What a pity

[Lyudmil Tsvetkov]

The 36th game of the TCEC final was for me the most interesting up until now, at least in terms of what could have happened.

SF found excellent attacking moves, with Ng3, Ng4, f5, Qe8-h5.

The only thing I wonder is why it did include e5 in the whole plan, wasting some time? Would not it be better to play 14...Qe8 straight after Komodo's 14.e4 (btw., the best move here)?

[d]r1b1qrk1/pp2p1b1/2pp3p/5pp1/2PPP1n1/PQNB1NP1/1P3PP1/R4RK1 w - - 0 2

How does white save this?

My opinion is that black should win this.

Please note, that the game after the suggested continuation seems to be too complicated for me, there are splurges of tactical fireworks everywhere, actually every single move seems to hold on tactics, but my assessment that black could have won is based on the following known facts:

- I judge the opening position as slightly favouring black; black almost certainly captures on g3, doubling important white shelter pawns. At the same time, black does not have obvious minus points in its structure.

- Komodo further played some inaccurate moves, like a3, Qb3, which should additionally aggravate white's situation

Are not the above 2 factors in combination sufficient to claim black should have a winning advantage after a correct play?

After 14...Qe8, followed by Qh5, white is almost mated. Actually, it would have been, if it were not for the white knight on f3. Black should only find a way of removing that knight, either through rook sacrifice on f3, by expelling it with a pawn, or maybe by attacking it with its light-coloured bishop. Other threats will also arise in the process, as white should always mind h2.

Of course, white has better development, some central pressure, etc., but is this enough to hold the game and counter the existing threats? Black can incorporate in the attack many other pieces and alternate tactical threats.

People claim that there are too many drawn games in TCEC. No, I think there are too many missed chances.

In any case, even if you do not find the win for black, I am absolutely certain you will have big time going through the different lines after 14...Qe8, as the position is simply gorgeous. In any case, the best possible position that could have arisen in TCEC so far.

[Henk]

There is no hope for these tactical monsters. Better look at good strategic playing engines that make tactical blunders we don't notice.

[Lyudmil Tsvetkov]

What a pity no one posts.

One might suppose it is Christmas time...

I would not like any concrete replies, the position is simply too complicated, at least for me, there are innumerable reasonable continuations, all leading to unbelievably beautiful and non-standard positions.

In such a case, I would not be inclined to pass a final judgement, although I am pretty sure black has big advantage and excellent winning chances.

I would be just happy if you check some lines for yourselves, not simply expecting engine output with dry scores, but also looking at/replaying some lines yourselves. I stand by my opinion that such positions happen very rarely.

For example, you might want to check what happens after 14...Qe8 15. Rae1 Qh5 16. e5 e6

[d]r1b2rk1/pp4b1/2ppp2p/4Pppq/2PP2n1/PQNB1NP1/1P3PP1/4RRK1 w - - 0 4

Great line, do not you think?

Or, after 16...de5 17. de5 Kh8 in the above main variation

[d]r1b2r1k/pp2p1b1/2p4p/4Pppq/2P3n1/PQNB1NP1/1P3PP1/4RRK1 w - - 0 5

Also not a bad position.

[zullil]

The 36th game of the TCEC final was for me the most interesting up until now, at least in terms of what could have happened.

SF found excellent attacking moves, with Ng3, Ng4, f5, Qe8-h5.

The only thing I wonder is why it did include e5 in the whole plan, wasting some time? Would not it be better to play 14...Qe8 straight after Komodo's 14.e4 (btw., the best move here)?

[d]r1b1qrk1/pp2p1b1/2pp3p/5pp1/2PPP1n1/PQNB1NP1/1P3PP1/R4RK1 w - - 0 2

How does white save this?

SF seems to be unable to see beyond a repetition of position in this line:

Code: Select all

info depth 45 seldepth 55 multipv 1 score cp 0 nodes 35608610123 nps 20293946 tbhits 0 time 1754642 pv c4c5 g8h8 e4f5 c8f5 d3f5 f8f5 b3b7 a8b8 b7c7 b8c8 c7b7 c8b8

[Lyudmil Tsvetkov]

Or, you might want to look at a line like 15.c5 Kh8 16.cd6 ed6 17.ef5 Qh5

[d]r1b2r1k/pp4b1/2pp3p/5Ppq/3P2n1/PQNB1NP1/1P3PP1/R4RK1 w - - 0 5

I do not think this is the strongest line for black, but is just interesting to follow.

Engines might think it is a straightforward draw, or even that white has advantage, but with deeper search the balance should swing to black.

In any case, it is an interesting position to follow tactically.

[Lyudmil Tsvetkov]

The 36th game of the TCEC final was for me the most interesting up until now, at least in terms of what could have happened.

SF found excellent attacking moves, with Ng3, Ng4, f5, Qe8-h5.

The only thing I wonder is why it did include e5 in the whole plan, wasting some time? Would not it be better to play 14...Qe8 straight after Komodo's 14.e4 (btw., the best move here)?

[d]r1b1qrk1/pp2p1b1/2pp3p/5pp1/2PPP1n1/PQNB1NP1/1P3PP1/R4RK1 w - - 0 2

How does white save this?

SF seems to be unable to see beyond a repetition of position in this line:

Code: Select all

info depth 45 seldepth 55 multipv 1 score cp 0 nodes 35608610123 nps 20293946 tbhits 0 time 1754642 pv c4c5 g8h8 e4f5 c8f5 d3f5 f8f5 b3b7 a8b8 b7c7 b8c8 c7b7 c8b8

Maybe play Qh5 instead of Bf5.

[zullil]

The 36th game of the TCEC final was for me the most interesting up until now, at least in terms of what could have happened.

SF found excellent attacking moves, with Ng3, Ng4, f5, Qe8-h5.

The only thing I wonder is why it did include e5 in the whole plan, wasting some time? Would not it be better to play 14...Qe8 straight after Komodo's 14.e4 (btw., the best move here)?

[d]r1b1qrk1/pp2p1b1/2pp3p/5pp1/2PPP1n1/PQNB1NP1/1P3PP1/R4RK1 w - - 0 2

How does white save this?

SF seems to be unable to see beyond a repetition of position in this line:

Code: Select all

info depth 45 seldepth 55 multipv 1 score cp 0 nodes 35608610123 nps 20293946 tbhits 0 time 1754642 pv c4c5 g8h8 e4f5 c8f5 d3f5 f8f5 b3b7 a8b8 b7c7 b8c8 c7b7 c8b8

Maybe play Qh5 instead of Bf5.

Better for White after Qc2:
[D]r1b2r1k/pp2p1b1/2pp3p/2P2Ppq/3P2n1/PQNB1NP1/1P3PP1/R4RK1 w - - 1 4

Code: Select all

+0.53 4. Qc2 Bxd4 5. cxd6 exd6 6. Rae1 Be5 7. Nd1 Bd7 8. Ne3 Nxe3 9. Rxe3 Bf6 10. Rfe1 Qf7 11. Bc4 d5 12. Bd3 Rae8 13. Ne5 Bxe5 14. Rxe5 Rxe5 15. Rxe5 b6 16. Qc3 Kg8 17. Qd4 Qg7 18. Qe3 Rf7 19. g4 Qf6 20. Bc2 Qd6 21. Bb3 Qc5 22. Qe1 Qd4 23. Qe2 Qf4 24. g3 Qd4 25. Kg2 h5 26. gxh5 Bxf5 27. Re8+ Kh7 (depth 41, 0:07:59)

[zullil]

Still better for White after Qc2:

Code: Select all

+0.39 4. Qc2 Bxd4 5. cxd6 exd6 6. Rae1 Be5 7. Nd1 Bf6 8. Ne3 Nxe3 9. Rxe3 Qf7 10. Rfe1 Bd7 11. Bc4 d5 12. Bd3 Qg7 13. Ne5 Rfe8 14. Nxd7 Rxe3 15. Rxe3 Qxd7 16. Re6 Kg7 17. Qe2 Qf7 18. Qe3 a6 19. b4 Rd8 20. g4 Rd7 21. g3 Re7 22. Kg2 Rxe6 23. fxe6 Qe7 24. Bb1 Kf8 25. Bc2 d4 26. Qe4 Qd6 27. Qg6 Qxe6 28. Qxh6+ Ke7 29. Bf5 Qf7 30. Kg1 Kd8 31. Kg2 (depth 48, 1:07:08)

[Lyudmil Tsvetkov]

Still better for White after Qc2:

Code: Select all

+0.39 4. Qc2 Bxd4 5. cxd6 exd6 6. Rae1 Be5 7. Nd1 Bf6 8. Ne3 Nxe3 9. Rxe3 Qf7 10. Rfe1 Bd7 11. Bc4 d5 12. Bd3 Qg7 13. Ne5 Rfe8 14. Nxd7 Rxe3 15. Rxe3 Qxd7 16. Re6 Kg7 17. Qe2 Qf7 18. Qe3 a6 19. b4 Rd8 20. g4 Rd7 21. g3 Re7 22. Kg2 Rxe6 23. fxe6 Qe7 24. Bb1 Kf8 25. Bc2 d4 26. Qe4 Qd6 27. Qg6 Qxe6 28. Qxh6+ Ke7 29. Bf5 Qf7 30. Kg1 Kd8 31. Kg2 (depth 48, 1:07:08)

Maybe there is a draw in the c5 line, after all.

I was considering mainly Rae1, and in most lines there black should have the advantage, especially if manages to play f5-f4.

The lines after f5-f4 are incredibly beautiful.

What a pity I could have been wrong this time.

Still, I think my assessment that 14...Qe8 is better than e5 should be correct, in terms of the overall performance of different lines.

As I acknowledged, this position is simply too complicated for me.

You never know what happens, even in terms of pawns tactics.

After c5, white can play cd6, or maybe black will play dc5, or maybe white will advance d5, or maybe e5, or maybe black will advance d5, you do not know if white will capture ef5, or black fe4, adn after f5-f4, is black will capture fg3, or maybe white gf4?

And this is only in terms of pawn tactics, what to say about existing piece tactics then?

And the worst thing is that most of the lines seem relevant.

That is why I said this position is too complicated for me. But I think such positions are most interesting.

[kbhearn]

Having had an engine running while this game was going on TCEC, SF was indeed stuck on 0.00 for so many different line choices. I think in a human game you're absolutely right to evaluate this as black advantage because it seemed much harder for white to hold the balance (often only 1 move and not an obvious one to find sufficient counterplay where black's moves seemed more natural and numerous)