Test Position bm Rxc3

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.
chrisw
Posts: 1853
Joined: Tue Apr 03, 2012 2:28 pm

Re: Test Position bm Rxc3

Post by chrisw » Tue May 07, 2019 4:24 pm

chrisw wrote:
Tue May 07, 2019 3:39 pm
MikeGL wrote:
Tue May 07, 2019 2:53 pm
chrisw wrote:
Tue May 07, 2019 11:17 am
30 Ra1 looks better, after Rc2 it seems black just gets a huge grip on the position
I tried to run analysis with MultiPV and just ends up with B vs R endgame, but a pawn up and slightly better for black.
Better space and possible K activity compared to K of white, but might require deeper analysis.

e4 is absolute key move, cutting blacks ability to pivot on that square.

30. Ra1 Bd3 31. e4 dxe4 32. Qa4 Nb3 33. Ra2 is better than the above Rad1 (which threatens nothing and allows black to improve his position)

after 33. Ra2 Nc1 is just drawing, Nc5 also a draw. NxN is draw, some positions are like your line, but a vital tempo ahead (various defences, including Ke3 and counter exchange sac, the tempo is important here)
a6 would be an attempt to keep pieces on the board, but it isn't enough and the position holds, I think.

Key was e4, it gave the king a blockade square on e3, prevented the mating threat lines with Qe4 and gained just enough time. Therefore draw, not decisive black win by strangulation.
Some fun lines in these variations .......
position after 30. Ra1 Bd3 31. e4 dxe4 32. Qa4 Nb3 33. Ra2 a6 (trying to keep white in a bind) 34. Ke3 Rb5

can white play Qxa6? Try without your engine ....


User avatar
MikeGL
Posts: 873
Joined: Thu Sep 01, 2011 12:49 pm

Re: Test Position bm Rxc3

Post by MikeGL » Thu May 09, 2019 1:23 am

How about 36.Nf1 instead of e4?
After deviation 31.Qa1 Na4 32.Nf1 Fish doesn't see any progress for Black anymore.
(
1. Qb3 Nc6 2. Bg5 f6 3. Bd2 Qd7 4. Nc3 O-O-O 5. Na4 Bd6 6. e3 Nge7 7. a3 h5 8.
Bb4 Kb8 9. Bxd6 cxd6 10. Nc3 h4 11. Nd2 Qe6 12. Ba6 Na5 13. Qb5 Rd7 14. Qxa5
bxa6 15. Qxa6 h3 16. g3 Rb7 17. Ra2 g5 18. O-O Rc8 19. Re1 Bg6 20. Qf1 Nc6 21.
Raa1 Ka8 22. Ra2 Kb8 23. f3 Na5 24. Qf2 f5 25. f4 g4 26. Qe2 Be8

27. Kf2 Rxc3 28. bxc3 Bb5 29. Qd1 Bc4 30. Rc2 Bd3 31. Ra2 Nc4 32. Qa1 Kc7 33. Kg1 a5 34. Qd1
Nb2

35. Qa1 Na4 36. Nf1
)


Analysis by CF OPT 180419 x64 POP NUMA:

36...De4 37.Tf2 Sb6 38.a4 Sc4 39.Sd2 De8 40.e4 Lxe4 41.Tef1 Tb6 42.Tc1 Dc6 43.Te1 Kd7 44.Tc1 Ke7 45.Te1
= (0.00) Tiefe: 54/34 00:06:21 8123MN

After some Forward- Backward with 24 threads of 12x3GHz CPU, 32G hash
Problem with this position is that after Rxc3 of black the score is not correctly given by AB engines.
27-41 moves... eval negative (total of 15 moves negative score according to AB engines)
42-47 eval ... 0.00 (total of 6 moves showing it is equal)
48... first gain in score was found. (or 21 moves away after Rxc3 before the +score for black would be detected by AB)
That's what happened in actual game, which means the winning moves are hidden with negative or 0.00 scores.
My point is that, 0.00 shown by AB engines doesn't always end up drawn after watching more than a hundred games between AB - NN engines.

Back to your line, after Nf1 black might continue with Bxf1 and the resulting position just resembles the actual game, almost similar with a boa constrictor play of black.
Complete game with your Nf1 possibility inserted at move 35 and black still has positional dominance to compensate for the exchange sac:


peter
Posts: 1760
Joined: Sat Feb 16, 2008 6:38 am
Full name: Peter Martan

Re: Test Position bm Rxc3

Post by peter » Thu May 09, 2019 6:11 am

MikeGL wrote:
Thu May 09, 2019 1:23 am
Complete game with your Nf1 possibility inserted at move 35 and black still has positional dominance to compensate for the exchange sac:
The line you give after 35.Qa1 Na4 36.Nf1
(
35. Qa1 Na4 36. Nf1 Bxf1 37. Kxf1 Qe4 38. Kg1 Rb3
)
is only 5 plies long, Mike, after those I have this:


Analysis by Cfish 060519 64 POPCNT NUMA:

39.Tf2 Sxc3 40.a4 Kd7 41.Dc1 Sxa4 42.Da1 Tb4 43.Da3 Ke6 44.Td1 Kf7 45.Te1 Kf6 46.Dc1 Tc4 47.Dd2 Sc3 48.Tc1 Sb5 49.Txc4 dxc4 50.Dxa5 Db1+ 51.Tf1 Dc2 52.Dd8+ Kf7 53.Dd7+ Kf8 54.Dc8+ Ke7 55.Db7+ Ke6 56.Dc8+ Kf7
= (0.00) Tiefe: 60/36 00:06:50 10444MN

How does Black continue to take advantage of its "positional dominance"? It can get the white a- and c- Pawn, but then the free black a-Pawn doesn't make it against White's quality of Rook against Night. Not as far as I can see. In best or worst case, depending from which side you look at it, White can give back quality in exchange against this remaining pawn at a- line. So what?
And then there are still the Queens on board, which gives White a second one chance to come into Black's back- ranks chasing King into checks repeated, as seen in output- line above, after Black exchanged Rooks and gave a- Pawn against setting free the one on c-line.

In the game after 35.Dc1(?), which might well be the latest game changing blunder, some other suboptimal moves followed up like 38.Rf2(?), even here Qa1 instead would have given better resistance still, making the real dominance by Black at this point then decisive in a hopeless way from White's side. Yet at this point (38.ff) White's already in Zugzwang, which is the real reason for SF not to see the danger coming sooner and better.

But up to this 35th move of White, the position is equal. White can build up a fortress, if it doesn't miss the chance for it.

What output and eval does LC0 show there at your hardware?
I have only weak installation of Leela, so better try out with yours, if you'd like to have some playout from there onwards, I'll be glad to check your LC0- lines with SF at my hardware with CPU much stronger then GPU.

27...Rxc3 is a good and a beautiful move, but after all, it's just setting up a trap. So I wouldn't call it a game changing single one best move.
Peter.

chrisw
Posts: 1853
Joined: Tue Apr 03, 2012 2:28 pm

Re: Test Position bm Rxc3

Post by chrisw » Thu May 09, 2019 7:58 am

Neither has this line been refuted...

e4 is absolute key move, cutting blacks ability to pivot on that square.

30. Ra1 Bd3 31. e4 dxe4 32. Qa4 Nb3 33. Ra2 is better than the above Rad1 (which threatens nothing and allows black to improve his position)

after 33. Ra2 Nc1 is just drawing, Nc5 also a draw. NxN is draw, some positions are like your line, but a vital tempo ahead (various defences, including Ke3 and counter exchange sac, the tempo is important here)
a6 would be an attempt to keep pieces on the board, but it isn't enough and the position holds, I think.

Key was e4, it gave the king a blockade square on e3, prevented the mating threat lines with Qe4 and gained just enough time. Therefore draw, not decisive black win by strangulation.

User avatar
MikeGL
Posts: 873
Joined: Thu Sep 01, 2011 12:49 pm

Re: Test Position bm Rxc3

Post by MikeGL » Thu May 09, 2019 8:08 am

Sorry, as already posted in reply to jp, this game is from Pohl and I don't have GPU too. Reason why I mentioned difficulty in analzying as moves should be sifted from Multi-PV output on possible Lc0 reply. So I'm basically using AB engine to search for possible NN moves.
Analysis by CF OPT 180419 x64 POP NUMA:

36...De4 37.Tf2 Sb6 38.a4 Sc4 39.Sd2 De8 40.e4 Lxe4 41.Tef1 Tb6 42.Tc1 Dc6 43.Te1 Kd7 44.Tc1 Ke7 45.Te1
= (0.00) Tiefe: 54/34 00:06:21 8123MN

After some Forward- Backward with 24 threads of 12x3GHz CPU, 32G hash
What happens after your above line 36...De4 37.Tf2, then
37...Bxf1 38.Rexf1 Rb3 because I traced with AB Multi-PV playing only best moves from both sides and I ended up with a win for black.
I told my wife that a husband is like a fine wine; he gets better with age. The next day, she locked me in the cellar.

peter
Posts: 1760
Joined: Sat Feb 16, 2008 6:38 am
Full name: Peter Martan

Re: Test Position bm Rxc3

Post by peter » Thu May 09, 2019 8:19 am

MikeGL wrote:
Thu May 09, 2019 8:08 am
What happens after your above line 36...De4 37.Tf2, then
37...Bxf1 38.Rexf1 Rb3 because I traced with AB Multi-PV playing only best moves from both sides and I ended up with a win for black.
Why not show this complete win then, Mike?
Or, even better, refute the output- line from my previous post, continuing just from the moves you had given before in your .pgn and now again up to 38...Rb3
(
rn1qkbnr/ppp2ppp/8/3p1b2/3P4/5N2/PP2PPPP/RNBQKB1R w KQkq - 0 1

1. Qb3 Nc6 2. Bg5 f6 3. Bd2 Qd7 4. Nc3 O-O-O 5. Na4 Bd6 6. e3 Nge7 7. a3 h5 8.
Bb4 Kb8 9. Bxd6 cxd6 10. Nc3 h4 11. Nd2 Qe6 12. Ba6 Na5 13. Qb5 Rd7 14. Qxa5
bxa6 15. Qxa6 h3 16. g3 Rb7 17. Ra2 g5 18. O-O Rc8 19. Re1 Bg6 20. Qf1 Nc6 21.
Raa1 Ka8 22. Ra2 Kb8 23. f3 Na5 24. Qf2 f5 25. f4 g4 26. Qe2 Be8 27. Kf2 Rxc3


28. bxc3 Bb5 29. Qd1 Bc4 30. Rc2 Bd3 31. Ra2 Nc4 32. Qa1 Kc7 33. Kg1 a5 34. Qd1
Nb2

35. Qa1 Na4 36. Nf1 Bxf1 37. Kxf1 Qe4 38. Kg1 Rb3
)
peter wrote:
Thu May 09, 2019 6:11 am

Analysis by Cfish 060519 64 POPCNT NUMA:

39.Tf2 Sxc3 40.a4 Kd7 41.Dc1 Sxa4 42.Da1 Tb4 43.Da3 Ke6 44.Td1 Kf7 45.Te1 Kf6 46.Dc1 Tc4 47.Dd2 Sc3 48.Tc1 Sb5 49.Txc4 dxc4 50.Dxa5 Db1+ 51.Tf1 Dc2 52.Dd8+ Kf7 53.Dd7+ Kf8 54.Dc8+ Ke7 55.Db7+ Ke6 56.Dc8+ Kf7
= (0.00) Tiefe: 60/36 00:06:50 10444MN
BTW sorry for german abbrevations, here translated to English:

39. Rf2 Nxc3 40. a4 Kd7 41. Qc1 Nxa4 42. Qa1 Rb4 43. Qa3 Ke6 44. Rd1 Kf7 45.
Re1 Kf6 46. Qc1 Rc4 47. Qd2 Nc3 48. Rc1 Nb5 49. Rxc4 dxc4 50. Qxa5 Qb1+ 51. Rf1
Qc2 52. Qd8+ Kf7 53. Qd7+ Kf8 54. Qc8+ Ke7 55. Qb7+ Ke6 56. Qc8+ Kf7
Peter.

User avatar
MikeGL
Posts: 873
Joined: Thu Sep 01, 2011 12:49 pm

Re: Test Position bm Rxc3

Post by MikeGL » Thu May 09, 2019 8:41 am

chrisw wrote:
Thu May 09, 2019 7:58 am
Neither has this line been refuted...

e4 is absolute key move, cutting blacks ability to pivot on that square.

30. Ra1 Bd3 31. e4 dxe4 32. Qa4 Nb3 33. Ra2 is better than the above Rad1 (which threatens nothing and allows black to improve his position)

after 33. Ra2 Nc1 is just drawing, Nc5 also a draw. NxN is draw, some positions are like your line, but a vital tempo ahead (various defences, including Ke3 and counter exchange sac, the tempo is important here)
a6 would be an attempt to keep pieces on the board, but it isn't enough and the position holds, I think.

Key was e4, it gave the king a blockade square on e3, prevented the mating threat lines with Qe4 and gained just enough time. Therefore draw, not decisive black win by strangulation.
I traced this and too many 0.00 but strange how I got +score for black at the end, maybe there are improvements here:
30. Ra1 Bd3 31. e4 dxe4 32. Qa4 Rb2 33. Qxa5 Rxd2+ 34. Ke3 Rb2 35. Qd8+ Kb7 36. Rad1 Rb3 37. Rxd3 Qc4 38. Red1 exd3 39.
Qxd6 d2 {-/+}
If you check with multi-pv infinite you would agree that I just traced and followed the top moves for white but ended a win for black.
Maybe some drawing lines are hidden.
I told my wife that a husband is like a fine wine; he gets better with age. The next day, she locked me in the cellar.

User avatar
MikeGL
Posts: 873
Joined: Thu Sep 01, 2011 12:49 pm

Re: Test Position bm Rxc3

Post by MikeGL » Thu May 09, 2019 8:46 am

Why not show this complete win then, Mike?
Or, even better, refute the output- line from my previous post, continuing just from the moves you had given before in your .pgn and now again up to 38...Rb3
(
rn1qkbnr/ppp2ppp/8/3p1b2/3P4/5N2/PP2PPPP/RNBQKB1R w KQkq - 0 1

1. Qb3 Nc6 2. Bg5 f6 3. Bd2 Qd7 4. Nc3 O-O-O 5. Na4 Bd6 6. e3 Nge7 7. a3 h5 8.
Bb4 Kb8 9. Bxd6 cxd6 10. Nc3 h4 11. Nd2 Qe6 12. Ba6 Na5 13. Qb5 Rd7 14. Qxa5
bxa6 15. Qxa6 h3 16. g3 Rb7 17. Ra2 g5 18. O-O Rc8 19. Re1 Bg6 20. Qf1 Nc6 21.
Raa1 Ka8 22. Ra2 Kb8 23. f3 Na5 24. Qf2 f5 25. f4 g4 26. Qe2 Be8 27. Kf2 Rxc3


28. bxc3 Bb5 29. Qd1 Bc4 30. Rc2 Bd3 31. Ra2 Nc4 32. Qa1 Kc7 33. Kg1 a5 34. Qd1
Nb2

35. Qa1 Na4 36. Nf1 Bxf1 37. Kxf1 Qe4 38. Kg1 Rb3
)

because I was using my weak smartphone earlier to check those lines, after 36.Nf1 the direct Bxf1 looks incorrect, that's why I traced your PC output just now but the Nb6 (Sb6) on your first post looks not top move of engines, it was only 6th.

(35. Qa1 Na4 36. Nf1 Qe4 37. Rf2 Bxf1 38. Rexf1 Rb3 39. Qa2 Rxc3 40. Re2 Qd3) this is the line I was referring with a win for black. Maybe there were improvements somewhere. This is just the top output of Multi-PV analysis.
I told my wife that a husband is like a fine wine; he gets better with age. The next day, she locked me in the cellar.

peter
Posts: 1760
Joined: Sat Feb 16, 2008 6:38 am
Full name: Peter Martan

Re: Test Position bm Rxc3

Post by peter » Thu May 09, 2019 9:37 am

MikeGL wrote:
Thu May 09, 2019 8:41 am
chrisw wrote:
Thu May 09, 2019 7:58 am
Neither has this line been refuted...

e4 is absolute key move, cutting blacks ability to pivot on that square.

30. Ra1 Bd3 31. e4 dxe4 32. Qa4 Nb3 33. Ra2 is better than the above Rad1 (which threatens nothing and allows black to improve his position)

after 33. Ra2 Nc1 is just drawing, Nc5 also a draw. NxN is draw, some positions are like your line, but a vital tempo ahead (various defences, including Ke3 and counter exchange sac, the tempo is important here)
a6 would be an attempt to keep pieces on the board, but it isn't enough and the position holds, I think.

Key was e4, it gave the king a blockade square on e3, prevented the mating threat lines with Qe4 and gained just enough time. Therefore draw, not decisive black win by strangulation.
I traced this and too many 0.00 but strange how I got +score for black at the end, maybe there are improvements here:
30. Ra1 Bd3 31. e4 dxe4 32. Qa4 Rb2 33. Qxa5 Rxd2+ 34. Ke3 Rb2 35. Qd8+ Kb7 36. Rad1 Rb3 37. Rxd3 Qc4 38. Red1 exd3 39.
Qxd6 d2 {-/+}
If you check with multi-pv infinite you would agree that I just traced and followed the top moves for white but ended a win for black.
Maybe some drawing lines are hidden.
MikeGL wrote:
Thu May 09, 2019 8:41 am
I traced this and too many 0.00 but strange how I got +score for black at the end, maybe there are improvements here:
30. Ra1 Bd3 31. e4 dxe4 32. Qa4 Rb2 33. Qxa5 Rxd2+ 34. Ke3 Rb2 35. Qd8+ Kb7 36. Rad1 Rb3 37. Rxd3 Qc4 38. Red1 exd3 39.
Qxd6 d2 {-/+}
Indeed there seem to be many improvements for both sides, I just start here with 33. backward, after this one 33.Qxa5(?)

1k6/p7/3pq3/Q4p2/3PpPp1/P1Pb2Pp/1r1N1K1P/R3R3 b - - 0 1

Analysis by CF EXT 060519 x64 POP NUMA:

33...e3+ 34.Kg1 Rxd2 35.Qd8+ Kb7 36.Rab1+ Bxb1 37.Rxb1+ Ka6 38.Qa8 Qb3 39.Qc6+ Ka5 40.Re1 Rg2+ 41.Kh1 Rb2 42.Qc7+ Qb6 43.Qd7 Ka6 44.Qa4+ Kb7 45.Kg1 Rg2+ 46.Kh1 Rxh2+ 47.Kg1 Rb2 48.Qd7+ Qc7 49.Qxc7+ Kxc7 50.Rxe3 Kc6 51.Kh1 Kd5 52.a4 Ra2 53.Re8 Rxa4 54.Re2 Ra3 55.Re3 Ra1+ 56.Kh2 Ra2+ 57.Kh1 a5 58.Re8 Ra3 59.Re3 Ra1+ 60.Kh2 Ra2+ 61.Kh1 a4 62.Re8 a3 63.Re3 Ra1+ 64.Kh2 Kc4 65.Re7 Rc1 66.Rc7+ Kd3 67.Ra7 Rc2+ 68.Kh1 a2 69.c4 Kxc4
-+ (-9.63) Depth: 41/85 00:01:55 3136MN

After 32...Rb2:

1k6/p7/3pq3/n4p2/Q2PpPp1/P1Pb2Pp/1r1N1K1P/R3R3 w - - 0 1

Analysis by CF EXT 060519 x64 POP NUMA:

33.Ke3 Nb7 34.Rac1 Qa2 35.Ra1 Qd5 36.c4 Qa5 37.Qxa5 Nxa5 38.c5 d5 39.Rab1 Bxb1 40.Rxb1 Nc4+ 41.Ke2 Rxb1 42.Nxb1 Kc7 43.a4 Kc6 44.Nc3 Na3 45.Kd2 a6 46.Nd1 Nc4+ 47.Ke2 Kd7 48.Ne3 Ke6 49.Nd1
= (0.00) Depth: 52/53 00:04:16 5639MN

That's much more hardware- time than shown,there was some Forward- Backward before copying and much more already now afterwards. I get other lines a lot, now and then some =/+ eval arises but gets 0.00 again with more time or going into the alternative lines.

Typical SF- bevaiour dealing with fortresses, seeing them-losing them out of horizon and hash again, no real progress without blunders in the lines shown in output.

Cfish sticks to 0.00 the rest of the way back to 30.Ra1, gone have some longer Forward Backward of the rest of these 3 moves still.

If I'll find something yet, I'll be back, but for now I'd say, that's as drawn as the rest of my lines, I have as best ones till now.
That's a fortress, if White blocks the black e-Pawn.
Peter.

peter
Posts: 1760
Joined: Sat Feb 16, 2008 6:38 am
Full name: Peter Martan

Re: Test Position bm Rxc3

Post by peter » Thu May 09, 2019 10:04 am

MikeGL wrote:
Thu May 09, 2019 8:46 am
because I was using my weak smartphone earlier to check those lines, after 36.Nf1 the direct Bxf1 looks incorrect, that's why I traced your PC output just now but the Nb6 (Sb6) on your first post looks not top move of engines, it was only 6th.

(35. Qa1 Na4 36. Nf1 Qe4 37. Rf2 Bxf1 38. Rexf1 Rb3 39. Qa2 Rxc3 40. Re2 Qd3) this is the line I was referring with a win for black. Maybe there were improvements somewhere. This is just the top output of Multi-PV analysis.
Don't see an output- line of mine with 37...Bxf1 but after 35. Qa1 Na4 36. Nf1 Qe4 37. Rf2 Bxf1 38. Rexf1 Rb3 39. Qa2 Rxc3 40. Re2 Qd3 it's drawn still.

8/2k5/3p4/p2p1p2/n2P1Pp1/P1rqP1Pp/Q3R2P/5RK1 w - - 0 1

Analysis by CF EXT 060519 x64 POP NUMA:

41.Rfe1 Rb3 42.Qc2+ Qxc2 43.Rxc2+ Kd8 44.Rec1 Nb6 45.Rc7 Nc4 46.Ra7 Rb2 47.Kh1 Kc8 48.Rf7 a4 49.Rxf5 Kd7 50.Rg5 Nd2 51.f5 Ke7 52.Rg8 Kf6 53.Rgc8 Ne4 54.R1c2 Rb3 55.R2c7 Kxf5 56.Kg1 Nd2 57.Kf2 Nc4 58.Re8 Rb2+ 59.Ke1 Nxa3 60.Rf7+ Kg6 61.Rff8 Rxh2 62.Rg8+ Kh7 63.Rh8+ Kg7 64.Rhg8+
= (0.00) Depth: 55/47 00:04:03 5867MN

Till I see a real winning line by myself or from anybody else, I'm outa here, Mike.

Thanks again for showing the position, it's very interesting and can be used well for testing, but 27...Rxc3 isn't a game changing single one best move, and it's drawn.
:)
Peter.

Post Reply