I've been meaning to develop my understanding of the Hyper-Accelerated Dragon (1.e4 c5 2.Nf3 g6) as Black, mainly by playing a few correspondence games and working through deep positional analyses. Most of what I read about this opening centers on the Maróczy Bind as White's typical response, as it's the most efficient restriction on Black's dynamic play. I haven't yet read any solid ideas for Black to play for an advantage against the Maróczy. Most sources that I've been able to find could basically be summed up as "Stay sharp and wait for White to lose concentration." However, this isn't realistic when it comes to correspondence games (particularly against centaurs). It seems that facing an ideal offense from White, Black's position gets irredeemably worse:

[FEN ""]
1.e4 c5 2.Nf3 g6 3.d4 cxd4 4.Nxd4 Nc6 5.c4 Nf6 6.Nc3 d6 7.Be2 Nxd4 8.Qxd4 Bg7 9.Be3 O-O 10.Qd2 Be6 11.O-O Qa5

In fact, there are several OTB games between high-rated players that feature this line, such as two games between Ivanchuk and Anand in 1994. Actually, after 12.b3, it seems that White has never lost an ICCF game from this position, nor does the database at 365Chess have any winning games for Black. Is the Hyper-Accelerated Dragon just not a viable opening for Black in a correspondence game? Can anybody point out where Black might improve their play along this line? Thanks in advance for your time.

  • 1
    Winning black side of correspondence game is too much for this line.
    – hoacin
    Feb 14, 2017 at 16:43

3 Answers 3


In Beating the Sicilian 3 by Nunn and Gallagher, they claim (emphasis is mine):

Although the Maroczy Bind is slightly passive for Black, players such as Larsen, Petursson and Velimirović have shown that by patiently waiting for a lapse of concentration from White this line can offer winning chances for Black. The theoretical opinion is that White should maintain a slight advantage, but no one should believe that this is a line in which White cannot lose.

As you know, in a correspondence game lapses of concentration are rarer than in over-the-board.

My generic advice would be to be alert to weaknesses induced by white's last move, and take advantage of that. If you want more specific advice, please post a new question requesting feedback on your middle/endgame play.


The correspondence chess database includes many wins for Black. Black typically wins by queenside attack or because his pawn structure is superior to Whites for the endgame because it is a bit over extended. The best way to prepare is to look at games from the best players you can find. There are plenty of opportunities for Black to win if he understands the position.


With B27 Sicilian (Hungarian): 1.e4 c5 2.Nf3 g6 365Chess.com lists black chances at 33% compared to Whites 35.9% (draw: 31.2%). That seems to be pretty even.

Carry on to the Sicilian Defence, Accelerated Dragon: 1.e4 c5 2.Nf3 g6 3.d4 cxd4 4.Nxd4 Nc6 and 365Chess.com has whites chances pretty much the same but black's chances dropping to 29.5% (draw: 35.4%).

White's 5. c4 (making it the Maróczy) seems to suggest play ahead is sharp. Blacks chances are restored at 35.5% but whites also improve at 39.6%.

Draw's chances fall to 25% so these percentage changes don't suggest black's chances are disadvantaged. Rather it suggests this position presents many traps for both white and black. So the recommendation to 'stay sharp' applies equally to both black and white, not just black.

I was able to find 11621 games with this position in my SCID Database (my SCID DB is 3.3M games). Here's a couple of games to look at:

Fischer used this opening to beat Bent Larsen:

[Event "Fischer - Larsen Candidates Semifinal"]
[Site "Denver, CO USA"]
[Date "1971.07.08"]
[EventDate "1971.??.??"]
[Round "2"]
[Result "0-1"]
[White "Bent Larsen"]
[Black "Robert James Fischer"]
[ECO "A30"]
[WhiteElo "?"]
[BlackElo "?"]
[PlyCount "108"]
[FEN ""]

1. c4 c5 2. Nf3 g6 3. d4 cxd4 4. Nxd4 Nc6 5. e4 Nf6 6. Nc3 d6
7. Be2 Nxd4 8. Qxd4 Bg7 9. Bg5 h6 10. Be3 O-O 11. Qd2 Kh7
12. O-O Be6 13. f4 Rc8 14. b3 Qa5 15. a3 a6 16. f5 Bd7 17. b4
Qe5 18. Rae1 Bc6 19. Bf4 Nxe4 20. Nxe4 Qxe4 21. Bd3 Qd4+
22. Kh1 Rce8 23. Be3 Qc3 24. Bxh6 Qxd2 25. Bxd2 Be5 26. Bf4
Bxf4 27. Rxf4 gxf5 28. Rxf5 Kg7 29. Rg5+ Kh6 30. h4 e6 31. Rf1
f5 32. Re1 Rf7 33. b5 axb5 34. cxb5 Bd7 35. g4 Ra8 36. gxf5
exf5 37. Bc4 Ra4 38. Rc1 Bxb5 39. Bxf7 Rxh4+ 40. Kg2 Kxg5
41. Bd5 Ba6 42. Rd1 Ra4 43. Bf3 Rxa3 44. Rxd6 Ra2+ 45. Kg1 Kf4
46. Bg2 Rb2 47. Rd7 b6 48. Rd8 Be2 49. Bh3 Bg4 50. Bf1 Bf3
51. Rb8 Be4 52. Ba6 Ke3 53. Rc8 Rb1+ 54. Kh2 Kf4 0-1

From the emergence of the Maróczy bind position white is encouraged NOT to castle queen-side as there are many traps black can exploit. From the Maróczy the main line continues with black 5. .. Bg7 or 5. .. Nh6

The Stockfish chess engine has 5. .. Nh6 as the stronger response with a 20 move analysis, but let it run for a while with deeper analysis (36 moves) and it really favours 5. .. Bg7:

36 +0.40 5.... Bg7 6.Be3 (1048.58)

It really doesn't matter because whichever move you play Nh6 or Bg7 either has the other move following quickly in the sequence anyway. In the above game Fischer played 5. .. Nh6 . Stockfish eventually (after a really long analysis) has the very counter-intuitive 5. .. d6 emerging as strongest:

37 +0.40 5.... d6 6.Be2 Nf6 7.Nc3 Nxd4 8.Qxd4 Bg7 9.Be3 O-O 10.Qd2 Qa5 11.O-O Be6 12.b3 Rfc8 13.Rfc1 Ng4 14.Bxg4 Bxg4 15.Bd4 Bxd4 16.Qxd4 Be6 17.a4 Qc5 18.Qxc5 dxc5 19.Nd5 Kf8 20.f3 Rd8 21.Kf2 Rac8 22.Ke3 f5 23.exf5 gxf5 (1513.20)

It's such a long advantage it's likely best stick with 5. .. Nh6 or Bg7 since even the best humans can't see that far ahead.

Some statistics for the Maróczy with a black 5. .. Bg7 response are as follows (SCID):

Most frequent players (White)

1: 14 1988-2000 64% 2605 Nijboer, Friso [2,3,7,88,105]
2: 12 1991-2013 67% 2609 Korneev, Oleg [24,35,63]
3: 10 1993-1998 70% 2570 Wojtkiewicz, Aleksander [55,101]
4: 9 1993-2014 83% 2599 Kotronias, Vasilios [37,61,102,105]
5: 8 1995-2012 69% 2732 Lékó, Péter [45,55]
6: 8 1980-1998 69% 2475 Székely, Péter

Most frequent players (Black)

1: 21 1991-1998 43% 2515 Nielsen, Peter Heine [55,79,88,94,100,105]
2: 20 1973-2006 55% 2575 Velimirovic, Dragoljub [18,23,34,41,45,49,50,54,...]
3: 17 1987-2014 29% 2395 Afek, Yochanan
4: 16 2000-2016 56% 2646 Guseinov, Gadir [2,62,63,78,89,96,102]
5: 14 2009-2013 39% 2492 Plat, Vojtech [9,29,30]
6: 13 2013-2016 54% 2661 Iturrizaga Bonelli, Eduardo [3,8,48,88,89,96,103,105]

Oldest game
=-=(20) Keres - Benkő, Moskou olm 1956.09 [38] (Draw)

Newest game
=-=(17) Kovalev 2596 - Guseinov 2623, Konya TUR 2016.07.21 [102] (Draw)

This is likely the original Maróczy match with Maróczy winning (from Chessgames):

[Event "Monte Carlo"]
[Site "Monte Carlo MNC"]
[Date "1904.02.11"]
[EventDate "1904.02.08"]
[Round "4"]
[Result "0-1"]
[White "Rudolf Swiderski"]
[Black "Geza Maroczy"]
[ECO "B38"]
[WhiteElo "?"]
[BlackElo "?"]
[FEN ""]
[PlyCount "96"]

1.e4 c5 2.c4 Nc6 3.Nf3 g6 4.d4 cxd4 5.Nxd4 Bg7 6.Be3 Nf6 7.Nc3
d6 8.Be2 Bd7 9.O-O O-O 10.h3 Nxd4 11.Bxd4 Bc6 12.Qd3 Nd7
13.Bxg7 Kxg7 14.b4 b6 15.Rfd1 a5 16.a3 axb4 17.axb4 Qc7 18.Nd5
Bxd5 19.Qxd5 Rxa1 20.Rxa1 f5 21.Qe6 Ne5 22.exf5 Rxf5 23.Ra8
Rf8 24.Rxf8 Kxf8 25.c5 dxc5 26.f4 Nf7 27.Bc4 Qxf4 28.bxc5 bxc5
29.Qc8+ Kg7 30.Qxc5 Qe5 31.Qc8 Nd6 32.Qg8+ Kh6 33.Qf8+ Kg5
34.Bf1 Qe3+ 35.Qf2 Qxf2+ 36.Kxf2 Kf4 37.Bd3 Ne4+ 38.Ke2 g5
39.Bc2 h5 40.Bb3 e5 41.Bf7 h4 42.Bc4 Nf6 43.Kf2 Ke4 44.Bf7 Kd3
45.Bg6+ e4 46.Bf5 Kd2 47.Bxe4 Nxe4+ 48.Kf3 Kd3 0-1

And here is Pál Benkő using it as black to keep Bobby Fischer at bay in New York in a draw. Note: my SCID DB has this game in 1960 but Chessgames has it taking place in 1961:

[Event "New York ch-US  Rd: 10"]
[Site "New York, NY USA"]
[Date "1961.01.01"]
[EventDate "?"]
[Round "10"]
[Result "1/2-1/2"]
[White "Robert James Fischer"]
[Black "Pal Benko"]
[ECO "B36"]
[WhiteElo "?"]
[BlackElo "?"]
[PlyCount "24"]
[FEN ""]

1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 g6 5. c4 Nf6 6. Nc3
Nxd4 7. Qxd4 d6 8. Be2 Bg7 9. Be3 O-O 10. Qd2 Be6 11. O-O Qa5
12. f4 Rfc8 1/2-1/2


I'm not sure what you mean by "nor does the database at 365Chess have any winning games for Black". I found many (and this one for example). It's not true this opening is necessarily weak for Black. Mikhail Tal (Oliver Reeh vs Mikhail Tal, Deutschland Cup (1989)) and Viswanathan Anand (Vassily Ivanchuk vs Viswanathan Anand, Tournament (thematic) 1994) have both had success with it.

Perhaps they know its opening theory better than others but it seems to be whoever knows their theory best has the greatest chance to win unsurprisingly.

  • I believe the Maroczy is 5.c4 not 5.c3.
    – bof
    Feb 17, 2017 at 1:30
  • Why do you believe that? I posted the Maróczy match. What did he play?
    – user34445
    Feb 17, 2017 at 3:35
  • The database at 365Chess has no winning games for the position after 12.b3. I played through one of the mainlines available to White as it appears in the ICCF database and found that if Black plays ideally against it (as best as I can tell), White can force Black into a position for which I cannot find any winning games for Black (specifically, the position I embedded in my question, plus 12.b3 from White).
    – Michael L.
    Feb 17, 2017 at 6:42
  • 1
    Thanks. I think it is these same tactical landmines that make this opening either one to be avoided or one to be pursued; depending, of course, upon the relative strength of you and your opponent's tactics.
    – user34445
    Feb 17, 2017 at 17:32
  • 2
    That's roughly the same as the question I asked, which is about a possible refutation of the Maróczy Bind suitable for computer-aided correspondence chess. In ICCF games, players often let their computers calculate out 35-40 moves, and they also have access to opening books and tablebases.
    – Michael L.
    Feb 17, 2017 at 18:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.