Positional draw?

[zullil]

White has just exchanged a rook for a knight and pawn, believing that he has achieved a "positional draw." Has he, or can Black win this with correct play?

Posting of general strategies and possible lines of play for discussion are welcome, but, please, let's not turn this into another online game.

[d]r5k1/5p2/r2p2pp/2p5/2P1PP1P/1B1P4/P1R3K1/8 b - - 0 41

[pgn]
1. c4 e5 2. Nc3 Nf6 3. e4 Bc5 4. Be2 d6 5. Nf3 Nc6 6. O-O O-O 7. d3 a5 8. Bg5 h6 9. Bd2 Bg4 10. h3 Bxf3 11. Bxf3 Nd4 12. g3 c6 13. Bg2 b5 14. Kh1 b4 15. Ne2 Qe7 16. Nxd4 Bxd4 17. Rb1 Nd7 18. Qe2 Rab8 19. f4 exf4 20. gxf4 Nc5 21. b3 Rb6 22. Qf3 Ra6 23. Qg3 Qf6 24. Rbd1 Re8 25. Be1 Nd7 26. Bf3 c5 27. Bg4 Nf8 28. Rb1 a4 29. Bd1 axb3 30. Bxb3 Rea8 31. Qh4 Ne6 32. Qxf6 Bxf6 33. Kg2 Nd4 34. Bf2 Ne2 35. Be1 Bc3 36. Rf2 Nd4 37. Bxc3 bxc3 38. Rc1 c2 39. h4 g6 40. Rcxc2 Nxc2 41. Rxc2
[/pgn]

[Lyudmil Tsvetkov]

White has just exchanged a rook for a knight and pawn, believing that he has achieved a "positional draw." Has he, or can Black win this with correct play?

Posting of general strategies and possible lines of play for discussion are welcome, but, please, let's not turn this into another online game.

[d]r5k1/5p2/r2p2pp/2p5/2P1PP1P/1B1P4/P1R3K1/8 b - - 0 41

[pgn]
1. c4 e5 2. Nc3 Nf6 3. e4 Bc5 4. Be2 d6 5. Nf3 Nc6 6. O-O O-O 7. d3 a5 8. Bg5 h6 9. Bd2 Bg4 10. h3 Bxf3 11. Bxf3 Nd4 12. g3 c6 13. Bg2 b5 14. Kh1 b4 15. Ne2 Qe7 16. Nxd4 Bxd4 17. Rb1 Nd7 18. Qe2 Rab8 19. f4 exf4 20. gxf4 Nc5 21. b3 Rb6 22. Qf3 Ra6 23. Qg3 Qf6 24. Rbd1 Re8 25. Be1 Nd7 26. Bf3 c5 27. Bg4 Nf8 28. Rb1 a4 29. Bd1 axb3 30. Bxb3 Rea8 31. Qh4 Ne6 32. Qxf6 Bxf6 33. Kg2 Nd4 34. Bf2 Ne2 35. Be1 Bc3 36. Rf2 Nd4 37. Bxc3 bxc3 38. Rc1 c2 39. h4 g6 40. Rcxc2 Nxc2 41. Rxc2
[/pgn]

Thanks for posting this Louis.

I am so tired, I can barely recognise the position.

Lanzo Nazaire shortened my life with at least 10 years.

[peter]

Hi Louis!

If you open a new thread about that and give a link in the other thread to me, I of course have to give my input, having contributed so much to the confusion about that, which I yet am sorry about too.

1.c4 e5 2.Sc3 Sf6 3.e4 Lc5 4.Le2 d6 5.Sf3 Sc6 6.0–0 0–0 7.d3 a5 8.Lg5 h6 9.Ld2 Lg4 10.h3 Lxf3 11.Lxf3 Sd4 12.g3 c6 13.Lg2 b5 14.Kh1 b4 15.Se2 De7 16.Sxd4 Lxd4 17.Tb1 Sd7 18.De2 Tab8 19.f4 exf4 20.gxf4 Sc5 21.b3 Tb6 22.Df3 Ta6 23.Dg3 Df6 24.Tbd1 Te8 25.Le1 Sd7 26.Lf3 c5 27.Lg4 Sf8 28.Tb1 a4 29.Ld1 axb3 30.Lxb3 Tea8 31.Dh4 Se6 32.Dxf6 Lxf6 33.Kg2 Sd4 34.Lf2 Se2 35.Le1 Lc3 36.Tf2 Sd4 37.Lxc3 bxc3 38.Tc1 c2 39.h4 g6 40.Tcxc2 Sxc2 41.Txc2

41...Kf8 42.Kf3 Ke7 43.Tg2 Tg8 44.Ke3 f6 45.Th2 Ta5 46.Kf3 Taa8 47.Th1 h5

48.Kg3 Th8 49.Kf2 g5 50.Kf3 g4+ 51.Kg2 f5 52.Tb1 Thf8 53.Tf1 Tab8 54.Kg3 Tb4 55.Td1 Te8 56.Kf2 Kf6 57.Ke3 Tg8 58.Tg1 g3 59.exf5 Kxf5 60.Kf3 Tb7 61.La4 Tg4 62.Txg3 Txf4+ 63.Ke2 Te7+ 64.Kd2 Txh4 65.Tg2 Ta7 66.Lb3 Tg4 67.Tf2+ Tf4 68.Th2 h4 69.Ke3 Tg4 70.Kf2 Kg5

is the one line I got from SF- playout overnight with 24 threads and 32 Gb hash at 10'/move.

That's what I posted in the other thread too, Lyudmil deviated from this with

48.Ke3
and
48...Th8 49.Th2 g5 50.Ld1 g4 51.f5 Ta3 52.Lb3 Kd7 53.Kf4 Tha8 54.Tc2 T3a6 55.Tb2 Kc7 56.Tb1 Ta5

was the next line of SF- autoplay at 5'/move, I broke that off not seeing as much progress anymore as I had expected to be possible.

Even if I still think Black should make the whole point out of this, I'd like at first to find a better way, not against 48.Ke3 instead of 48.Kg3, but against (above 51.)f5 that could not be played in the first line because of giving check in the moment of (50)...g4+ there.

[d]r6r/4k3/3p1p2/2p3pp/2P1PP1P/3PK3/P6R/3B4 b - - 0 50

That's the position, that makes the difference to the first line. After 50...g4 here in the second one without giving check, White can pass with 51.f5 and the further breakthrough of the kingside- Pawns isn't as easy as before.

After White's 51.f5 in the second line there still is progress in the engine's eval from 1.65 there to 2.15 at the "end" 56...Ra5, but I don't see the positional progress as well as I'd like to.

So here I stand at the moment without any sure prove of Black's winning neither of a positional draw of course, which I still think at least much to much said at move nr. 41 and much less probable to me.
Maybe I should simply go on after 56th move, yet I think there should be something better to be found earlier.

BTW, the moves from 41 to 47 were "played" against Lyudmil, (therefore the second paragraph in the first line) so they should be in discussion too, that was far away from a thorough analysis by my side.

That might be all for today for me, more probably not until tomorrow, any ideas welcome so I maybe wouldn't have to have much more man- and hardware- time by myself for that position , which maybe is of no specific interest except it's one you think you should be able to prove won or drawn but you can't make it easily.
And of course as for engine's weaknesses it's of some interest too, zugzwang is something to me that doesn't appear qualitatively only, it has quantitative aspects for both sides always too, I guess that's a reason of the problem of programming the engines nullmove- pruning fitting for as many positions as possible.
Just look at the end of the first line, White's fully lost there because of zugzwang and it appears all the way down the line as well, even if not as heavily and as clearly to be seen

[zullil]

Hi Louis!

If you open a new thread about that and give a link in the other thread to me, I of course have to give my input,

Yes, that was my plan. Thanks for all the analysis, to get this thread started.

[peter]

I said not until tomorrow, well for me that was yesterday already here at my home.

48.Ke3
and
48...Th8 49.Th2 g5 50.Ld1 g4 51.f5 Ta3 52.Lb3 Kd7 53.Kf4 Tha8 54.Tc2 T3a6 55.Tb2 Kc7 56.Tb1 Ta5

...

[d]r6r/4k3/3p1p2/2p3pp/2P1PP1P/3PK3/P6R/3B4 b - - 0 50

That's the position, that makes the difference to the first line. After 50...g4 here in the second one without giving check, White can pass with 51.f5 and the further breakthrough of the kingside- Pawns isn't as easy as before.

Instead of 53...Rha8 (?!) 53...Kc7!

1.c4 e5 2.Nc3 Nf6 3.e4 Bc5 4.Be2 d6 5.Nf3 Nc6 6.0–0 0–0 7.d3 a5 8.Bg5 h6 9.Bd2 Bg4 10.h3 Bxf3 11.Bxf3 Nd4 12.g3 c6 13.Bg2 b5 14.Kh1 b4 15.Ne2 Qe7 16.Nxd4 Bxd4 17.Rb1 Nd7 18.Qe2 Rab8 19.f4 exf4 20.gxf4 Nc5 21.b3 Rb6 22.Qf3 Ra6 23.Qg3 Qf6 24.Rbd1 Re8 25.Be1 Nd7 26.Bf3 c5 27.Bg4 Nf8 28.Rb1 a4 29.Bd1 axb3 30.Bxb3 Rea8 31.Qh4 Ne6 32.Qxf6 Bxf6 33.Kg2 Nd4 34.Bf2 Ne2 35.Be1 Bc3 36.Rf2 Nd4 37.Bxc3 bxc3 38.Rc1 c2 39.h4 g6 40.Rcxc2 Nxc2 41.Rxc2

41...Kf8 42.Kf3 Ke7 43.Rg2 Rg8 44.Ke3 f6 45.Rh2 Ra5 46.Kf3 Raa8 47.Rh1 h5 48.Ke3 Rh8 49.Rh2 g5 50.Bd1 g4 51.f5 Ra3 52.Bb3 Kd7 53.Kf4

53...Kc7 54.Ke3 Kb6 55.Rb2 Ka5 56.Bd1 Ra8 57.Rb5+ Ka6 58.Rb2 Ra7 59.Bb3 Rg7 60.Rg2 Ka5 61.Rg1 g3!

[d]8/6r1/3p1p2/k1p2P1p/2P1P2P/rB1PK1p1/P7/6R1 w - - 0 62

If 61.Rg3
61...Kb4!

The principle: to prevent White's fortress the black King goes Queenside, threatening to support the re- exchange of the qualtiy against the white a-Pawn, to hold against that White has to come with the Rook and in some lines even with the King to the Qeenside too and that than opens the way for the black g-Pawn to march again, if it is on g3 the way for a black Rook over g4 is open.

Good night, White!

[peter]

And now, Louis, after you have seen my analysis, especially the last one posting,

http://www.talkchess.com/forum/viewtopi ... 14&t=56059

, what do you think after backward solving my lines with your machine?
Is this rather a "positional draw" or a "fortress" or just a Black win?

BTW, I'd gladly play that out against you and your machine, we could do it by private message or email not to start a flame war again before having relevant reults.

You'd have three trials with White, as for you I'm sure you'd resign as soon as you'd see the evals climbing significantly enough, and so we wouldn't need so many moves per game probably, cause I for sure would resign (accept the draw) as soon as we both saw the evals stagnate and didn't see any obvious positional progress anymore.

We can do it at any kind of "blitz- corr." or long TC of your choice, for such educated and arranged single corr.- games I'd always find the time besides of my other games.


But that just in case, you don't see the evals you get by reanalysing my lines as clear as I do.
If you yet would come to the same conclusion as I, just please tell us so, and then I could close this case for me myself too.

I don't mind, if Lyudmil accepts your or my or any engines evals, I don't have the ambition to prove anything to him or anybody else, I'd just like to be sure for myself and I'd like to know your opinion about that position you brought to this second thread, for what I anyhow would like to thank you too.


Edit: of course, if you just want to give the one or the other line as example of positional draw of your own or your SF's analysis, no problem for me to have a look at those just right here and now.
I think we'd come to a shared conclusion about such rather quickly too

[Lyudmil Tsvetkov]

I will look further into your other line, but to see how propensive to different fortress/positional draws the position is, just let me mention a very easy draw I missed in the real game continuation after black's g5:

[d]6r1/5p2/3p1k1p/2p3p1/2P1PP1P/rB1PK3/P5R1/8 w - - 0 12

Here white should play Kf3, and after Raa8 (the black rook on g8 is hanging), fg5 hg5 h5 Ke5 Kg4 Kd4 Rg3, we have this:

[d]r5r1/5p2/3p4/2p3pP/2PkP1K1/1B1P2R1/P7/8 b - - 0 16

which is a positional draw, actually more like a fortress draw.

White just plays Rg3-h3-f3 and back, black can do nothing to penetrate and a sacrifice on b3 does not work.

The above fortress was also possible in other similar lines, it was easy to find, but of course I missed it because of Lanzo's intervention that threw me off balance.

So the line in the game for black is no good.

Overall, from the main diagram position can arise a vast multitude of positional/fortress draws, at least 5 to 10 different varieties. This only underscores how drawish the position is.

I will not say now where the other draws are, as I am sure SF could still fall into one of those.

[zullil]

And now, Louis, after you have seen my analysis, especially the last one posting,

http://www.talkchess.com/forum/viewtopi ... 14&t=56059

, what do you think after backward solving my lines with your machine?
Is this rather a "positional draw" or a "fortress" or just a Black win?

I haven't had time yet to study these lines. I'll return to them when I can.

What I've seen so far suggests that Black wins after all. The general strategy seems to be:
1) create a passed pawn on the kingside, preferably on the g-file.
2) keep White from making any useful break in the center.
3) Bring the black king to b4, to support an exchange on b3 and to penetrate at c3.

[Lyudmil Tsvetkov]

I said not until tomorrow, well for me that was yesterday already here at my home.

48.Ke3
and
48...Th8 49.Th2 g5 50.Ld1 g4 51.f5 Ta3 52.Lb3 Kd7 53.Kf4 Tha8 54.Tc2 T3a6 55.Tb2 Kc7 56.Tb1 Ta5

...

[d]r6r/4k3/3p1p2/2p3pp/2P1PP1P/3PK3/P6R/3B4 b - - 0 50

That's the position, that makes the difference to the first line. After 50...g4 here in the second one without giving check, White can pass with 51.f5 and the further breakthrough of the kingside- Pawns isn't as easy as before.

Instead of 53...Rha8 (?!) 53...Kc7!

1.c4 e5 2.Nc3 Nf6 3.e4 Bc5 4.Be2 d6 5.Nf3 Nc6 6.0–0 0–0 7.d3 a5 8.Bg5 h6 9.Bd2 Bg4 10.h3 Bxf3 11.Bxf3 Nd4 12.g3 c6 13.Bg2 b5 14.Kh1 b4 15.Ne2 Qe7 16.Nxd4 Bxd4 17.Rb1 Nd7 18.Qe2 Rab8 19.f4 exf4 20.gxf4 Nc5 21.b3 Rb6 22.Qf3 Ra6 23.Qg3 Qf6 24.Rbd1 Re8 25.Be1 Nd7 26.Bf3 c5 27.Bg4 Nf8 28.Rb1 a4 29.Bd1 axb3 30.Bxb3 Rea8 31.Qh4 Ne6 32.Qxf6 Bxf6 33.Kg2 Nd4 34.Bf2 Ne2 35.Be1 Bc3 36.Rf2 Nd4 37.Bxc3 bxc3 38.Rc1 c2 39.h4 g6 40.Rcxc2 Nxc2 41.Rxc2

41...Kf8 42.Kf3 Ke7 43.Rg2 Rg8 44.Ke3 f6 45.Rh2 Ra5 46.Kf3 Raa8 47.Rh1 h5 48.Ke3 Rh8 49.Rh2 g5 50.Bd1 g4 51.f5 Ra3 52.Bb3 Kd7 53.Kf4

53...Kc7 54.Ke3 Kb6 55.Rb2 Ka5 56.Bd1 Ra8 57.Rb5+ Ka6 58.Rb2 Ra7 59.Bb3 Rg7 60.Rg2 Ka5 61.Rg1 g3!

[d]8/6r1/3p1p2/k1p2P1p/2P1P2P/rB1PK1p1/P7/6R1 w - - 0 62

If 61.Rg3
61...Kb4!

The principle: to prevent White's fortress the black King goes Queenside, threatening to support the re- exchange of the qualtiy against the white a-Pawn, to hold against that White has to come with the Rook and in some lines even with the King to the Qeenside too and that than opens the way for the black g-Pawn to march again, if it is on g3 the way for a black Rook over g4 is open.

Good night, White!

The line you mention is not optimal on many occasions both for black and white, but, the way you play for black, white achieves a fortress with Rg2-g3 (the g4 pawn should be blocked by the rook and not allowed to move forward), and then moving the king to the queen side to guard b4 penetration square.

[d]1r6/r7/1k1p1p2/2p2P1p/2P1P1pP/1BKP2R1/P7/8 w - - 0 10

This is another positional, actually more like fortress draw, possible from the position on the main diagram.

[Lyudmil Tsvetkov]

And now, Louis, after you have seen my analysis, especially the last one posting,

http://www.talkchess.com/forum/viewtopi ... 14&t=56059

, what do you think after backward solving my lines with your machine?
Is this rather a "positional draw" or a "fortress" or just a Black win?

I haven't had time yet to study these lines. I'll return to them when I can.

What I've seen so far suggests that Black wins after all. The general strategy seems to be:
1) create a passed pawn on the kingside, preferably on the g-file.
2) keep White from making any useful break in the center.
3) Bring the black king to b4, to support an exchange on b3 and to penetrate at c3.

Too many demands on black, white will of course not stay in the same place all the time.