What a move 11.a3!! Aronian-Carlsen_Norway Chess 2017

[Lyudmil Tsvetkov]

ok, I am fed up with it.

so far, 264 analyses, 264 draws.

did nit find a single line where white can win.

e4 Nd7 Bf4 (Ba5) f5

Ba5 Na6, followed by f5

g4 Nd7 (but f5 also holds)

h4 f5

f4 f5

Kf1(f3) f5

always fortresses.

not a single line remained I have not checked.

On spotting this position for the first time, I evaluated it immediately with 0.0, did not doubt for a second, it is so simplistic, but your eagerness to prove Svidler and Aronian analyse better than me completely shook me, so I started analysing things that should not be analysed.

obvious draws are obvious draws.

I bet, if you show 10 top players this line, all of them will immediately say: draw.

[Lyudmil Tsvetkov]

Thanks, Michael! Incredible, that black can lose this position so quickly. Although the rook exchange surely must help black, it still seems very difficult to hold the ending.

In the meantime I played this one on my Quad this morning. Another quick white win by SF over Houdini:

[pgn][Event "Rapid 60m"]
[Site "Chess"]
[Date "2017.06.17"]
[Round "1"]
[White "Stockfish 260517 64 POPCNT"]
[Black "Houdini 5.01 x64-popc"]
[Result "1-0"]
[ECO "D45"]
[Annotator "1.55;1.53"]
[PlyCount "89"]
[TimeControl "3600"]

1. d4 d5 2. c4 c6 3. Nf3 Nf6 4. Nc3 e6 5. e3 a6 6. b3
Bb4 7. Bd2 Nbd7 8. Bd3 O-O 9. O-O Qe7 10. Bc2 Rd8 11. a3 Bxa3 12. Rxa3 Qxa3 13.
c5 Ne4 14. Bxe4 dxe4 15. Ng5 h6 16. Ngxe4 a5 17. Qc2 a4 18. bxa4 Qb4 19. Nd6
Qa5 20. Nd5 Qxa4 21. Ne7+ Kf8 22. Qxa4 Rxa4 23. Nexc8 Ra8 24. Rb1 Rdxc8 25.
Nxc8 Rxc8 26. Rxb7 Rb8 27. Rxb8+ Nxb8 {Both last book move} 28. Kf1 {1.55/42
198} f6 {1.53/32 284 (Nd7)} 29. Ke2 {1.62/41 67} Nd7 {1.50/32 93} 30. Ba5 {
1.62/42 19} Kf7 {1.56/35 118 (h5)} 31. Kd3 {1.70/40 71} g5 {1.60/33 108} 32.
Bc7 {1.78/39 20 (Kc4)} h5 {1.68/36 122} 33. Kc4 {1.96/43 95} Ke8 {1.79/37 125
(Nf8)} 34. Bd6 {2.52/41 62 (e4)} f5 {1.85/36 100} 35. f3 {2.61/41 20 (Kd3)} Nf6
{2.70/36 339} 36. Be5 {2.95/40 67} Nd5 {2.77/38 127} 37. Kd3 {3.22/41 70} g4 {
2.84/31 24 (Nb4+)} 38. fxg4 {4.93/48 231 (Bd6)} hxg4 {3.62/36 160} 39. e4 {
5.28/41 48} Ne7 {4.50/34 147} 40. Ke3 {5.47/36 15} Ng6 {5.71/36 238} 41. h3 {
5.59/54 46} gxh3 {6.17/39 143 (fxe4)} 42. gxh3 {6.77/42 51} Kf8 {6.97/35 110
(Kf7)} 43. exf5 {9.36/42 140} exf5 {7.54/32 95} 44. Bd6+ {10.12/37 65} Kf7 {
8.19/36 80} 45. Kf3 {10.22/36 14} 1-0[/pgn]

well, obviously, already f6 is ridiculous, g5 even more so.

with such 3rd grade play, obviously black will lose.

[Lyudmil Tsvetkov]

I received this game from a friend. Komodo loses in only 9 moves after 27.Rxb8 Nxb8. Amazing. For me, this is enough now on this subject.

My conclusion: Aronian's 11.a3! is the best move in the position and offers white excellent chances. Best answer is 11... Bd6, to keep a playable position. Too risky is 11... Bxa3? which leads to a clear advantage for white in all lines and black must avoid many pitfalls to lose quickly. I am looking forward to Aronian's annotation of this game!

[pgn][Event "Eindspel"]
[Site "Test-PC"]
[Date "2017.06.16"]
[Round "1"]
[White "Stockfish 8"]
[Black "Komodo 10.4"]
[Result "1-0"]
[ECO "D45"]
[Annotator "1.61;1.38"]
[PlyCount "71"]
[TimeControl "5400"]

1. d4 d5 2. c4 c6 3. Nf3 Nf6 4. Nc3 e6 5. e3 a6 6. b3
Bb4 7. Bd2 Nbd7 8. Bd3 O-O 9. O-O Qe7 10. Bc2 Rd8 11. a3 Bxa3 12. Rxa3 Qxa3 13.
c5 Ne4 14. Bxe4 dxe4 15. Ng5 h6 16. Ngxe4 a5 17. Qc2 a4 18. bxa4 Qb4 19. Nd6
Qa5 20. Nd5 Qxa4 21. Ne7+ Kf8 22. Qxa4 Rxa4 23. Nexc8 Ra8 24. Rb1 Rdxc8 25.
Nxc8 Rxc8 26. Rxb7 Rb8 27. Rxb8+ Nxb8 {Both last book move} 28. Kf1 {1.61/42
290} f6 {1.38/34 126 (Nd7)} 29. Ke2 {1.61/44 162} e5 {1.60/34 93 (h5)} 30. Kd3
{2.14/43 81 (Bc3)} Ke7 {1.97/37 254} 31. g4 {2.39/46 166 (Bc3)} g5 {2.27/39
304 (exd4)} 32. Kc4 {4.37/40 87} Kd7 {3.71/37 83 (Ke6)} 33. Ba5 {5.04/42 120}
exd4 {4.18/43 246 (Ke7)} 34. exd4 {6.37/41 75} Kc8 {5.23/44 223} 35. d5 {
10.81/45 204 (Bc3)} Nd7 {6.76/44 131} 36. Bc3 {11.33/37 40 adjud.} 1-0[/pgn]

again, f6 is ridiculous, e5 even more so.

on Kf1, f5 holds very easily.

till when are you going to repeat that mantra?

a3 is weak, I already demonstrated a straightforward, and completely forced drawing line, did not you notice it?

Bd6 loses the game, but no time to discuss now, as, if can not agree on an obvious simplistic endgame, how are we going to decide on such complicated positions?

c5 is much stronger, b6 loses immediately, the e5 line is much better, white gest an advantage also after it, but for the same reason I will not post my output: we will never agree on anything.

[Lyudmil Tsvetkov]

a3 is weak, as both Ba3 and Bd6 in reply seemingly lead to a draw.
Carlsen made some mistakes tactically, to lose the game afterwards.

best moves is obviously c5, suggested by both SF and Komodo, and after e5 a3 Bc3 Bc3 e4 Nd2 Nf8, white has seemingly only not-losing, but very strong move, Re1(defending the e3 pawn and preparing either f2-f3 or f2-f4, otherwise white can not get rid of the nasty clamp in the form of the e4,e3,f2,g2 pawns, and will necessarily lose in due time) Bg4 f3 ef3 gf3 Bh5 Qe2:

[d]r2r1nk1/1p2qppp/p1p2n2/2Pp3b/3P4/PPB1PP2/2BNQ2P/R3R1K1 b - - 0 8

and, although SF shows some black edge here, white has excellent winning chances, but game should probably still end with a draw with perfect play.

SF will see white is much better on another 10-20 moves, of course...

instead, in both SF and Komodo's choice of this line, avoiding to play f3 or f4, white should lose.

[d]r1br1nk1/1p2qppp/p1p2n2/2Pp4/3Pp3/PPB1P3/2BN1PPP/R2Q1RK1 w - - 0 5

only choice for white is to play here Re1, otherwise white loses by force.

Have you at least changed your verdict on the merits of a3 over c5? I would like to know how an engine shoot out at respectable time control from the c5 move. If the imperical evidence shows a3 leads to better score than c5 then it would be good to admit it.

no, c5 is much better than a3, a3 was an obvious blunder.

I can more or less conclusively demonstrate c5 leads to bigger advantage for white than a3, but will not do that, as, if we can not agree on very simple endgames, how are we going to do so in an extremely complicated position with myriads of variations?

an engine shooutout, no matter the TC, after c5 would be meaningless, for first engines do not understand the position, and secondly, it is so rich in branching lines, that practically anything can happen.

[yanquis1972]

you continue to claim fortress, but i really think you are missing things that computers are seeing, hence moves you feel ridiculous. in your fortress position, how does black hold against the threat of g4?

if there are several variables in play (& not one simple idea i can replicate) when you have time could you demonstrate against SF or komodo?

[Lyudmil Tsvetkov]

apologies for switching positions again, but i am beginning to think the core issue is that lyudmil is incorrect or overly optimistic about what constitutes a fortress. i don't mean that as an insult, given that i don't really know what a fortress is.

[d]4k3/6B1/2p1p3/2P2p1p/3PnP2/4PK2/6PP/8 w - -

you say this is a simple fortress draw, but i don't believe engine - engine play (at least with only 5-man TBs) would confirm it as such (maybe at very long time control; ie not simple...although this is only my suspicion; i'm just now running a 15min game)...there is no question the computer fumbles around (making it all the more difficult for me to comprehend whats going on) but how does black draw? how can it be a fortress when the king & minor piece remain at very least potentially active? most baffling to me is the pawn break with h3 & g4; i gather it doesn't work immediately, but it seems to be a serious threat...

i will post a couple games shortly; the first i forgot to turn resign off, but the eval was near +10. the second just started & is quite different as the knight re-established itself on d5 instead of e4...which i assume is a blunder since white now threatens e4 as well. so perhaps 3 games...

black draws by doing nothing, that is how fortresses are handled in general, doing nothing.

e3 pawn is blocked by the knight, so immobile
c5 and f4 pawns are blocked and immobile
d4 pawn is immobile, because the d5 square is controlled by 2 enemy pawns, c6 and e6, while the only other friendly pawn on adjacent file, able to support it, e3, is blocked by the knight

thus, only try to do something is push h3, followed by g4.
but, on h3, black has Nd2 check Ke2, and back to e4, Ne4, with repetition, so pushing g4 is tactically impossible.

in case white tries to prevent knight check from c3 with a manoeuvre like Bg5-h4-e1, controlling c3, when the bishop gets to e1, the black knight returns to f6, and g4 is again impossible, as the knight will control it too.

saccing the g pawn after h3 Nd2 Ke2 Ne4 g4 hg4 hg4 fg4, gives black a dangerous passer with simple draw

alternatively, attempting to bring the white king to the queen side easily fails, simply to placing the black king on d7, and the white king can not pass. in the meantime, black has always passing moves with the knight, so no zugzwang.

an obvious draw for me, spot it in less than a second.

however, engines might very well blunder, even if all games. they simply do not understand fortresses, so they will prefer lines with artificially lower score, that is losing instead.

[Lyudmil Tsvetkov]

you continue to claim fortress, but i really think you are missing things that computers are seeing, hence moves you feel ridiculous. in your fortress position, how does black hold against the threat of g4?

if there are several variables in play (& not one simple idea i can replicate) when you have time could you demonstrate against SF or komodo?

I already posted plan, please see other message.

white can do nothing to improve.

[Lyudmil Tsvetkov]

will be briefly back, for people not to say I have abandoned the forum, as well as my beliefs.

How could people ever say this? In many of the threads you participate in you sooner or later announce that you won't post again, but then post only minutes later. You are well known for these antics.

Also why would it bother you if people say that? Step back and really think about that, you might learn something about yourself. There is no shame in admitting a mistake, realizing ones limitations and moving on.

I am perfectly fine, john, please believe me, it is not possible to throw me off balance with such posts.

And yet you return.

I simply missed at first that black can trade rooks because of the mate threat on the first rank.

Yes first true statement. Players of your strength tend to miss these mate-in-one threats.

This time, I spent couple of hours analysing every single possible meaningful continuation, and, as expected there is draw everywhere. absoluetly everywhere. I checked and rechecked this, in at least 10 different continuations.

At least 10 different continuations... Wow! That must be a lot for you. Congrats for trying soooo hard.

however, black should proceed with a very stringent plan: when playing f5, the f5 pawn should be supported by a pawn on g6, to capture gf5 after ef5, otherwise black loses.

No he should not, this is one of the weaker defensive set-ups and looses. As you probably already fear, because you do not even provide any lines. I understand that for a player of your caliber move-orders do not matter but unfortunately they do.

this is the simple relevant continuation repeating in almost all lines, but it perfectly holds, no doubt Carlsen would have found it imediately, as I myself pick it on the spot as only alternative.

You and Carlsen, great minds think alike. Sure you would have picked it immediatly, but one moment didn't you say this

I simply missed at first that black can trade rooks because of the mate threat on the first rank.

Hmmm, strange concept of immediatly you have.

every other possible configuration, as said, might be losing at some point, but engines do not quite get that, choosing other moves.

And yet, many people on here have been telling you that in dozens of posts.

again, this is position engines do not understand, almost all of their moves are wrong, as only possible way to draw is going for a fortress.

Lyudmil, the only fortress here is your mind.

as said, in the simple one-piece endgame, black should first move g6, before pushing f5.

Simple for simple minded people.

My challenge still stands - why not play the position against me move-by-move and stand by your beliefs?

I guess it is about time to ban this one from the forum...

[d]1n3k2/5pp1/2p1p2p/2P5/3P4/4P3/3B1PPP/6K1 w - - 0 2

DRAW

show me a possible winning continuation, and I will refute it.

[Guenther]

ok, I am fed up with it.

so far, 264 analyses, 264 draws.

If this would be true, we also had to doubt about your ability to run a chess program at all... that is a completely ridiculous result from the given position.

[Lyudmil Tsvetkov]

also, why do not you send a message to Carlsen, Svidler and Aronian, they have web pages, will certainly reply in a while?

and you know what, all of them will say: draw!

you already start being obnoxious, by not showing any lines, but just picking single moves here and there, sometimes not even single moves, just single stings.

show us your analysis.