Gligoric-Smyslov, 1959

[zullil]

Qd7

[d]8/3q3k/r5p1/3nPr1p/1BpPQP1P/p1P3K1/R7/5R2 w - - 0 2

SF likes Rxa3:
[d]8/3q3k/r5p1/3nPr1p/1BpPQP1P/R1P3K1/8/5R2 b - - 0 2

[Lyudmil Tsvetkov]

Qd7

[d]8/3q3k/r5p1/3nPr1p/1BpPQP1P/p1P3K1/R7/5R2 w - - 0 2

SF likes Rxa3:
[d]8/3q3k/r5p1/3nPr1p/1BpPQP1P/R1P3K1/8/5R2 b - - 0 2

OK, after Ra3, I recapture Ra3 and, supposing SF will take again with the bishop, Ba3, I play Nc3:

[d]8/3q3k/6p1/4Pr1p/2pPQP1P/B1n3K1/8/5R2 w - - 0 4

Well, play opportunities are exhausted at this point, when you ask SF, if it agrees with the above line and the draw, then maybe, before I ask you to allow me one more chance to win the game (you know I am usually better in my second tries ) , probably you could briefly tell us where could black have gotten a lead in score, as I remember that at one point you suggested SF already exhibited a slight plus for black.

[peter]

probably you could briefly tell us where could black have gotten a lead in score, as I remember that at one point you suggested SF already exhibited a slight plus for black.

In other words you ask Stockfish, where it thinks your winning chances against Stockfish would have been better before?

[Lyudmil Tsvetkov]

probably you could briefly tell us where could black have gotten a lead in score, as I remember that at one point you suggested SF already exhibited a slight plus for black.

In other words you ask Stockfish, where it thinks your winning chances against Stockfish would have been better before?

No, I want to illustrate black got the advantage at some point, according to SF.

SF defended accurately this time, but not the next one...

[zullil]

OK, after Ra3, I recapture Ra3 and, supposing SF will take again with the bishop, Ba3, I play Nc3:

[d]8/3q3k/6p1/4Pr1p/2pPQP1P/B1n3K1/8/5R2 w - - 0 4

Well, play opportunities are exhausted at this point, when you ask SF, if it agrees with the above line and the draw, then maybe, before I ask you to allow me one more chance to win the game (you know I am usually better in my second tries ) , probably you could briefly tell us where could black have gotten a lead in score, as I remember that at one point you suggested SF already exhibited a slight plus for black.

SF gives the following in response to Nxc3:

Code: Select all

info depth 55 seldepth 29 multipv 1 score cp 0 nodes 40173838784 nps 21782194 hashfull 958 tbhits 33583012 time 1844343 pv e4f3 d7d4 f1e1 c3d5 e1e4 d4g1 g3h3 c4c3 e5e6 d5f6 e4e3 c3c2 f3b7 h7h8 b7c8 h8g7 c8c7 g7h8 c7c8

So we agree to a draw?

I have not keep any log of this game. But I know that SF has not evaluated in favor of Black in a very long time. I think the last such position was this:

[d]b7/1r3q1k/r2B2pb/p2nPp1p/2pP1P1P/2P3QN/R1B4K/5R2 w - - 0 3

I'll let SF work on this, to see if it evaluates in favor of Black again.

[Lyudmil Tsvetkov]

OK, after Ra3, I recapture Ra3 and, supposing SF will take again with the bishop, Ba3, I play Nc3:

[d]8/3q3k/6p1/4Pr1p/2pPQP1P/B1n3K1/8/5R2 w - - 0 4

Well, play opportunities are exhausted at this point, when you ask SF, if it agrees with the above line and the draw, then maybe, before I ask you to allow me one more chance to win the game (you know I am usually better in my second tries ) , probably you could briefly tell us where could black have gotten a lead in score, as I remember that at one point you suggested SF already exhibited a slight plus for black.

SF gives the following in response to Nxc3:

Code: Select all

info depth 55 seldepth 29 multipv 1 score cp 0 nodes 40173838784 nps 21782194 hashfull 958 tbhits 33583012 time 1844343 pv e4f3 d7d4 f1e1 c3d5 e1e4 d4g1 g3h3 c4c3 e5e6 d5f6 e4e3 c3c2 f3b7 h7h8 b7c8 h8g7 c8c7 g7h8 c7c8

So we agree to a draw?

I have not keep any log of this game. But I know that SF has not evaluated in favor of Black in a very long time. I think the last such position was this:

[d]b7/1r3q1k/r2B2pb/p2nPp1p/2pP1P1P/2P3QN/R1B4K/5R2 w - - 0 3

I'll let SF work on this, to see if it evaluates in favor of Black again.

A draw is a draw, nothing to do above.

So, please, Louis, just do one more run on the position after Rb7, maybe at sufficiently big depth, in order for us to be clear what SF thinks about it.

Btw., was this the only time SF showed black advantage, and what was its size in centipawns?

I mean, when did the positive score for black disappear, on the very next move, 2 moves later on, or so?

[zullil]

So, please, Louis, just do one more run on the position after Rb7, maybe at sufficiently big depth, in order for us to be clear what SF thinks about it.

Btw., was this the only time SF showed black advantage, and what was its size in centipawns?

I mean, when did the positive score for black disappear, on the very next move, 2 moves later on, or so?

I'm sorry, but I do not have a log of evaluations. Last time, SF played Rg1 here, with a small score in favor of Black. This I know because the information appears in a post in this thread---page 3, I think. Don't recall any other scores in favor of Black at all. In the current search of the position after Rb7, SF prefers Ba4 at depth 46, with a score of 0.00. But the search is still running...

[zullil]

[d]b7/1r3q1k/r2B2pb/p2nPp1p/2pP1P1P/2P3QN/R1B4K/5R2 w - - 0 3

So, please, Louis, just do one more run on the position after Rb7, maybe at sufficiently big depth, in order for us to be clear what SF thinks about it.

Decided to re-search the position above using MultiPV=3. This time (with a hash full of positions from the game just played), SF sees no advantage at all for either side:

Code: Select all

info depth 46 seldepth 63 multipv 1 score cp 0 nodes 54435179701 nps 20889149 hashfull 936 tbhits 4366357 time 2605907 pv c2a4 a6b6 d6a3 b7b8 a2e2 b6b1 e2f2 b8b6 a4c2 b1f1 f2f1 d5b4 c2a4 b4d3 h3g5 h6g5 g3g5 a8e4 g5d8 f7b7 d8g5 b7f7

info depth 46 seldepth 63 multipv 2 score cp 0 nodes 54435179701 nps 20889149 hashfull 936 tbhits 4366357 time 2605907 pv f1a1 a5a4 c2a4 a6a4 a2a4 b7b2 h2g1 d5c3 g3c3 f7d5 a4a7 a8b7 a7b7 b2b7 c3d2 b7b3 a1a7 h7h8 a7a3 b3a3 d6a3 d5f3 d2g2 f3a3 g2g6 a3a1 g1g2 a1a8 g2f2 a8a2 f2e3 a2b3 e3f2 b3b2 f2e3 b2b3

info depth 46 seldepth 63 multipv 3 score cp 0 nodes 54435179701 nps 20889149 hashfull 936 tbhits 4366357 time 2605907 pv d6c5 a5a4 c2a4 b7b8 f1a1 f7b7 a1g1 b7f7 g1a1

Based on this "backward analysis", my conclusion is the position above is a draw. Will now back up to your original position and search one more time...

[peter]

No, I want to illustrate black got the advantage at some point, according to SF.

SF defended accurately this time, but not the next one...

Yet I don't understand you quite well, Lyudmil.
If you want to illustrate that SF evaluates the position incorrectly, why than start over again just where it gives you the best chances to with its eval?

Everybody saw you just wait for SF's blunder to turn this dead draw to a win, so your wishful thinking shall be encouraged by SF itself now.

That's pure gold-digging, Lyudmil, but you'll find plate only again, if you don't believe me, why not ask Stockfish?


Seriously, couldn't you do so on your own? Just let it run for that time longer Louis engine is stronger than yours.

If you at first had the intention to move only every forty days, your game should become rather high-classed too anyway.

[Lyudmil Tsvetkov]

[d]b7/1r3q1k/r2B2pb/p2nPp1p/2pP1P1P/2P3QN/R1B4K/5R2 w - - 0 3

So, please, Louis, just do one more run on the position after Rb7, maybe at sufficiently big depth, in order for us to be clear what SF thinks about it.

Decided to re-search the position above using MultiPV=3. This time (with a hash full of positions from the game just played), SF sees no advantage at all for either side:

Code: Select all

info depth 46 seldepth 63 multipv 1 score cp 0 nodes 54435179701 nps 20889149 hashfull 936 tbhits 4366357 time 2605907 pv c2a4 a6b6 d6a3 b7b8 a2e2 b6b1 e2f2 b8b6 a4c2 b1f1 f2f1 d5b4 c2a4 b4d3 h3g5 h6g5 g3g5 a8e4 g5d8 f7b7 d8g5 b7f7

info depth 46 seldepth 63 multipv 2 score cp 0 nodes 54435179701 nps 20889149 hashfull 936 tbhits 4366357 time 2605907 pv f1a1 a5a4 c2a4 a6a4 a2a4 b7b2 h2g1 d5c3 g3c3 f7d5 a4a7 a8b7 a7b7 b2b7 c3d2 b7b3 a1a7 h7h8 a7a3 b3a3 d6a3 d5f3 d2g2 f3a3 g2g6 a3a1 g1g2 a1a8 g2f2 a8a2 f2e3 a2b3 e3f2 b3b2 f2e3 b2b3

info depth 46 seldepth 63 multipv 3 score cp 0 nodes 54435179701 nps 20889149 hashfull 936 tbhits 4366357 time 2605907 pv d6c5 a5a4 c2a4 b7b8 f1a1 f7b7 a1g1 b7f7 g1a1

Based on this "backward analysis", my conclusion is the position above is a draw. Will now back up to your original position and search one more time...

Thanks Louis.

No need to go to the original position, I see no way black can improve its play.

But I bet the position after Rb7 is won for black.
Maybe we give it one more try, even if nothing else, at least we will certainly learn some useful things in the process.

If your SF kindly accepts, I would be very happy if you could do another search of the position after black's Rb7, and post SF's reply, then I will give my, to see where this leads us the second time.

I bet the second time SF will have a much much more difficult time.
No one can convince me this position is a draw.