Endgame

[tpoppins]

According to 'Yet another chess problem database', this problem is from 1981:

http://yacpdb.org/#286033

Matouš, Mário
L'Italia Scacchistica (346) 1981
3rd Prize

Solution:

1. Nc2+ $1 (1. Nxb3+ $2 Kb1) (1. Be5 $2 Kb1 $1) 1...
bxc2 2. Rf1+ $1 Bxf1 3. Kc1 Re7 4. Bd2 Rc7 5. Bb4 $1
Rc8 $1 6. Be7 Rc6 7. Bf8 Rg6 8. Bc5 Rg4 9. Ba3 1-0

I do not know what $ symbol stands for in this context.

Numeric Annotation Glyphs

[tpoppins]

YACPDB's solution leads to mate in 11. However, on the second move Black is not forced to take the rook. Then ...Kb2 seems to be best, which can drag things out until move 40.

[zullil]

I do not understand why any engine can´t solve this endgame.
[d]8/6r1/8/8/2bN1B2/1p3R2/p2K4/k7 w - - 0 1
White play and win.
All the engines I used only find a draw.

Cfish-dev with 6-man endgame tables finds a win in 5 seconds:

info depth 33 seldepth 8 multipv 1 score cp 0 nodes 15619042 nps 3273745 hashfull 3 tbhits 574916 time 4771 pv d4b3 a1b1 f3c3 g7d7 d2e1 c4b3 c3b3
info depth 34 currmove d4b3 currmovenumber 1
info depth 34 currmove d4c2 currmovenumber 2
info depth 34 seldepth 20 multipv 1 score cp 530 lowerbound nodes 15686519 nps 3281698 hashfull 3 tbhits 578367 time 4780 pv d4c2
info depth 34 currmove d4c2 currmovenumber 1
info depth 34 seldepth 20 multipv 1 score cp 851 lowerbound nodes 15687481 nps 3281899 hashfull 3 tbhits 578399 time 4780 pv d4c2
info depth 34 currmove d4c2 currmovenumber 1
info depth 34 seldepth 20 multipv 1 score cp 978 lowerbound nodes 15688465 nps 3282105 hashfull 3 tbhits 578436 time 4780 pv d4c2
info depth 34 currmove d4c2 currmovenumber 1
info depth 34 seldepth 20 multipv 1 score cp 996 lowerbound nodes 15694984 nps 3282782 hashfull 3 tbhits 578767 time 4781 pv d4c2
info depth 34 currmove d4c2 currmovenumber 1
info depth 34 seldepth 20 multipv 1 score cp 1021 lowerbound nodes 15745095 nps 3285018 hashfull 3 tbhits 581180 time 4793 pv d4c2
info depth 34 currmove d4c2 currmovenumber 1
info depth 34 seldepth 20 multipv 1 score cp 1450 lowerbound nodes 15770222 nps 3288203 hashfull 3 tbhits 582569 time 4796 pv d4c2
info depth 34 currmove d4c2 currmovenumber 1
info depth 34 seldepth 20 multipv 1 score cp 15382 lowerbound nodes 15772063 nps 3287901 hashfull 3 tbhits 582651 time 4797 pv d4c2
info depth 34 currmove d4c2 currmovenumber 1
info depth 34 currmove d4b3 currmovenumber 2
info depth 34 currmove f4e5 currmovenumber 3
info depth 34 currmove f3c3 currmovenumber 4
info depth 34 currmove d4c6 currmovenumber 5
info depth 34 currmove f3d3 currmovenumber 6
info depth 34 currmove f3f2 currmovenumber 7
info depth 34 currmove f3g3 currmovenumber 8
info depth 34 currmove d2e3 currmovenumber 9
info depth 34 currmove f3f1 currmovenumber 10
info depth 34 currmove d4e2 currmovenumber 11
info depth 34 currmove f3e3 currmovenumber 12
info depth 34 currmove d4e6 currmovenumber 13
info depth 34 currmove f4b8 currmovenumber 14
info depth 34 currmove d2c3 currmovenumber 15
info depth 34 currmove f3h3 currmovenumber 16
info depth 34 currmove f4g5 currmovenumber 17
info depth 34 currmove f4g3 currmovenumber 18
info depth 34 currmove d4b5 currmovenumber 19
info depth 34 currmove d4f5 currmovenumber 20
info depth 34 currmove d2e1 currmovenumber 21
info depth 34 currmove f4d6 currmovenumber 22
info depth 34 currmove d2d1 currmovenumber 23
info depth 34 currmove f4c7 currmovenumber 24
info depth 34 currmove f4e3 currmovenumber 25
info depth 34 currmove f4h6 currmovenumber 26
info depth 34 currmove d2c1 currmovenumber 27
info depth 34 currmove f4h2 currmovenumber 28
info depth 34 currmove f3b3 currmovenumber 29
info depth 34 seldepth 48 multipv 1 score cp 15387 nodes 22039653 nps 4323195 hashfull 5 tbhits 824523 time 5098 pv d4c2 b3c2 f3f1 a1b2 f4e5 b2b3 f1f3 b3a4 e5g7 c4b3 f3f4 a4b5 d2c1 b5c5 g7c3 c5c6 f4b4 b3e6 c1c2 e6f5 c2b2 f5b1 b2b3 c6c5 b4a4 c5d5 a4a5 d5c6 b3b2 c6b6 b2c1 b6c6 c3d4 c6d6 c1b2 d6c6 a5c5 c6d7 d4e5 d7e6 e5b8 e6f6 b2a1 f6e7

[thewhip]

I want to thank all those who contributed to solve this endgame.
The contributions have helped me to use engines that I had not used before.
I will continue looking for endgames that it is difficult to find the solution to the engines.

[zullil]

I want to thank all those who contributed to solve this endgame.
The contributions have helped me to use engines that I had not used before.
I will continue looking for endgames that it is difficult to find the solution to the engines.

Here are some positions that engines find very difficult!

[chessmobile]

Naum 4.2 finds Nc2+ very fast at depth 19. This with default settings and no TB access

[MikeGL]

True. But interesting to note that the toughest ones on that suite have the same study theme.
The trapping of either Q, K or both of them and then maneouvering your own king to gobble pawns for the win.
I remember a similar theme, first posted at Rybka forum:

[d]4q1kr/p6p/1prQPppB/4n3/4P3/2P5/PP2B2P/R5K1 w - - 0 1

Similar theme, a trapped K+R and then Q and R defending its trapped (almost mated) K.
These types of themes are the most difficult it seems, partly because the solution requires long PV.

I want to thank all those who contributed to solve this endgame.
The contributions have helped me to use engines that I had not used before.
I will continue looking for endgames that it is difficult to find the solution to the engines.

Here are some positions that engines find very difficult!

[tpoppins]

I remember a similar theme, first posted at Rybka forum:

4q1kr/p6p/1prQPppB/4n3/4P3/2P5/PP2B2P/R5K1 w - - 0 1

Similar theme, a trapped K+R and then Q and R defending its trapped (almost mated) K.
These types of themes are the most difficult it seems, partly because the solution requires long PV.

Gusev vs. Auerbach, Chelyabinsk 1946