Is there anymore unsolved positions at all!?

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

Moderators: hgm, Rebel, chrisw

zullil
Posts: 6442
Joined: Tue Jan 09, 2007 12:31 am
Location: PA USA
Full name: Louis Zulli

Re: Is there anymore unsolved positions at all!?

Post by zullil »

h1a8 wrote: Sat Aug 22, 2020 5:12 am
Omg. You need to trust me on this. If I'm wrong then please never listen to anything I say. But please give me the benefit of the doubt. I know this position extremely well now. It is not simple at all. Once Bb6 then moving the knight to c6 depends on the position of the black and queen. For example, if the black king is on h7 then moving the knight to c6 is a draw.
Now assume, the black king is on g8 (queen is always on e8) then Nc6 is winning. But the win still isn't simple. Black has got several trucks to draw.

Now don't get me wrong, after Bb6 the cpu can win this against any entity. The problem is that it is not simple for humans. The study says it is mate in 50 after Nxd7

Don't believe me that it isn’t easy? Play 5 games against stockfish with no take backs. How many did you draw?
Sorry for not reading carefully. So you don't doubt the correctness of the study. Your assertion is that finishing correctly from here is not easy for humans.

[d]4q2n/2KN1p1k/1B1PpPp1/4P1P1/8/8/6P1/8 b - - 0 35
zullil
Posts: 6442
Joined: Tue Jan 09, 2007 12:31 am
Location: PA USA
Full name: Louis Zulli

Re: Is there anymore unsolved positions at all!?

Post by zullil »

zullil wrote: Sat Aug 22, 2020 11:56 am
h1a8 wrote: Sat Aug 22, 2020 5:12 am
Omg. You need to trust me on this. If I'm wrong then please never listen to anything I say. But please give me the benefit of the doubt. I know this position extremely well now. It is not simple at all. Once Bb6 then moving the knight to c6 depends on the position of the black and queen. For example, if the black king is on h7 then moving the knight to c6 is a draw.
Now assume, the black king is on g8 (queen is always on e8) then Nc6 is winning. But the win still isn't simple. Black has got several trucks to draw.

Now don't get me wrong, after Bb6 the cpu can win this against any entity. The problem is that it is not simple for humans. The study says it is mate in 50 after Nxd7

Don't believe me that it isn’t easy? Play 5 games against stockfish with no take backs. How many did you draw?
Sorry for not reading carefully. So you don't doubt the correctness of the study. Your assertion is that finishing correctly from here is not easy for humans.

[d]4q2n/2KN1p1k/1B1PpPp1/4P1P1/8/8/6P1/8 b - - 0 35
The stated mate-in-50 seems a bit off:

info depth 93 seldepth 55 multipv 1 score mate -26 nodes 364489056624 nps 73757409 hashfull 1000 tbhits 9364803855 time 4941728 pv e8a8 d7b8 a8g2 d6d7 g2d5 b8c6 d5d2 b6c5 d2g5 d7d8q g5f4 c5d4 f4d2 c7b7 d2c1 c6e7 c1b1 b7c7 b1c1 c7d7 c1d1 d8g8 h7h6 g8h8 h6g5 e7c6 g5g4 h8g7 d1b1 d7c7 g6g5 g7f7 g4g3 f7e7 b1f5 f6f7 f5f1 f7f8q f1f4 f8f4 g5f4 e7g5 g3f3 g5g1 f3e4 g1g6 e4f3 c6b4 f3e2 g6d3 e2e1 b4c2
h1a8
Posts: 508
Joined: Fri Jun 04, 2010 7:23 am

Re: Is there anymore unsolved positions at all!?

Post by h1a8 »

zullil wrote: Sat Aug 22, 2020 11:56 am
Sorry for not reading carefully. So you don't doubt the correctness of the study. Your assertion is that finishing correctly from here is not easy for humans.

[d]4q2n/2KN1p1k/1B1PpPp1/4P1P1/8/8/6P1/8 b - - 0 35
Yes. But I now know how to win after going over all the different variations where black has drawing tricks. The study doesn't cover those wrinkles. Qa8 and queen gets out and has a few tricks (tricks involving Qxg5 and tricks involving where there is no Qxg5).

Actually it starts right before Nc6. If the black Queen is on e8 and black king on h7 when white moves Nc6 then it is a draw. So white has to move the g pawn (or triangulate the king) in order to get queen on e8 and king on 8th rank (or queen away from e8) before moving Nc6.The study doesn't cover that wrinkle.

[d]3Nq2n/1K1p1p1k/1B1PpPp1/4P1P1/8/8/6P1/8 w - - 19 29

Nc6 is a draw

Here's the kicker. Once the king gets to b7 the stockfish still doesn't know exactly what to do. In some cases it pushes the g pawn twice or moves the king around a few times before moving nc6 and winning.
User avatar
MikeB
Posts: 4889
Joined: Thu Mar 09, 2006 6:34 am
Location: Pen Argyl, Pennsylvania

Re: Is there anymore unsolved positions at all!?

Post by MikeB »

Collingwood wrote: Sat Aug 22, 2020 10:18 am <snip>

Please consider using the Post Reply button or editing out stuff if you use the Quote button.
We could use a snip button ...highlight the text you dont want and click "snip" , ...make it easier to snip.
Image
Dann Corbit
Posts: 12541
Joined: Wed Mar 08, 2006 8:57 pm
Location: Redmond, WA USA

Re: Is there anymore unsolved positions at all!?

Post by Dann Corbit »

MikeB wrote: Sat Aug 22, 2020 7:00 pm
Collingwood wrote: Sat Aug 22, 2020 10:18 am <snip>

Please consider using the Post Reply button or editing out stuff if you use the Quote button.
We could use a snip button ...highlight the text you dont want and click "snip" , ...make it easier to snip.
I highlight the text and use the backspace key. It works as a snip button that way.
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.
Alayan
Posts: 550
Joined: Tue Nov 19, 2019 8:48 pm
Full name: Alayan Feh

Re: Is there anymore unsolved positions at all!?

Post by Alayan »

Collingwood wrote: Sat Aug 22, 2020 10:18 am Guys, I don't want to be the formatting police, but it's really hard to read posts when you quote all of the previous posts. The last post has quotes nested 10 deep.

Please consider using the Post Reply button or editing out stuff if you use the Quote button.
This. Badly formatted posts with multiple nested quotes make it a pain to follow a thread.
User avatar
Thomas Lagershausen
Posts: 328
Joined: Mon Jun 11, 2007 6:59 pm

Re: Is there anymore unsolved positions at all!?

Post by Thomas Lagershausen »

The skill to build a positional dominat position.

Anand shows it against Kasparov in game 10 of the wcch 95.

[d]r3rbk1/1p3ppp/1B2qn2/PQ2p3/2P1P3/5B1P/6P1/1R1R3K w - - 1 27 bm Rd5 !
TL
zullil
Posts: 6442
Joined: Tue Jan 09, 2007 12:31 am
Location: PA USA
Full name: Louis Zulli

Re: Is there anymore unsolved positions at all!?

Post by zullil »

Thomas Lagershausen wrote: Sun Aug 23, 2020 10:42 am The skill to build a positional dominat position.

Anand shows it against Kasparov in game 10 of the wcch 95.

[d]r3rbk1/1p3ppp/1B2qn2/PQ2p3/2P1P3/5B1P/6P1/1R1R3K w - - 1 27 bm Rd5 !
Perhaps a comforting move for humans in 1995, but Cfish-dev-PureNNUE says it's about not even in the top 20:

info depth 43 seldepth 85 multipv 1 score cp 249 nodes 35752780122 nps 22944319 hashfull 1000 tbhits 2413981 time 1558241 pv b1c1 e8c8 c4c5 h7h5 d1f1 e6e7 f3e2 f6e4 f1f7 e4g3 h1h2 g8f7 b5b3 f7g6 b3g3 g6h6 g3f3 g7g6 f3e4 e7d7 e2d3 c8c6 d3b5 a8c8 e4e5 h6h7 b5c6 c8c6 c1e1 d7f7 e1e3 f7d7 e3e2 d7f7 e2e1 f7d7 e1e3 f8g7 e5e4 d7d4 e4d4 g7d4 e3e7 h7h6 e7b7 d4c5 b6d8 c6d6 d8c7 d6d3 h3h4 c5f2 c7f4 g6g5 h4g5 h6g6 b7b5 d3a3 g2g3 h5h4 g3h4 f2h4 h2g2 g6h5 b5d5 h4e1 f4c7

info depth 43 seldepth 88 multipv 2 score cp 240 nodes 35752780122 nps 22944319 hashfull 1000 tbhits 2413981 time 1558241 pv c4c5 e8c8 b1c1 h7h5 d1f1 e6e7 f3e2 f6e4 f1f7 e4g3 h1h2 g8f7 b5b3 f7g6 b3g3 g6h6 g3f3 g7g6 f3e4 e7d7 e2d3 c8c6 d3b5 a8c8 e4e5 f8g7 e5e3 h6h7 c1e1 d7d5 b5c6 c8c6 e1e2 d5g8 e3e4 g8b8 h2g1 b8c8 e4d5 g7f6 g1h1 h7h6 e2e1 f6g5 h1g1 g5f6 d5f7 c8b8 e1e4 b8c8 g1h1 h5h4 f7d5 f6g5 e4e1 c8f5 d5f5 g6f5

info depth 43 seldepth 80 multipv 3 score cp 169 nodes 35752780122 nps 22944319 hashfull 1000 tbhits 2413981 time 1558241 pv d1f1 e8c8 c4c5 e6d7 b5b4 h7h6 b1c1 d7d3 f1e1 d3d7 e1d1 d7e7 b4b3 g7g6 d1f1 h6h5 f3e2 f8h6 f1f6 e7f6 c1f1 f6g5 b3f7 g8h8 f7b7 c8f8 e2f3 a8b8 b7a7 g5f6 a7d7 h6e3 b6c7 b8b7 c5c6 b7b2 d7d5 e3f4 c7b6 f8c8 c6c7 b2b6 d5d7 c8c7 d7c7 b6c6 c7b7 f6d6 f1d1 d6c7 a5a6 c6c1 b7c7 c1c7 h3h4 f4e3 f3e2 e3d4 d1f1 h8h7 h1h2 c7c2 e2b5
info depth 44 currmove b1c1 currmovenumber 1
info depth 44 currmove d1f1 currmovenumber 2
info depth 44 currmove d1c1 currmovenumber 3
info depth 44 currmove d1d3 currmovenumber 4
info depth 44 currmove h1h2 currmovenumber 5
info depth 44 currmove b6g1 currmovenumber 6
info depth 44 currmove c4c5 currmovenumber 7
info depth 44 currmove b6f2 currmovenumber 8
info depth 44 currmove b6c7 currmovenumber 9
info depth 44 currmove d1d2 currmovenumber 10
info depth 44 currmove b1b2 currmovenumber 11
info depth 44 currmove b6e3 currmovenumber 12
info depth 44 currmove b5d7 currmovenumber 13
info depth 44 currmove d1d8 currmovenumber 14
info depth 44 currmove b1a1 currmovenumber 15
info depth 44 currmove h1g1 currmovenumber 16
info depth 44 currmove d1d6 currmovenumber 17
info depth 44 currmove b6a7 currmovenumber 18
info depth 44 currmove d1d7 currmovenumber 19
info depth 44 currmove g2g3 currmovenumber 20
info depth 44 currmove b5b4 currmovenumber 21
info depth 44 currmove d1d5 currmovenumber 22
info depth 44 currmove b6d8 currmovenumber 23
info depth 44 currmove b5c6 currmovenumber 24
User avatar
Thomas Lagershausen
Posts: 328
Joined: Mon Jun 11, 2007 6:59 pm

Re: Is there anymore unsolved positions at all!?

Post by Thomas Lagershausen »

Be thankfull.

The position from Anand against Kasparov is the unsolved position you looked for.

It is clear that current Sf-NNUE is too weak to understand it.

Nevermind.
TL
zullil
Posts: 6442
Joined: Tue Jan 09, 2007 12:31 am
Location: PA USA
Full name: Louis Zulli

Re: Is there anymore unsolved positions at all!?

Post by zullil »

Thomas Lagershausen wrote: Sun Aug 23, 2020 4:33 pm Be thankfull.

The position from Anand against Kasparov is the unsolved position you looked for.

It is clear that current Sf-NNUE is too weak to understand it.

Nevermind.
Why do you persist in believing that humans play chess well? :D