Leela crushes any resistance in this position

[jp]

Too bad nothing is at stake because no actual game is being played, so the truth will never be known.

The truth will certainly not be known merely through playing a game, but we can arrange a CC game if there's the will. We do need someone with a respectable machine to help with White, though. It's highly likely to be a draw with best play by White.

[jp]

I don't think it's enough for Black to play accurately. White has also to play inaccurately for Black to win.

Then we can ask how can white address the problem bishop on h2 without compromising the stability of the position. In order for white to being able to hold, it has to be addressed. Otherwise all indications suggest white will remain a piece down practically speaking.

It's Black who needs to break down White's position to win. White just needs to grovel for a draw.

You get to positions like

[d]8/8/2b5/3k4/3B1pP1/p7/6PK/8 w - - 0 21

Draw!


Or this position:

[d]8/pNb5/4k3/P7/5p2/2P2P2/1n1K1P1B/8 b - - 2 24

In this position, the White bishop is still stuck, but it's still a draw!


Wherever you go, you find draws.

[Nordlandia]

It's not necessarily a won position, as long white play precisely, it may turn out that white can hold.

Leela selfplay ended in a quick draw. And SF Dev with -30 surprisingly held against Leela. It seems that negative contempt is required for SF to be able to hold.

[pgn][Event "i7-5960X 4.5GHz / 2x 2080 Supe"]
[Site "i7-5960X 4.5GHz / 2x 2080 Supe"]
[Date "2020.07.03"]
[Round "?"]
[White "lc0 v0.26.0"]
[Black "lc0 v0.26.0"]
[Result "1/2-1/2"]
[SetUp "1"]
[FEN "r4rk1/pp6/1bn1R2p/6pb/5p2/1NP2N1P/PP3PPB/3R2K1 b - - 0 6"]
[PlyCount "9"]
[EventDate "2020.??.??"]
[TimeControl "1800+30"]

6... Rad8 {0.52/19 31} 7. Rxd8 {-0.49/18 31} Rxd8 {0.51/22 32} 8. Nbd2 {
-0.49/21 32} Bf7 {0.51/22 33} 9. Re2 {-0.49/22 39} Bxa2 {0.51/23 40} 10. g3 {
-0.50/24 22} Bc7 {adjudication +0.52/24 0.26s, Draw by adjudication: user
decision} 1/2-1/2

[/pgn]

[pgn][Event "i7-5960X 4.5GHz / 2x 2080 Supe"]
[Site "i7-5960X 4.5GHz / 2x 2080 Supe"]
[Date "2020.07.03"]
[Round "?"]
[White "stockfish_20070107_x64_bmi2"]
[Black "lc0 v0.26.0"]
[Result "1/2-1/2"]
[SetUp "1"]
[FEN "r4rk1/pp6/1bn1R2p/6pb/5p2/1NP2N1P/PP3PPB/3R2K1 b - - 0 6"]
[PlyCount "35"]
[EventDate "2020.??.??"]
[TimeControl "1800+30"]

6... Rad8 {0.51/19 31} 7. Rde1 {-0.35/38 91s} Bxf3 {0.88/14 48} 8. gxf3 {
-0.27/40 48} Kg7 {0.93/14 31} 9. a4 {-0.40/47 165s} Rf6 {1.25/13 73s} 10. h4 {
-0.50/46 66s} Rd5 {1.31/14 63s} 11. Kg2 {-0.97/41 212s} Rxe6 {1.49/15 70s} 12.
Rxe6 {-0.26/41 68s} Bc7 {1.52/14 40} 13. Nd4 {-0.17/55 33} Re5 {1.89/16 77s}
14. Rxe5 {-0.39/56 178s} Bxe5 {1.98/16 41} 15. Nxc6 {-0.39/71 33} bxc6 {
2.41/14 91s} 16. h5 {-0.39/76 73s} Kf6 {3.62/12 181s} 17. Kf1 {-0.39/81 65s} c5
{3.62/12 0.21s} 18. Ke2 {-0.39/88 57} c4 {3.63/11 52} 19. Kf1 {-0.39/89 59} Ke6
{3.86/10 148s} 20. Kg2 {-0.39/91 46} Bd6 {3.86/10 75s} 21. Kh3 {-0.39/90 47}
Bc5 {3.76/11 71s} 22. Bg1 {-0.39/87 54} Kd6 {3.58/12 77s} 23. Kg4 {-0.39/89
357s} Ke5 {adjudication +3.58/13 52s, Draw by adjudication: user decision}
1/2-1/2

[/pgn]

[Nordlandia]

SF -33 here.

[pgn][Event "?"]
[Site "https://lichess.org/gpGRu6ab"]
[Date "2020.07.03"]
[Round "?"]
[White "stockfish_20070107_x64_bmi2"]
[Black "lc0 v0.26.0"]
[Result "0-1"]
[WhiteElo "?"]
[BlackElo "?"]
[Variant "From Position"]
[TimeControl "1800+30"]
[ECO "?"]
[Opening "?"]
[Termination "Unknown"]
[FEN "r4rk1/pp6/1bn1R2p/6pb/5p2/1NP2N1P/PP3PPB/3R2K1 b - - 0 6"]
[SetUp "1"]
[Annotator "lichess.org"]

6... Rad8 7. Rde1 Bxf3 8. gxf3 Kg7 9. h4 Rf6 10. Kg2 Rxe6 11. Rxe6 Rd5 12. a4 Bc7 13. Nd4 Re5 14. Rxe5 Bxe5 15. Nb3 Kg6 16. hxg5 hxg5 17. Kf1 Bd6 18. Nd2 Ne5 19. b4 b6 20. Ke2 Kh5 21. Bg3 Bc7 22. Ne4 fxg3 23. fxg3 Kg6 24. Nd2 Nd7 25. f4 gxf4 26. gxf4 Bxf4 27. Nb3 Ne5 28. Nd4 Kf6 29. Kd1 Ke7 30. Ne2 Be3 { Black wins. } 31. Ng3 0-1[/pgn]

[abgursu]

This position reminds me some Capablanca games.

[Ovyron]

We do need someone with a respectable machine to help with White, though.

Nope. A skilled centaur will have the same performance. Respectable machines are for those that haven't figured out how to analyze with their engines.

[jp]

Nope. A skilled centaur will have the same performance. Respectable machines are for those that haven't figured out how to analyze with their engines.

My point is that we can always organize CC games (the results of which you severely over-value but are still of some value as data), but I (for example) cannot take on White by myself when the computing power available to me right now is less than someone's smartphone.

[jp]

It's not necessarily a won position, as long white play precisely, it may turn out that white can hold.

Leela selfplay ended in a quick draw. And SF Dev with -30 surprisingly held against Leela. It seems that negative contempt is required for SF to be able to hold.

<snip>
bxc6 {2.41/14 91s} 16. h5 {-0.39/76 73s} Kf6 {3.62/12 181s} 17. Kf1 {-0.39/81 65s} c5
{3.62/12 0.21s} 18. Ke2 {-0.39/88 57} c4 {3.63/11 52} 19. Kf1 {-0.39/89 59}
<snip>

Do those numbers mean {evaluation/depth time} or something else?

All your games have Leela Black, right? I guess if SF takes Black and Leela White, White will get crushed?

[Ovyron]

(the results of which you severely over-value but are still of some value as data)

It's not about the result of the game, but all the data accumulated that each player will have, and can tell others about their conclusions.

[jp]

I should have written "products", not "results". It's useful data but not definitive.

It's better to have it than not to have it, so who here can be persuaded to play White? I'm pretty sure this is a draw.