Possiblity of an cooked study

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

Moderators: hgm, Harvey Williamson, bob

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
Post Reply
User avatar
Nordlandia
Posts: 2215
Joined: Fri Sep 25, 2015 7:38 pm
Location: Sortland, Norway

Possiblity of an cooked study

Post by Nordlandia » Fri Mar 15, 2019 3:06 pm

It is possible that this study is unsound.

Usually RN vs RBN is a tablebase win. There are some chances of that RNP vs RBN is a tablebase win. I'd suspect analysing that endgame with 7-piece syzygy can give us better answer than analysing it with 6-piece syzygy.

J. Rusinek | 1st price L' Italia Scacchistica 1976




Longest RN v RBN endgame - 2N1nB2/8/4k3/8/6r1/7R/3K4/8 b - - | White mates in 231.

peter
Posts: 1711
Joined: Sat Feb 16, 2008 6:38 am
Full name: Peter Martan

Re: Possiblity of an cooked study

Post by peter » Sat Mar 16, 2019 12:03 am

Nordlandia wrote:
Fri Mar 15, 2019 3:06 pm
I'd suspect analysing that endgame with 7-piece syzygy can give us better answer than analysing it with 6-piece syzygy.
I don't think so, Jon, if you cannot find a way to reach 7men position at all without losing instead of winning. Author's theme is a fortress, if you want to refute it, you simply have to give a line with any progress at all, with our without tbs. Or you just have to get 50- moves boundary into horizon and then keep it in hash backwarding. Of course that's a long way, so you won't make it fully, but probably not much better with 7men Syzygys instead of 6.

8/8/1r5k/8/7K/3RP3/2b1n3/1N6 w - - 0 1

Analysis by CF EXT 120319 x64 POP NUMA:

1.Rd2 Rb4+ 2.Kh3 Bf5+ 3.Kg2 Rg4+ 4.Kf2 Ng1 5.Rd6+ Kg7 6.Nd2 Nh3+ 7.Kf3 Rg1 8.Rb6 Bg6 9.e4 Bh5+ 10.Ke3 Re1+ 11.Kd4 Nf2 12.Ke5 Bg6 13.Rb7+ Kh6 14.Rb6 Re2 15.Rd6 Nh3 16.Rd5 Bh7 17.Rd3 Ng5 18.Kf4 Rf2+ 19.Ke3 Rf7 20.Rd5 Rf8 21.Rb5 Bg6 22.Rb6 Re8 23.Ra6 Rc8 24.Kf4 Rf8+ 25.Ke3 Re8 26.Rc6 Rb8 27.Kf4 Ra8 28.Kg4 Rf8 29.Ra6
-/+ (-0.85) Depth: 50/56 00:02:04 2977MN, tb=69724990

Black will try to win of course 50 moves long after last pawn-move.

So positions like this one



won't arise before at least 50 moves and there Black can exchange the Bishop against the Knight reaching a remis- 7men position or go on shuffling.

http://www.yacpdb.org/#search/ODgxcjVrO ... MS8xLzA=/1
Peter.

User avatar
Nordlandia
Posts: 2215
Joined: Fri Sep 25, 2015 7:38 pm
Location: Sortland, Norway

Re: Possiblity of an cooked study

Post by Nordlandia » Sat Mar 16, 2019 9:24 am

Peter: not all EG studies enforce 50-move rule. This study can be an exception. The likelihood for black to winning is minimally.

User avatar
Nordlandia
Posts: 2215
Joined: Fri Sep 25, 2015 7:38 pm
Location: Sortland, Norway

Re: Possiblity of an cooked study

Post by Nordlandia » Sat Mar 16, 2019 11:16 am

Peter: here 7-piece syzygy will definitely give better answer than 6-piece syzygy.

No 19384 | V. Kirillov & E. Kudelich | Honourable mention | EG October 2013 Supplement, page 398



Possible cook :arrow:

Winning 8-man EGTB position.

http://www.viewchess.com/cbreader/2017/ ... 23453.html

peter
Posts: 1711
Joined: Sat Feb 16, 2008 6:38 am
Full name: Peter Martan

Re: Possiblity of an cooked study

Post by peter » Sat Mar 16, 2019 11:46 am

Nordlandia wrote:
Sat Mar 16, 2019 9:24 am
Peter: not all EG studies enforce 50-move rule.
Of course not, Jon, just the fortresses among them often do so
:)
Peter.

User avatar
Nordlandia
Posts: 2215
Joined: Fri Sep 25, 2015 7:38 pm
Location: Sortland, Norway

Re: Possiblity of an cooked study

Post by Nordlandia » Sat Mar 16, 2019 1:03 pm

Peter: this is an classic when the 50-move rule saves white.


8/1p6/1p6/kPp2P1K/2P5/N1Pp4/q2P4/1N6 w - - 0 1



peter
Posts: 1711
Joined: Sat Feb 16, 2008 6:38 am
Full name: Peter Martan

Re: Possiblity of an cooked study

Post by peter » Sat Mar 16, 2019 3:42 pm

Nordlandia wrote:
Sat Mar 16, 2019 11:16 am
No 19384 | V. Kirillov & E. Kudelich | Honourable mention | EG October 2013 Supplement, page 398



Possible cook :arrow:

Winning 8-man EGTB position.

http://www.viewchess.com/cbreader/2017/ ... 23453.html
You brought that here not long ago.

viewtopic.php?p=790544#p790544

Thanks for the names of the authors now, but the link to the cbreader doesn't deal with this one.
Peter.

Paloma
Posts: 733
Joined: Thu Dec 25, 2008 8:07 pm

Re: Possiblity of an cooked study

Post by Paloma » Sat Mar 16, 2019 5:35 pm

> You brought that here not long ago.

Short-term memory :)

peter
Posts: 1711
Joined: Sat Feb 16, 2008 6:38 am
Full name: Peter Martan

Re: Possiblity of an cooked study

Post by peter » Sat Mar 16, 2019 7:14 pm

Paloma wrote:
Sat Mar 16, 2019 5:35 pm
> You brought that here not long ago.

Short-term memory :)
Not quite or not only, rather database- storage too
:)
Peter.

peter
Posts: 1711
Joined: Sat Feb 16, 2008 6:38 am
Full name: Peter Martan

Re: Possiblity of an cooked study

Post by peter » Sat Mar 16, 2019 7:50 pm

Nordlandia wrote:
Sat Mar 16, 2019 11:16 am
Having made a new Backward of the lines, I had stored in February already, I got an even better output now:

1K6/8/2P3P1/1R6/8/5n2/4rrPk/6b1 w - - 0 1

Analysis by CF EXT 120319 x64 POP NUMA:

1.c7 Rxg2 2.c8Q Rxg6 3.Rh5+ Kg3 4.Qh3+ Kf2 5.Rf5 Rg3 6.Qh6 Kg2 7.Qc6 Rg8+ 8.Kb7 Re7+ 9.Ka6 Re3 10.Rc5 Rg4 11.Rc4 Rg7 12.Kb5 Bf2 13.Rc2 Rg4 14.Rc4 Rg5+ 15.Ka4 Ree5 16.Rc3 Be3 17.Rd3 Rg4+ 18.Ka3 Bc5+ 19.Ka2 Re2+ 20.Kb3 Be3 21.Ka3 Kf2 22.Qc7 Re4 23.Qc8 Re7 24.Rc3 Ra7+ 25.Kb3 Raa2 26.Kc4 Ra4+ 27.Kb3 Rf4 28.Ka3 Nd4 29.Rc5 Rf3 30.Kb4 Rb2+ 31.Kc4 Rc2+ 32.Kd5 Rd2 33.Qh8 Ne6+ 34.Kc6 Nxc5 35.Qh2+ Kf1 36.Qh1+ Ke2 37.Kb5
-+ (-52.96) Depth: 58/72 00:05:25 8219MN, tb=215013784

Still with 6men Syzygys only
:)
Peter.

Post Reply