deep.09

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

Moderators: hgm, Rebel, chrisw

peter
Posts: 3187
Joined: Sat Feb 16, 2008 7:38 am
Full name: Peter Martan

Re: deep.09

Post by peter »

Dirt wrote:
peter wrote:Rather think the fortress is stable, do you see Black not losing soon after not giving the Queen- sac
I sure feel stupid. How does black not give up the queen?
Sorry Greg, you're fully right of course, so it's not to be called a sac of course.
:)
What I meant was, if Black doesn't force White to take the Queen on that special square (or let it go untaken), it's over soon.
After 1...Qf5 a fortress arises not to be overcome, is it
:?:
Peter.
jhellis3
Posts: 546
Joined: Sat Aug 17, 2013 12:36 am

Re: deep.09

Post by jhellis3 »

Ok, it is not quite as obvious as it first looks, but eventually white can prevail.

The solution is for white to target this position:
[d]1r6/6r1/2p1k3/p1pbPp1p/Pp1pRPpP/1P1P2K1/2P2Q2/4R3 b - - 37 22

The rook is taboo and white is just going to sac it for the d and c pawns, put the queen on c5 leading to a position similar to this

[d]r7/3r4/2p1k3/p1QbPp1p/Pp3PpP/1P1P2K1/2P5/4R3 w - - 3 26

So the question becomes is if this position is won for White.

Eventually, I get to this position (which looks reasonably promising for white):
[d]r7/r4k2/2Q5/4Pp1p/P2P1PpP/6K1/8/8 b - - 0 34


The problem is that black just has too many breaks to prevent, and can not stop the white king from coming to g3 (if it plays g3 itself it walks into a lost bishop vs rook endgame after a white queen sac).
Dann Corbit
Posts: 12550
Joined: Wed Mar 08, 2006 8:57 pm
Location: Redmond, WA USA

Re: deep.09

Post by Dann Corbit »

jhellis3 wrote:Yeah, as far as I can see black is just busted here... there is no "best move".
After the key move, we have this position:
[d]6r1/6r1/2p1k1pp/p1pbPq2/Pp1p1PpP/1P1P2NR/1KPQ3R/8 w - -
which leads to a fortress:

Code: Select all

FEN: 6r1/6r1/2p1k1pp/p1pbPq2/Pp1p1PpP/1P1P2NR/1KPQ3R/8 w - - 0 1 

Houdini_4_Pro_CA_x64B Tactical:
  4/10	00:00	         941	0	+6.52	Nxf5 gxf5 Rg3 Re8 Rg1
  5/12	00:00	       2,235	2,235,000	+6.47	Nxf5 gxf5 Rg3 Rh8 h5 Re8 Qe1
  6/12	00:00	       3,172	1,586,000	+6.50	Nxf5 gxf5 Rg3 Re8 h5 Rge7 Rg1
  7/16	00:00	       4,866	2,433,000	+6.61	Nxf5 gxf5 Rg3 Re8 Qe1 Rge7 Rg1 Bf3
  8/16	00:00	       7,921	2,640,000	+6.56	Nxf5 gxf5 Rg3 Re8 Qe1 Rge7 Rg1 h5 Re2 Bf3 Rd2
  9/22	00:00	      29,238	4,176,000	+6.58	Nxf5 gxf5 Rg3 h5 Qe1 Bf3 Rf2 Bd5 Rd2 Re8 Rg1 Bf3 Qg3 Rge7 Re1
 10/24	00:00	      92,853	6,632,000	+6.60	Nxf5 gxf5 Rg3 Rf7 Qe1 Re8 Rg1 Rfe7 Re2 Kd7 Rd2 Kc7 Qg3 h5 Re1 Bf3
 11/24	00:00	     174,096	7,254,000	+6.60	Nxf5 gxf5 Rg3 h5 Qe1 Re7 Rg1 Kd7 Rd2 Kc7 Qg3 Rge8 Re1 Bf3
 12/24	00:00	     250,639	7,371,000	+6.54	Nxf5 gxf5 Rg3 h5 Re2 Re7 Re1 Kd7 Qf2 Kc7 Rgg1 Rd8 Rb1 Bf3 Ra1
 13/29	00:00	     349,878	7,289,000	+6.61	Nxf5 gxf5 Rg3 h5 Re2 Re7 Re1 Kd7 Qf2 Kc7 Rgg1 Bf3 Rb1 Rge8 Ra1 Re6
 14/29	00:00	     449,916	7,375,000	+6.52	Nxf5 gxf5 Rg3 h5 Re2 Re7 Qe1 Kd7 Rd2 Kc7 Rg1 Rge8 Qg3 Rd7 Re1 Rde7 Rde2 Bf3
 15/29	00:00	     638,820	7,696,000	+6.56	Nxf5 gxf5 Rg3 h5 Re2 Re7 Qe1 Kd7 Rg1 Kc7 Qg3 Rge8 Ree1 Bf3 Ra1 Re6 Rge1 Bd5
 16/29	00:00	     941,102	7,975,000	+6.50	Nxf5 gxf5 Rg3 h5 Re2 Re7 Qe1 Kd7 Rg1 Kc7 Qg3 Rge8 Ree1 Re6 Ra1 Bf3 Qf2 Bd5 Qd2 Bf3
 17/30	00:00	   1,390,666	8,038,000	+6.51	Nxf5 gxf5 Rg3 h5 Re2 Re7 Qe1 Kd7 Rg1 Kc7 Qg3 Rge8 Ree1 Re6 Ra1 Bf3 Qf2 R6e7 Rae1 Bd5 Rb1 Re6 Ra1 Bf3 Qg3 Bd5 Qf2 Bf3
 18/38	00:00	   3,216,066	9,241,000	+6.52	Nxf5 gxf5 Rg3 h5 Qe1 Re8 Rg1 Kd7 Rf2 Kc7 Rff1 Rge7 Qg3 Re6 Rd1 Bf3 Rde1 Bd5 Rd1 Bf3
 19/38	00:00	   4,391,831	10,096,000	+6.52	Nxf5 gxf5 Rg3 h5 Qe1 Re8 Rg1 Kd7 Rf2 Kc7 Rff1 Rge7 Qg3 Re6 Rd1 Bf3 Rde1 Bd5 Rd1 Bf3
 20/38	00:00	   6,973,349	11,320,000	+6.50	Nxf5 gxf5 Rg3 h5 Qe1 Re8 Rg1 Kd7 Rf2 Kc7 Rff1 Rge7 Qg3 Re6 Rd1 Bf3 Rde1 Bd5 Ra1 Bf3 Qf2 R6e7 Rab1 Re6 Qd2 R6e7 Qf2 Re6
 21/38	00:00	   9,725,151	11,702,000	+6.50	Nxf5 gxf5 Rg3 h5 Qe1 Re8 Rg1 Kd7 Rf2 Kc7 Rff1 Rge7 Qg3 Re6 Rd1 Rg8 Rd2 Rge8 Rdd1 Rg8
 22/40	00:01	  16,709,442	12,187,000	+6.53	Nxf5 gxf5 Rg3 h5 Qe1 Re8 Rg1 Kd7 Rf2 Kc7 Rff1 Rge7 Qg3 Re6 Rd1 Rg8 Ra1 Rge8 Rab1 Bf3 Rbe1 Bd5 Re2 Bf3 Ree1 Bd5
 23/41	00:02	  27,766,371	12,357,000	+6.53	Nxf5 gxf5 Rg3 h5 Qe1 Re8 Rg1 Kd7 Rf2 Kc7 Rff1 Rge7 Qg3 Re6 Rd1 Rg8 Ra1 Rb8 Rae1 Rg8 Rb1 Rge8 Ra1 Bf3 Rae1 Bd5 Qf2 Bf3 Ra1 R6e7 Qg3 Re6 Rae1
 24/41	00:03	  43,082,465	12,560,000	+6.51	Nxf5 gxf5 Rg3 h5 Qe1 Re8 Rg1 Kd7 Rf2 Kc7 Rff1 Rge7 Qg3 Re6 Rb1 Bf3 Rbe1 Bd5 Qf2 Bf3 Rb1 R6e7 Ra1 Re6 Qg3 Bd5 Rab1 Bf3
 25/47	00:05	  64,545,853	13,058,000	+6.51	Nxf5 gxf5 Rg3 h5 Qe1 Re8 Rg1 Kd7 Rf2 Kc7 Rff1 Rge7 Qg3 Re6 Rb1 Bf3 Rbe1 Bd5 Qf2 Bf3 Rb1 R6e7 Ra1 Bd5 Qg3 Re6 Qh2 Bf3 Qg3 Bd5 Qh2
 26/48	00:07	  99,403,444	13,088,000	+6.51	Nxf5 gxf5 Rg3 h5 Qe1 Re8 Rg1 Kd7 Rf2 Kc7 Rff1 Rge7 Qg3 Re6 Rb1 Bf3 Rbe1 Bd5 Qf2 Bf3 Ra1 Bd5 Qg3 Bf3 Rae1 Bd5
 27/48	00:12	 163,886,246	13,330,000	+6.52	Nxf5 gxf5 Rg3 h5 Qe1 Re8 Rg1 Kd7 Rf2 Kc7 Rff1 Rge7 Qg3 Re6 Rb1 Bf3 Rbe1 Bd5 Qf2 Bf3 Rb1 Rc8 Qe1 Rg8 Ra1 Rge8 Qg3 Bd5 Rae1 R8e7 Rb1 Re8 Rbd1 Rd8 Ra1 Rde8 Rae1
 28/59	01:34	1,207,708,107	12,785,000	+6.61	Nxf5 gxf5 Rg3 h5 Re2 Re7 Re1 Rge8 Rf1 Rg8 Kc1 Rf7 Kd1 Re7 Ke1 Reg7 Kf2 Kd7 Qe2 Rg6 Qe1 Kc7 Qd2 R6g7 Ra1 Kb6 Re1 Kc7 Reg1 Kb6 Ra1 Kc7 Rc1 Kb6 Rcg1 Re7 Rf1 Kc7 Rgg1 Bf3 Kg3 Kb6 Ra1
 29/60	02:56	2,300,445,135	13,028,000	+6.75	Nxf5 gxf5 Rg3 h5 Re2 Re7 Re1 Rge8 Rf1 Rg8 Qf2 Reg7 Re1 Re7 Rgg1 Kd7 Qg3 Kc7 Kc1 Rge8 Kd1 Rg8 Kd2 Bf3 Kc1 Rge8 Kd2 Re6 Ref1 Bd5 Ke2 Kb6 Ke1 Rg8 Kf2 Rg7 Re1 Rge7 Rb1 Bf3 Rbe1 Bd5 Rb1
 30/60	03:28	2,742,116,139	13,138,000	+6.75	Nxf5 gxf5 Rg3 h5 Re2 Re7 Re1 Rge8 Qe2 Kd7 Rgg1 Kc7 Qf2 Re6 Kc1 R6e7 Qg3 Bf3 Kd2 Bd5 Ke2 Rg8 Kf2 Reg7 Rb1 Kb6 Rbd1 Bf3 Rdf1 Bd5 Rd1 Bf3
 31/60	04:34	3,603,004,333	13,114,000	+6.75	Nxf5 gxf5 Rg3 h5 Re2 Re7 Re1 Rge8 Qe2 Kd7 Rgg1 Kc7 Qf2 Re6 Kc1 R6e7 Kd2 Bf3 Qh2 Rg8 Qg3 Rd8 Ref1 Bd5 Ke1 Rg8 Kf2 Reg7 Rc1 Bf3 Rce1 Bd5 Rb1 Bf3 Rbf1 Bd5 Rc1
 32/62	06:41	5,246,954,446	13,058,000	+6.75	Nxf5 gxf5 Rg3 h5 Re2 Re7 Re1 Rge8 Qe2 Kd7 Rgg1 Kc7 Qf2 Re6 Kc1 R6e7 Qg3 Kb6 Ref1 Rg8 Kd2 Kc7 Ke1 Kb7 Kf2 Reg7 Rb1 Kc7 Rbe1 Kb6 Ra1 Kb7 Rad1 Kb6 Rde1 Kc7 Rb1 Bf3 Rbf1 Bd5 Rc1 Bf3 Rce1 Bd5 Rb1
 33/63	11:32	9,191,256,907	13,275,000	+6.75	Nxf5 gxf5 Rg3 h5 Re2 Re7 Re1 Rge8 Qe2 Kd7 Rgg1 Kc7 Qf2 Re6 Kc1 R6e7 Qg3 Kb6 Ref1 Rg8 Kd2 Kc7 Ke1 Kb7 Kf2 Reg7 Rb1 Kc7 Rbe1 Kb6 Ra1 Kb7 Rad1 Kb6 Rde1 Kc7 Rb1 Bf3 Rbf1 Bd5 Rc1 Bf3 Rce1 Bd5 Rb1
 34/65	19:22	15,588,253,811	13,407,000	+6.75	Nxf5 gxf5 Rg3 h5 Re2 Re7 Re1 Rge8 Qe2 Kd7 Rgg1 Kc7 Qf2 Re6 Kc1 R6e7 Qg3 Bf3 Kd2 Bd5 Ref1 Kd7 Ke2 Rg7 Kf2 Reg8 Re1 Kc7 Rc1 Bf3 Rce1 Bd5 Rc1
 35/68	38:19	31,337,941,004	13,626,000	+6.75	Nxf5 gxf5 Rg3 h5 Re2 Re7 Re1 Rge8 Qe2 Kd7 Rgg1 Kc7 Qf2 Re6 Kc1 R6e7 Qg3 Bf3 Kd2 Bd5 Ref1 Rg8 Rf2 Rge8 Kc1 Rg8 Rh2 Rge8 Rf2 Rg8
whereas gxh3 loses.

The score +6.75, which would indicate -6.75 for black on the previous ply should really be zero after the key move.

You will see that gxh3 has a score that drops with nearly every iteration after a certain point.
Taking ideas is not a vice, it is a virtue. We have another word for this. It is called learning.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
Dann Corbit
Posts: 12550
Joined: Wed Mar 08, 2006 8:57 pm
Location: Redmond, WA USA

Re: deep.09

Post by Dann Corbit »

This is the correct solution:

[d]6r1/6r1/2p1k1pp/p1pbP2q/Pp1p1PpP/1P1P2NR/1KPQ3R/8 b - - acd 36; acs 2299; acn 31337941004; ce 0; bm Qf5; pm Qf5; pv Qf5 Nxf5 gxf5 Rg3 h5 Re2 Re7 Re1 Rge8 Qe2 Kd7 Rgg1 Kc7 Qf2 Re6 Kc1 R6e7 Qg3 Bf3 Kd2 Bd5 Ref1 Rg8 Rf2 Rge8 Kc1 Rg8 Rh2 Rge8 Rf2 Rg8;
Taking ideas is not a vice, it is a virtue. We have another word for this. It is called learning.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
jhellis3
Posts: 546
Joined: Sat Aug 17, 2013 12:36 am

Re: deep.09

Post by jhellis3 »

Did you even look at my post Dann... there is no fortress...
Dann Corbit
Posts: 12550
Joined: Wed Mar 08, 2006 8:57 pm
Location: Redmond, WA USA

Re: deep.09

Post by Dann Corbit »

jhellis3 wrote:Did you even look at my post Dann... there is no fortress...
I guess I don't understand your post.

[d]6r1/6r1/2p1k1pp/p1pbP2q/Pp1p1PpP/1P1P2NR/1KPQ3R/8 b - -

In this pv, where does white deviate to win:
Qf5 Nxf5 gxf5 Rg3 h5 Re2 Re7 Re1 Rge8 Qe2 Kd7 Rgg1 Kc7 Qf2 Re6 Kc1 R6e7 Qg3 Bf3 Kd2 Bd5 Ref1 Rg8 Rf2 Rge8 Kc1 Rg8 Rh2 Rge8 Rf2 Rg8;
Taking ideas is not a vice, it is a virtue. We have another word for this. It is called learning.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
jhellis3
Posts: 546
Joined: Sat Aug 17, 2013 12:36 am

Re: deep.09

Post by jhellis3 »

White does this... as in my post...

[d]1r6/6r1/2p1k3/p1pbPp1p/Pp1pRPpP/1P1P2K1/2P2Q2/4R3 b - - 37 22
Dann Corbit
Posts: 12550
Joined: Wed Mar 08, 2006 8:57 pm
Location: Redmond, WA USA

Re: deep.09

Post by Dann Corbit »

jhellis3 wrote:White does this... as in my post...

[d]1r6/6r1/2p1k3/p1pbPp1p/Pp1pRPpP/1P1P2K1/2P2Q2/4R3 b - - 37 22
OK, let's play it out.

I assume we start like this:

[Event "Edited position"]
[Site "DCORBIT"]
[Date "2016.08.23"]
[Round "-"]
[White "-"]
[Black "-"]
[Result "*"]
[FEN "6r1/6r1/2p1k1pp/p1pbP2q/Pp1p1PpP/1P1P2NR/1KPQ3R/8 b - - 0 1"]
[SetUp "1"]

{--------------
. . . . . . r .
. . . . . . r .
. . p . k . p p
p . p b P . . q
P p . p . P p P
. P . P . . N R
. K P Q . . . R
. . . . . . . .
black to play
--------------}
1... Qf5 2. Nxf5 gxf5
*

Your move?
Taking ideas is not a vice, it is a virtue. We have another word for this. It is called learning.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
jhellis3
Posts: 546
Joined: Sat Aug 17, 2013 12:36 am

Re: deep.09

Post by jhellis3 »

There is no need for that... black can not do anything but shuffle and can not prevent the white setup. The only hope is better black rook placement. So pick the two most ideal square on black's side of the board for black's rooks and go from there.

I will check guarding the d-file... it might work (hold a draw)... but it is still not probably not a fortress.
jhellis3
Posts: 546
Joined: Sat Aug 17, 2013 12:36 am

Re: deep.09

Post by jhellis3 »

Ok, I did finally find a set up for black which appears to hold (King on c7, bishop on e6, rooks on b and d files), if he can achieve it, but I do not think there is time for that....

Here is the current line I have from here:

[d]6r1/6r1/2p1k3/p1pbPp1p/Pp1p1PpP/1P1P2R1/1KPQ3R/8 w - - 0 4

Qf5 Nxf5 gxf5 Rg3 h5 Re2 Re7 Re1 Rge8 Re3

[d]4r3/4r3/2p1k3/p1pbPp1p/Pp1p1PpP/1P1PR3/1KPQ4/4R3 b - - 5 6

I would advise not taking the rook or moving the king....