of chess art -- mate in 2,
helpmates, retroanalytic problems, etc. NB even programming a computer to
solve a retro problem is still a difficult problem.)
michael adams
> >
> > Micky.
>
> post after quoting the entire post.
>
> --
> Adios Amigo
>
> Carl Tillotson
>
> Lancashire Chess Association
> homepage: http://www.lancashirechess.demon.co.uk/
>
> Virtual Access 4.02 build 244 (32-bit)
> Using Win95
'gards micky.