Author: Dann Corbit
Date: 16:59:36 08/29/01
Go up one level in this thread
Selected goo about this problem...
The original problem statement:
=====================================================================
# Position: 9
# Move: W
r....r.k Weinstein - Elyoseph, Israel 1992
....bppb This one is really beautiful and should resist computers
..n.p..p for quite a long time, maybe until next century?
p.n.P... 1.Ng5!! hxg5 2.hxg5! Rac8 3.Nf6!! Nb8 (black can also
.p.p.BNP try 3...gxf6, which loses if white play precisely : 4.gxf6
...P.NP. 4...Rfe8 (best defense) 5.Qh5 Rg8 6.Rxc5! Bg6! 7.Qh4 Bxc5
qP..QPB. 8.Be4 Ne7 9Kg2 Qd5 10.Bxd5+-)
..RR..K. 4.Qh5 Bxf6 5.gxf6 gxf6 6.Rxc5 Rxc5 7.Be4 f5 8.Kg2 Rg8 9.Rh1
9...Rg7 10.Bh6 Nd7 11.Bxg7+ Kxg7 12.Qxh7+ 1-0
In fact, Pierre is not sure this one is 100% correct. Any improvements
are wellcome!
=====================================================================
Deep Thought said:
Position #9
r4r1k/4bppb/2n1p2p/p1n1P3/1p1p1BNP/3P1NP1/qP2QPB1/2RR2K1 w - -
+---+---+---+---+---+---+---+---+
8 | *R| | | | | *R| | *K|
+---+---+---+---+---+---+---+---+
7 | | | | | *B| *P| *P| *B|
+---+---+---+---+---+---+---+---+
6 | | | *N| | *P| | | *P|
+---+---+---+---+---+---+---+---+
5 | *P| | *N| | P | | | |
+---+---+---+---+---+---+---+---+
4 | | *P| | *P| | B | N | P |
+---+---+---+---+---+---+---+---+
3 | | | | P | | N | P | |
+---+---+---+---+---+---+---+---+
2 | *Q| P | | | Q | P | B | |
+---+---+---+---+---+---+---+---+
1 | | | R | R | | | K | |
+---+---+---+---+---+---+---+---+
a b c d e f g h
Weinstein - Elyoseph, Israel 1992
1. Ng5!! hxg5 2. hxg5! Rac8 3. Nf6!! Nb8
3. ...gxf6 4. gxf6 Rfe8 5. Qh5 Rg8 6. Rxc5! Bg6! 7. Qh4 Bxc5
8. Be4 Ne7 9. Kg2 Qd5 10. Bxd5 +-
4. Qh5 Bxf6 5. gxf6 gxf6 6. Rxc5 Rxc5 7. Be4 f5 8. Kg2 Rg8 9. Rh1 Rg7
10. Bh6 Nd7 11. Bxg7+ Kxg7 12. Qxh7+
After 9 minutes, it played 1. Nf6, expecting 1. ...Rfc8 2. Nxh7 Kxh7 3. Ra1.
The score was slightly nagative for white. On longer searches, it went up
to half a pawn and creeping up.
=====================================================================
Bruce analyzed the heck out of it:
Subject: Nolot #9
From: Bruce Moreland
E-mail: brucemo@seanet.com
Message Number: 151420
Date: January 22, 2001 at 20:01:22
[D]r4r1k/4bppb/2n1p2p/p1n1P3/1p1p1BNP/3P1NP1/qP2QPB1/2RR2K1 w - - 0 1
After a trillion nodes (19 plies) my program can't solve this.
It wants to play Nxh6 with a dubious +1 score (up an exchange for a pawn, and
probably losing a pawn), but it can't find the key, which is Ng5.
If this is solvable, it is much harder than many of the others.
bruce
=====================================================================
Ren Wu gave it a pounding also:
Subject: Re: Nolot #9
From: Ren Wu
E-mail: renw@iname.com
Message Number: 151534
Date: January 23, 2001 at 12:55:42
In Reply to: Nolot #9
Message ID: 151420
Posted by: Bruce Moreland
At: brucemo@seanet.com
On: January 22, 2001 at 20:01:22
On January 22, 2001 at 20:01:22, Bruce Moreland wrote:
>[D]r4r1k/4bppb/2n1p2p/p1n1P3/1p1p1BNP/3P1NP1/qP2QPB1/2RR2K1 w - - 0 1
>
>After a trillion nodes (19 plies) my program can't solve this.
>
>It wants to play Nxh6 with a dubious +1 score (up an exchange for a pawn, and
>probably losing a pawn), but it can't find the key, which is Ng5.
>
>If this is solvable, it is much harder than many of the others.
>
>bruce
After about 10 hours, my program gives
17. -13 22963.6 1097898436 g4f6 f8c8 f6h7 h8h7 c1a1 a2b3 f3d2 b3b2 g2c6 c8c6
d1b1 b2b1 d2b1 h7g8
18. 0 36290.5 2101713756 g4f6 f8c8 f6h7 h8h7 c1a1 a2b3 f3d2 b3b2 d1b1 b2c2
b1c1 c2b2 c1b1
And think it is a draw. This is about same to Deep Thought's result. I have to
stop it after 15 hours, and it still work on 19th ply. I'll let it run longer at
weekends.
And yes, this one seems very hard.
Ren
=====================================================================
CAP data:
Subject: Re: Nolot #9
From: Dann Corbit
E-mail: dcorbit@solutionsiq.com
Message Number: 151594
Date: January 23, 2001 at 16:52:24
In Reply to: Nolot #9
Message ID: 151420
Posted by: Bruce Moreland
At: brucemo@seanet.com
On: January 22, 2001 at 20:01:22
The current CAP record looks like this:
[D]r4r1k/4bppb/2n1p2p/p1n1P3/1p1p1BNP/3P1NP1/qP2QPB1/2RR2K1 w - - acd 17; acn
-1192429640; acs 15000; ce 0; pv Nf6 Rac8 Nxh7 Kxh7 Ra1 Qb3 Nd2 Qxb2 Rdb1 Qc3
Rc1 Qb2 Rcb1;
So Nf6 is what is chosen and looks drawish.
I also did a follow on with an experimental program where I forced the Ng5
choice (since it is experimental, the results are somewhat questionable):
[D]r4r1k/4bppb/2n1p2p/p1n1P1N1/1p1p1BNP/3P2P1/qP2QPB1/2RR2K1 b - - acd 20; ce
177; pv hxg5 Bxg5 Na4 Bxc6 Bxg5 hxg5 Nc3 Qf3 Nxd1 Bxa8 Nxb2 Be4 Bxe4 Qxe4 Qb3 g6
fxg6 Qxg6 Qxd3 Qxe6;
20 plies here does not mean the same thing as with your programs. It is not
full width to that depth. Anyway, since it looks very good for black (up 1 and
3/4 pawns) the real solution must be very clever.
At any rate, I am dying to see the actual game so that I can understand how to
win with Ng5. Maybe one of the resident chess whizzes like Come or Vincent can
help.
=====================================================================
Vincent's insight:
Subject: Re: Nolot #9
From: Vincent Diepeveen
E-mail: diep@xs4all.nl
Message Number: 151605
Date: January 23, 2001 at 17:48:45
In Reply to: Nolot #9
Message ID: 151420
Posted by: Bruce Moreland
At: brucemo@seanet.com
On: January 22, 2001 at 20:01:22
On January 22, 2001 at 20:01:22, Bruce Moreland wrote:
>[D]r4r1k/4bppb/2n1p2p/p1n1P3/1p1p1BNP/3P1NP1/qP2QPB1/2RR2K1 w - - 0 1
>
>After a trillion nodes (19 plies) my program can't solve this.
>
>It wants to play Nxh6 with a dubious +1 score (up an exchange for a pawn, and
>probably losing a pawn), but it can't find the key, which is Ng5.
>
>If this is solvable, it is much harder than many of the others.
>
>bruce
Diep never solved it so far. Very important for this position
is extending on mate thread which the current diep version is
not always doing.
For sure white must give up 2 pieces first before he has any chance
of mating black. after ng5 real soon nf6 sacrafice has to follow.
After some moves done i get very soon a reasonable score for Ng5 like
-0.30 or something.
However in the root position after some time diep says white wins with
Rxc5 and not Ng5 and every ply score goes up for Rc5 from -0.0x to +1.xx
after a few tens of minutes, so diep will most likely go for Rxc5 for
a long period of time.
Definitely Ng5 is winning, not a hair on my head is doubting that.
This problem IS a tactical shot for sure, but a very deep one...
What time do you need for #8 Bruce?
Diep nowadays is lucky with extensions and finds it
after 2 and a half minute with a 2.xx score.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.