Nezhmetdinov - Chernikov - 1962 - Queen Sac

[Lyudmil Tsvetkov]

btw., as people would want win demonstrated, even if we take the first line suggested by me, even with the simplifications, after b4 gf5 (we did not continue this line) Nf5:

[d]r4k2/ppr2p1p/8/5N2/1P1BP3/2P5/1q4PP/R4RK1 b - - 0 14

Komodo evaluates this as already positive for white only a couple of moves later.

just check it, and you trust engines do not you?

here is the first Komodo selfplay game until white reaches +50cps score:

[pgn][Event "analysis"]
[Site "Microsoft"]
[Date "2016.12.13"]
[Round "?"]
[White "Komodo 8, owner"]
[Black "Komodo 8 "]
[Result "*"]
[Annotator "owner"]
[SetUp "1"]
[FEN "r4k2/ppr2p1p/8/5N2/1P1BP3/2P5/1q4PP/R4RK1 b - - 0 14"]
[PlyCount "18"]
[TimeControl "600+2"]

{512MB, OWNER-PC} 14... a5 {-0.02/15 2} 15. Ne3 {0.03/14 1} Ra6 {0.09/16 2} 16.
bxa5 {0.03/16 1} Rd7 {0.09/18 2} 17. Rad1 {0.08/17 1} Rg6 {0.23/18 2} 18. Rf2 {
0.17/17 1} Qb3 {0.23/18 1} 19. Rfd2 {0.18/18 1} Ke8 {0.18/18 1} 20. Nd5 {0.11/
18 1} f5 {0.09/18 1} 21. Nb6 {0.09/18 1} Rxb6 {0.41/15 1} 22. axb6 {0.36/16 1}
fxe4 {0.42/16 1} 23. Ra1 {0.40/16 1} *

[/pgn]

if you continue more, score will increase further.

[Lyudmil Tsvetkov]

another try, this time with Komodo 10.1. the only move I suggest Komodo is h4:

[pgn][Event "analysis"]
[Site "Microsoft"]
[Date "2016.12.13"]
[Round "?"]
[White "Komodo 10.1, owner"]
[Black "Komodo 10.1 "]
[Result "*"]
[Annotator "owner"]
[SetUp "1"]
[FEN "r4k2/ppr2p1p/8/5N2/1P1BP3/2P5/1q4PP/R4RK1 b - - 0 14"]
[PlyCount "16"]
[TimeControl "600+2"]

{512MB, OWNER-PC} 14... Rc6 {-0.08/16 2} 15. h4 {10} Rg6 {-0.48/20 11} 16. Ne3
{-0.39/19 1} Rg3 {-0.35/16 1} 17. Kh2 {-0.14/17 1} Rxe3 {0.00/17 1} 18. Bxe3 {
-0.08/20 1} Qxc3 {-0.06/17 1} 19. Bh6+ {0.00/19 1} Ke7 {0.00/16 0} 20. Bg5+ {
0.00/18 0} Ke8 {0.06/17 0} 21. Ra5 {0.06/18 1} b6 {0.26/16 1} 22. Raf5 {0.41/
18 1} *

[/pgn]

again, after 20 plies, score rises to some 50cps for white, though initially Komodo believes black has the edge.

[Lyudmil Tsvetkov]

now, from the other main line position we stopped at with John:

[pgn][Event "Blitz 10m+2s"]
[Site "Microsoft"]
[Date "2016.12.13"]
[Round "?"]
[White "Komodo 10.1, owner"]
[Black "Komodo 10.1 "]
[Result "*"]
[Annotator "owner"]
[SetUp "1"]
[FEN "5k2/1p1r1p1p/r2N1Pp1/p1P1P3/3B4/1P5P/4q1P1/1R1R2K1 b - - 0 1"]
[PlyCount "18"]
[TimeControl "600+2"]

{512MB, OWNER-PC} 1... h5 {-0.97/17 2} 2. Re1 {-0.66/20 3} Qd2 {-0.35/21 3} 3.
Red1 {-0.17/23 3} Qc2 {-0.24/21 2} 4. Rdc1 {0.00/23 2} Qd2 {-0.15/23 4} 5. Rd1
{-0.09/24 3} Qb4 {-0.14/21 2} 6. Be3 {-0.13/23 1} Kg8 {-0.17/22 2} 7. Rdc1 {-0.
11/20 1} Qh4 {-0.06/22 2} 8. Rc4 {0.34/21 2} Qg3 {0.52/22 1} 9. Re4 {0.74/23 3}
Ra8 {0.71/21 1} 10. e6 {0.73/22 1} *

[/pgn]

as, you see, Komodo starts with 100 cps black edge and ends with 100cps white edge. it takes 20 plies or so.

so seemingly, every possible position favours white very much or is simply won, but you need some depth.

[beram]

another try, this time with Komodo 10.1. the only move I suggest Komodo is h4:

[pgn][Event "analysis"]
[Site "Microsoft"]
[Date "2016.12.13"]
[Round "?"]
[White "Komodo 10.1, owner"]
[Black "Komodo 10.1 "]
[Result "*"]
[Annotator "owner"]
[SetUp "1"]
[FEN "r4k2/ppr2p1p/8/5N2/1P1BP3/2P5/1q4PP/R4RK1 b - - 0 14"]
[PlyCount "16"]
[TimeControl "600+2"]

{512MB, OWNER-PC} 14... Rc6 {-0.08/16 2} 15. h4 {10} Rg6 {-0.48/20 11} 16. Ne3
{-0.39/19 1} Rg3 {-0.35/16 1} 17. Kh2 {-0.14/17 1} Rxe3 {0.00/17 1} 18. Bxe3 {
-0.08/20 1} Qxc3 {-0.06/17 1} 19. Bh6+ {0.00/19 1} Ke7 {0.00/16 0} 20. Bg5+ {
0.00/18 0} Ke8 {0.06/17 0} 21. Ra5 {0.06/18 1} b6 {0.26/16 1} 22. Raf5 {0.41/
18 1} *

[/pgn]

again, after 20 plies, score rises to some 50cps for white, though initially Komodo believes black has the edge.

This is ridiculous you let Komodo only play till depth 16
In this line after 16. Ne3 Komodo at depth 26 wants to play Rd8 with +0,75 for black!

[beram]

now, from the other main line position we stopped at with John:

[pgn][Event "Blitz 10m+2s"]
[Site "Microsoft"]
[Date "2016.12.13"]
[Round "?"]
[White "Komodo 10.1, owner"]
[Black "Komodo 10.1 "]
[Result "*"]
[Annotator "owner"]
[SetUp "1"]
[FEN "5k2/1p1r1p1p/r2N1Pp1/p1P1P3/3B4/1P5P/4q1P1/1R1R2K1 b - - 0 1"]
[PlyCount "18"]
[TimeControl "600+2"]

{512MB, OWNER-PC} 1... h5 {-0.97/17 2} 2. Re1 {-0.66/20 3} Qd2 {-0.35/21 3} 3.
Red1 {-0.17/23 3} Qc2 {-0.24/21 2} 4. Rdc1 {0.00/23 2} Qd2 {-0.15/23 4} 5. Rd1
{-0.09/24 3} Qb4 {-0.14/21 2} 6. Be3 {-0.13/23 1} Kg8 {-0.17/22 2} 7. Rdc1 {-0.
11/20 1} Qh4 {-0.06/22 2} 8. Rc4 {0.34/21 2} Qg3 {0.52/22 1} 9. Re4 {0.74/23 3}
Ra8 {0.71/21 1} 10. e6 {0.73/22 1} *

[/pgn]

as, you see, Komodo starts with 100 cps black edge and ends with 100cps white edge. it takes 20 plies or so.

so seemingly, every possible position favours white very much or is simply won, but you need some depth.

Well I certainly agree on your last qoute ''you need some depth''
and when you go in depth with SF or K or H this line is just skindeep

Black plays g5! instead of h5 and can allready choose to draw after Rde1 Qd3 Red1 Qe2 but he can also choose to play on with Qc2 Rdc1 Qg6

[d] 5k2/1p1r1p1p/r2N1P2/p1P1P1p1/3B4/1P5P/4q1P1/1R1R2K1 w - - 0 1

[Lyudmil Tsvetkov]

another try, this time with Komodo 10.1. the only move I suggest Komodo is h4:

[pgn][Event "analysis"]
[Site "Microsoft"]
[Date "2016.12.13"]
[Round "?"]
[White "Komodo 10.1, owner"]
[Black "Komodo 10.1 "]
[Result "*"]
[Annotator "owner"]
[SetUp "1"]
[FEN "r4k2/ppr2p1p/8/5N2/1P1BP3/2P5/1q4PP/R4RK1 b - - 0 14"]
[PlyCount "16"]
[TimeControl "600+2"]

{512MB, OWNER-PC} 14... Rc6 {-0.08/16 2} 15. h4 {10} Rg6 {-0.48/20 11} 16. Ne3
{-0.39/19 1} Rg3 {-0.35/16 1} 17. Kh2 {-0.14/17 1} Rxe3 {0.00/17 1} 18. Bxe3 {
-0.08/20 1} Qxc3 {-0.06/17 1} 19. Bh6+ {0.00/19 1} Ke7 {0.00/16 0} 20. Bg5+ {
0.00/18 0} Ke8 {0.06/17 0} 21. Ra5 {0.06/18 1} b6 {0.26/16 1} 22. Raf5 {0.41/
18 1} *

[/pgn]

again, after 20 plies, score rises to some 50cps for white, though initially Komodo believes black has the edge.

This is ridiculous you let Komodo only play till depth 16
In this line after 16. Ne3 Komodo at depth 26 wants to play Rd8 with +0,75 for black!

maybe I just left much lower TC just fro this particular one, others are with higher depth. this does not change the fact though that multiple tries with different engines only bring same results, clearly favouring white.

PS. I hate to ask the mods here, but what happened with their solemn promise to lock the longer benchmark thread, in case in gets continued? we already have a new thread for this, and I guess in the last 24 hours at least couple of messages have been posted in the old thread.

Are not (solemn) mod promises to be abided by? Should we trust our mods?

[Lyudmil Tsvetkov]

here another one: this time without suggesting Komodo any moves:

[pgn][Event "Shootout (Komodo864-bit, Blitz 10m+2s) "]
[Site "?"]
[Date "2016.12.14"]
[Round "?"]
[White "Komodo 10.1, owner"]
[Black "Komodo 10.1 "]
[Result "*"]
[Annotator "owner"]
[SetUp "1"]
[FEN "r4k2/ppr2p1p/8/5N2/1P1BP3/2P5/1q4PP/R4RK1 b - - 0 1"]
[PlyCount "14"]

1... Rc6 2. Ne3 b5 3. Rab1 Qa2 4. Ra1 Qb2 5. Rab1 Qa2 6. Rbe1 Rd8 7. Nd5 Rd7 8.
Rf2 {0.30/19 12} *

[/pgn]

again, score changes in white favour.

one thing you can not understand, Bram, is that you can not prove that a won position is drawn or lost: in this way you would simply do a disservice the position, the commnuity, as well as yourself.

and the white position is won, Nezhmetdinov's move was objectively the best, as well as an elegant solution.

I would suggest to you and maybe some other people, to do the following: check the position the way I am doing it: instead of doing a full search from the root position with engines, simply force move by move from them with the go or whatever button, leaving the engine a couple of seconds' thought before pushing again. In this way you can suggest move to the engine from time to time in critical positions, moves the engine itself does not see, apart from that the engine's horizon will come closer with each new ply.

this is a much better strategy than just leaving the engine calculate at the root; even 2 seconds search with renewed searches on every ply is much much more efficient than a full seacrh from the root, as, as many experienced authors will tell you, the more distant from the root you get, the more inaccurate the search becomes.

so, please just follow my advice, and you will see that in at least 80% of cases the score turns to white's advantage.

it would not be good to do a disservice to the position, especially this position.

[Lyudmil Tsvetkov]

now, from the other main line position we stopped at with John:

[pgn][Event "Blitz 10m+2s"]
[Site "Microsoft"]
[Date "2016.12.13"]
[Round "?"]
[White "Komodo 10.1, owner"]
[Black "Komodo 10.1 "]
[Result "*"]
[Annotator "owner"]
[SetUp "1"]
[FEN "5k2/1p1r1p1p/r2N1Pp1/p1P1P3/3B4/1P5P/4q1P1/1R1R2K1 b - - 0 1"]
[PlyCount "18"]
[TimeControl "600+2"]

{512MB, OWNER-PC} 1... h5 {-0.97/17 2} 2. Re1 {-0.66/20 3} Qd2 {-0.35/21 3} 3.
Red1 {-0.17/23 3} Qc2 {-0.24/21 2} 4. Rdc1 {0.00/23 2} Qd2 {-0.15/23 4} 5. Rd1
{-0.09/24 3} Qb4 {-0.14/21 2} 6. Be3 {-0.13/23 1} Kg8 {-0.17/22 2} 7. Rdc1 {-0.
11/20 1} Qh4 {-0.06/22 2} 8. Rc4 {0.34/21 2} Qg3 {0.52/22 1} 9. Re4 {0.74/23 3}
Ra8 {0.71/21 1} 10. e6 {0.73/22 1} *

[/pgn]

as, you see, Komodo starts with 100 cps black edge and ends with 100cps white edge. it takes 20 plies or so.

so seemingly, every possible position favours white very much or is simply won, but you need some depth.

Well I certainly agree on your last qoute ''you need some depth''
and when you go in depth with SF or K or H this line is just skindeep

Black plays g5! instead of h5 and can allready choose to draw after Rde1 Qd3 Red1 Qe2 but he can also choose to play on with Qc2 Rdc1 Qg6

[d] 5k2/1p1r1p1p/r2N1P2/p1P1P1p1/3B4/1P5P/4q1P1/1R1R2K1 w - - 0 1

one more try with Komodo, following the approach I specified with renewed searches, bringing the horizon closer:

[pgn][Event "analysis"]
[Site "Microsoft"]
[Date "2016.12.14"]
[Round "?"]
[White "Komodo 10.1 "]
[Black "Komodo 10.1, owner"]
[Result "*"]
[Annotator "owner"]
[SetUp "1"]
[FEN "5k2/1p1r1p1p/r2N1P2/p1P1P1p1/3B4/1P5P/4q1P1/1R1R2K1 w - - 0 1"]
[PlyCount "27"]
[TimeControl "600+2"]

{512MB, OWNER-PC} 1. Re1 {-0.76/20 4} Qc2 {-0.87/20 4} 2. Rec1 {-0.84/20 4} Qg6
{-0.77/20 3} 3. Rd1 {-0.82/19 4} Qc2 {-0.64/19 4} 4. Rdc1 {-0.45/20 3} Qg6 {-0.
71/22 6} 5. Rd1 {-0.66/20 3} Ra8 {-0.70/19 2} 6. Rbc1 {-0.54/20 3} h5 {-0.59/
20 2} 7. Rc4 {-0.53/21 3} Kg8 {-0.65/21 4} 8. Bf2 {-0.66/19 3} Rf8 {-0.59/21 3}
9. Rcc1 {-0.53/22 3} Kh8 {-0.53/22 2} 10. Rd4 {-0.47/22 1} Kh7 {-0.38/20 3} 11.
Be3 {-0.46/20 3} Rb8 {-0.54/20 3} 12. Kh2 {-0.44/20 2} Rdd8 {-0.05/17 3} 13.
Rcc4 {0.00/19 2} Re8 {0.07/19 4} 14. Re4 {0.20/20 3} *

[/pgn]

as you see, only 26 plies later the score gets from negative to positive for white, very deep search, you can not see that from the root with normal search.

I hope you see how white gets the edge; and it should get the edge each and every time, simply because white position is much much superior. I get white plus or significant white plus in more than 80% of cases, and I absolutely have no reason to tell the untruth.

black can not win this, no matter what it plays; on the contrary, white can win in an awful lot of lines.

[Milos]

again, score changes in white favour.

one thing you can not understand, Bram, is that you can not prove that a won position is drawn or lost: in this way you would simply do a disservice the position, the commnuity, as well as yourself.

and the white position is won, Nezhmetdinov's move was objectively the best, as well as an elegant solution.

I would suggest to you and maybe some other people, to do the following: check the position the way I am doing it: instead of doing a full search from the root position with engines, simply force move by move from them with the go or whatever button, leaving the engine a couple of seconds' thought before pushing again. In this way you can suggest move to the engine from time to time in critical positions, moves the engine itself does not see, apart from that the engine's horizon will come closer with each new ply.

You are really worse troll than ARB. Do you really think that ppl are stupid and don't know how to analyse with engines so they need you to come and show them score after 10 moves played at bullet????
That is not even a half of one sample of MC. Do you even have an idea what MC means????
The only thing that matches your ignorance is your arrogance.

Ok, lets solve this once and for all. We play the position from the 12.Qxf6, you play white, I play black. One day per move TC.
If you win provide me with you account number and I will transfer you 1000 (ONE THOUSAND) EUROS.
However, if I manage to draw, I will provide you my bank account number and you will transfer me 100 (ONE HUNDRED) EUROS.
So basically I am giving you 1:10 odds and once in a life time opportunity to earn 1000 Euros in the easiest possible way. You basically won it, just need to demonstrate it over the board. No ifs and buts, enough of this ridiculous "analysis" of yours, you just need to provide your moves.
So my reply is 12...Ne2. Your turn.
[d]r1b2rk1/pp1ppp1p/5Qp1/q7/4P3/1BN1B3/PPP1nPPP/R4RK1 w - - 1 13

So please, either accept this challenge and put money where your mouth is, or shut up and stop trolling this forum forever. I think this is more than fair offer!

[yanquis1972]

hi lyudmil, been a bit awol w/ a bad cold but good to see theres still interest/debate here.

i have to say tho i agree w/ bram, i simply don't think you are letting komodo reach sufficient depth. i certainly havent done deep analysis, but letting it reach even what i'd consider fairly low depth from these positions (pretty fast TC on quad core PC, say), i don't think you'll see these + evals for white. i can at least say that i never did, even after playing (according to the machine) objectively inferior moves to reduce material & hopefully prove a draw more speedily. granted i was almost never able to suggest superior alternative moves or come up with a plan myself which is where your & others input is extremely helpful.

but, doing my best to execute (with engine help) any plans, even allowing them with black, the machine would usually find an escape route. its definitely a troublesome position but given enough depth i still believe its an easy draw for an engine w/ black.

will try to do some engine v engine from these positions myself to show you what i mean; or perhaps ill be surprised myself.