Yes, strangely, Honfi played firstly 19..Be6 with a win and later he has played 19..Nf6 like me and lost in the same way I did.

ArKheiN wrote on 03/07/10 at 22:49:24:


In his Chess Publishing notes to Hoerling-Engstroem, Barcelona 2003.

Another 19...Be6 game is Ostojic-Honfi, Monte Carlo 1968.  It is cited in ECO as "with compensation"; the Russian "Modern Chess Opening Encyclopedia" and the first edition of David Levy's Dragon book gave it as slightly better for Black.

The Qa5 Dragon ship is letting in water in several places, but I am not sure that this is one of them.

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 g6 6.Be3 Bg7 7.f3 0-0 8.Qd2 Nc6 9.Bc4 Bd7 10.h4 Qa5 11.Bb3 Rfc8 12.0-0-0 Ne5 13.Kb1! Nc4 14.Bxc4 Rxc4 15.Nb3! Qc7 16.h5 Rxc3 17.Qxc3 Qxc3 18.bxc3 Nxh5 is a typical Dragon exchange sac position. The main thing is to play according to the principles for Black of such positions, in which he is the exchange down for a pawn (h-pawn), White has shattered queenside pawns and the queens have been exchanged.

First some observations:

a) In no way can this be described as an attacking position for Black. This is a not-quite-endgame position in which Black has compensation for the exchange, but little chance of imediate activity. First he has to focus on defence and prophylaxis. Any winning chances he achieves will tend to be "on the break".

b) g4 can be tactically dangerous for Black (so needs watching) but weakens the f3 pawn and the f4 square, thus offers Black chances of counterplay.

c) The shattered white queenside pawns are quite hard to attack (and so are only nominally "weak"); the problem for White is that they are hard to mobilize.

d) Black has a potential passed pawn on the h-file.

Rules of thumb for Black:

1) Try to avoid further piece exchanges, especially the exchange of the remaining black rook. The more pieces are still on the board, the less the advantage of the exchange makes itself felt. The black rook should therefore operate a policy of "avoiding confrontation" with its enemy counterparts.

2) Try to prevent White's semi-open files from becoming fully open. The white rooks are likely to cause havoc if they can invade the rear of the black position and White will also find it easier to force the exchange of Black's remaining rook if there is a fully open file.

3) Part of Black's compensation consists of the bishop pair; try to retain it, even if one of the bishops lacks immediate prospects of action.

Conclusions (provisional)
Even though Ward declared himself happy with his position in his game against M.Strange, it seems unwise to answer 19 Bd4 with 19...Nf6, since it allows 20 e5; thus 19...Nf6 conforms to principle 3 but contravenes principle 2.

19...Be6 is perhaps better (played by Dragon experts Gashimov and Cebalo) but contravenes principle 3, as does 19...Bb5.

That leaves 19...f6, which conforms to all the principles but feels counter-intuitive since it voluntarily shuts down Black's Dragon bishop. It has only been tried once, in a fairly low level game (Black won) but I suggest it deserves further investigation, since if it stands up tactically it is likely to be the best move strategically.

Some key games attached.

About Qa5 variation, the move is no longer played at higher levels since there appears to be problems in several lines at the moment.

Hello!

I want to give a "sad" story about myself I had 2 days ago. I will give the "complete" story that may not interest everyone, so for the reader interested by concrete dragon discussion, he just can read the chess part of my discussion if he wants.

There is a small local open (1h30 for 36 moves and 30 KO) with 7 rounds where after 5 games I had 4/5 points, and the favourite ELO (22~~) was leading with 4,5/5. We had to play on game 6, me as Black and of course, despite being a bit weaker than him, I wanted to play for a win (but of course I am aware that playing sharp as Black increase losing chances as well). Usually I always play for a win but with Black I am quite solid and dynamic in the beginning of the game. I had played this opponed about 1,5 years ago and I  have played the accelerated Dragon, I faced a Maroczy, I used my favourite Gurgenitze system and I managed to draw very easily after a quick Nd5 by White, exchanging everything and playing 2 bishops vs 2 bishops with no structural damage for both side, 99% of draw in many masters games as well.

I am an "expert" of the accelerated Dragon as Black, almost invincible around my level of play, I really like this defense since the day I leart that the Maroczy was not the ultimate torture I believed in the past because the Gurgenidze is quite easy to play (under the 26XX level of course). But when it comes about playing for a win as Black against the Maroczy, here this is a nightmare. The Schliemann, I only use it at corr chess and OTB against him (despite the fact that he plays the Maroczy, he is a dynamical and tactical player like me) would not be a very good idea, and he knows that opening. So I begin to think about every openings. This is not the first time I want to construct a true winning repertoire as Black when I need. I used to play the Najdorf (and a bit of Kan) a lot OTB before the accelerated Dragon but I didn't want because he plays the White side the way I do against theses defenses, while it's an advantage because of the understanding of playing a position for both side, I would not have been so happy to play "against myself". So, many unorthodox defenses cames into my mind, Alekhine, 1..Nc6, Caro-Kann with ..g6 just to name a few. But I feel something is wrong when you begin to think like I did at this moment. Then I had a serious idea, I remembered I still own for a few years the books "winning with the dragon 2 by Ward and easy guide to the dragon by Golubev. I had them mostly to prepare as White somes years ago, because I had not the balls to play it as Black at that time because for me Dragon = ton of theory just to not get mated by sac sac mat. But since that time I had experience with accelerated Dragon, so I know how to moves pieces a bit in theses structures, and I know not much so double-edged games against 1.e4. I think I already made my choice and I had 2 days for prepare for the game. I read my sources very intensively and I really enjoyed the idea to play my first Dragon! I was quite good prepared with the time I had. What I think I have read somewhere is that the ..Qa5 Dragon is refuted now, but hey, my opponent can't imagine I will play a Dragon, he is not a GM, and the refutation may be hard to remember or hard to find over the board! In fact I think my idea of picking the Dragon as a winning attempt was very good (but in fact, playing the accelerated Dragon, my pet opening was maybe best because what I forgot, is that he may want to play for a win to secure first place of the tournament, and overplaying in the Maroczy is dangerous while the Nc3 is more double-edged and I know quite well my repertoire, but maybe he had a serious preparation in mind because he really thought I would play the accelerated Dragon).

Ok sorry for the big talk

The chess part

The game begin.

2200 vs me

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 g6 (my first official "true Dragon") 6.Be3 Bg7 7.f3 0-0 8.Qd2 Nc6 9.Bc4 Bd7 10.h4 Qa5 11.Bb3 Rfc8 12.0-0-0 Ne5 13.Kb1! where I was aware of the 13..b5 move but I did not work it and Im not sure it's my taste. One day I had the White side of this position and my IM opponent shocked me with 13..Rxc3?!? a classical sacrifice but without having one pawn in hand so this might be a bit doubtful objectively but that's not so easy to play. I have been overplayed in the long run and lost, and my opponent said he even beat somes masters with that move. But I did not want to try this move from Black side this time) So I continued with Ward's recommanded 13..Nc4 14.Bxc4 Rxc4 15.Nb3! Qc7 and now I waited more 16.g4 or 16.Bd4 where I remembered my work with the book. I remember that Golubev in the easy guide speak of theses 2 moves only as being best. But my opponent played the natural 16.h5! where at first sight I remembered the rule "when you can with a pawn you can sacrifice the exchange on c3" which is almost always true, a rule I learnt from Ward. For that reason I really dismissed 16.h5 with no analysis. But the more I looked mentaly at 16..Rxc3 the least I liked the Black side of the ending, so I begin to think about 16..Nxh5 but I was worryied by 17.Nd5, I thought about 16..a5 but I felt this was not as good as with the bishop on d4, I thought of 16..Be6 which looked reasonable but his attack seemed so quick and against 16..Rc8 there was probably Bd4...

The nightmare began here for me as I didn't know what to chose and I couldn't remember what Ward said about that in his book(I just remembered Golubev doesn't mention it). I spend at least 20 minutes here or even 30 and I played 16..Rxc3 17.Qxc3 Qxc3 18.bxc3 Nxh5, my first idea since the first second and now the played natural 19.Bd4 I saw coming. I still didn't feel comfortable (I don't sacrifice exchange every day, I am a new Dragon player, but this is not the reason of my bad feeling because in many ending examples I undestood Black's compensation when I worked home or when I had the White side). I think this position is very important for the soundness of Black's system. Black is an exchange down for a pawn, and I don't see clears ways to improve my position. I felt my only plan in that position was to survive in a position at least +/= for humans, against a player a little stronger, and with about 45 minutes for the first time control against 1h15. Add to that the fact that I was disappointed by how it was so easy for him to stop my works and you get a suboptimal play. I have played the quite natural 18..Nf6 and he opened the file for his rooks with 20.e5! dxe5 (Ne8 was tempting but not obviously better) 21.Bxe5 and my Bd7 has to move, and I didn't like many squares because of the Nb3 jumps, so I played 21..Bb5 where he played 22.Rh4! supporting the push c3-c4 or with the idea of doubling in the d-file and threatening the 8th rank. Time pressure increases after each move, position become worse after each move, I moved pieces around and I lost an ending the exchange down, what followed is not instructive, I just have been outplayed. I even could't show my chess and I had no fun so quickly...

We had not much time to analyze but he said to me he got this position (after 16.h5) 3 times in serious play (one GM and one MI at least) so the position was not unknown to him, this is another factor I didn't know before, I know ..Qa5 is not a bug surprise today but I couldn't imagine he knows that well how to play the White side. (In fact he lost against the masters so this is a good news for Black but I think he got promising positions). I was really disappointed by the fact I have been outplayed by very easy White moves while many GM as White lost badly or haven't played 16.h5. I was disappointed because I couldn't remember during the game, what Ward say about that move. So I ran at home, opened WWTD2 and I understood why I could'nt remember of that move, it's not mentionned at all if I am not wrong. So I opened CB and my 16..Rxc3 proved to be clearly the main move with a score of 45% for Black in my base. Another surprise: Ward himself played the Black side of 16.h5 against a 2250 in the same way I did (but he played 21..Bc6 instead of my 21..Bb5 and he did draw quite easily but I feel White could do better that the game. Still, a draw as a GM against a 2250 is not an happy result, but from the position it's probably a fair result for the Master! Last surprise, I found the game from years ago of my opponent against the GM (0-1), where the GM used 19..Be6(!) I understand that with this move, Black can't allow White to open the d file in the same way my opponent did. So 19..Be6 is probably better that 19..Nf6, and I watched my recent ECO and it gives 19..Be6 indeed, with the sign "enough compensation".

But I wanted to win, and here I still feel I am just struggling to survive. I believe the position to be easier to play White let's say +/= for human, with very few winning prospects to Black if White is not crazy. This is not a WWTD spirit.

But I am not accusing Ward for my loss! I still enjoyed a lot my work and I may use the Dragon again but before I want to be sure I can't get neutralized so easily like this... I don't care of the risk of losing if the game was tense and fun like many Dragon lines. I don't care if a line is refuted after 10 precises moves by White if it's not findable OTB because I would use it only to unprepared opponents.

So there is my question, to Ward, and other Dragon experts: Can I continue to use his repertoire without being beated again and again without a chance? Is there an (easy) refutation of the Qa5 system? What do you think of 16.h5? You must know I have no subscription to the chesspub theory and did not read the whole dragon forum! And I don't want to learn a totally new way of playing the Dragon just to use it a few time.

Thank you for taking the time of the lecture, and sorry for my far from perfect english!

Thank you for the futures responses!