At the suggestion of Bruce on the "Death of KG" thread, I have taken the opportunity to start a thread on the Belgrade. I'm currently playing some games in the Belgrade on IECC, so I cannot post them yet but I shall do when I've finished them.

Anyway, for those who have never heard of this crazy opening, it is reached after 1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 (the "tame" four knights) 4.d4 exd4 5.Nd5!? (If someone can insert a diagram here that would be great)

In this position, black has a number of tries. He usually declines the gambit pawn, and I think this is probably wise. Nonetheless, the main tries are:

A) 5...Bc5
This line is often not mentioned, and sometimes only recieves a footnote, but I think it deserves close attention. The point is 6.Bf4 is met by O-O!, and 7.Bxc7 Qe8! is probably better for black. So instead best is probably 6.Bg5, when black usually chooses between 6...Be7 (which I don't feel is correct), or 6...d6. Quite a few players, including Mr Monson, have been known to play 6.Bf4 and only on 6...d6 play 7.Bg5, as this prevents ...Be7, but if black is aware of the O-O possibility, I'm not sure whether this is good for white.

B) 5...Nb4(!)
This move is often given a ! in theoretical manuals. White usually replies 6.Bc4, and then 6...Nbxd5 7.exd5 Bb4+ 8.Bd2 Qe7+ 9.Qe2 Bxd2+! 10.Kxd2 Qxe2+ 11.Kxe2 c5! 12.dxc6 bxc6 13.Nxd4 d5 gives black the advantage. I'd be interested to see what improvements Bruce has for white here (I do not own a copy of his book unfortunately). 6.Nxd4 Nxe4 7.Nb5 Nxd5 8.Qxd5 Qe7 9.Nxc7+?! is an interesting piece sac from junior which I don't believe is sound, but seems quite dangerous after 9...Kd8 10.Bf4 d6 11.O-O-O Kxc7 12.Qc4+ Kb8 13.Qd4, even if the attack is insufficient. 6.Nxf6+ Qxf6 7.Bb5 Bc5 8.O-O O-O 9.e5 Qb6 10.Be2 d6 is also better for black, so 5...Nb4 may be the way for black to cast doubt on the gambit's validity.

C) 5...Be7
This move is often considered the safest way for black to get a good game. 6.Nxd4 Nxd5 7.exd5 Nxd4 8.Qxd4 O-O intending ...Bf6 is better for black, so white needs to find something else. 6.Bc4 O-O 7.O-O d6 8.Nxd4 Nxd4 9.Qxd4 Nxd5 10.Bxd5 Bf6 11.Qd3 is a very slight improvement on this for white, but black's position still looks very comfortable. 6.Bf4 O-O! 7.Bxc7 Qe8 is similar to the idea I mentioned in line A above, though Igor Polovodin has shown a willingness to defend this position as white on more than one occassion. 8.Be2 d6 9.Nxf6+ Bxf6 10.Bxd6 Be7 11.Bxe7 Qxe7 12.Nd2 f5! gives black mighty good play for his material investment, 8....d6! being the novelty found and used by Felix Izeta Txabarri to defeat Bellon Lopez in 1998.

D) 5...Nxe4
This is rarely played, and sensibly so - it is extremely difficult for black to hold the position. I don't know the critical lines that well, so I wont attempt any analysis of these lines yet.

So what do you guys think? Does anyone play the white side and know of any improvements? Does anyone like the black side and have any more lines? Or do you just feel like saying your two pennies worth? Here's your chance!

Regards,
Craig

I do not agree with TopNotch. If you want to learn about open piece play and the balance between development/activity and material, there is no better way than practising open gambits, even if they are a bit obscure.
Of course it is very possible, that you feel the need to practice something strategically more complex after a while.

TopNotch (who ever this is) has missed the point on a whole host of issues:

First, whatever "his thoughts" on the Belgrade Gambit may be, they should hardly be given any significance in this case since he failed to provide even one variation to the discussion in support of his blanket castigation.

Second, this discussion is on the Belgrade Gambit and some questions some players have regarding some of the variations, not "opening systems"!  As such, TopNotch's insinuation that one who studies and strives to obtain the Belgrade Gambit in their games must somehow be conducting their opening preparation haphazardly, without an eye toward openings that will serve them "for a lifetime," or "build[ing] [their] entire Opening Repertoire around an offbeat system that could fall based on the discovery of a single powerful novelty" is simply asinine!

But HAD we actually been talking about such "opening systems" I could easily have pointed out that I have utilized 1.Nc3 as a "system" for many years, and that through this "system" one will frequently encounter the following line: 1.Nc3 e5 2.e4 Nc6 3.Nf3 Nf6 4.d4 exd4 5.Nd5, etc., and several variants among this move order, all leading to the same Belgrade Gambit.  Of course he may also encounter 1...d5, 1...c5, 1...g6 and several other possibilities, each of which white, using this "system," must be prepared for. 

In short, constructing your opening repertoire in a manner that improves the odds of reaching positions (or gambits) that you enjoy and understand is a recipe for success and should be encouraged.

Finally--and this is perhaps most important for those "amateurs" TopNotch was evidently trying to convince (are we to presume TopNotch is a "professional"?)--it is a categorical fallacy to suggest that the inclusion of off-beat gambits like the Belgrade (and the BG has been found to be "bullet proof" at the highest levels of correspondence chess for the last 40 years) in your repertoire is somehow a recipe for disaster.  In fact, the reverse is true.  Indeed, it is precisely through knowing such gambits--and using them--that offers lower rated players their best opportunities for success against higher rated players. 

What's interesting though is how TopNotch was on the right track when he said:

"they [the Bishop's Opening and Vienna Game] are strategically complex and unbalanced enough to allow the player that understands the typical resultant positions better to prevail."

But evidently this same standard of "understand[ing] the typical resultant positions" is insufficient if we're talking about the Belgrade Gambit or other gambits that happen to be out of the mainstream sights, even when said opening has been proven sound.

Bruce Monson

Craig wrote:
A) 5...Bc5
This line is often not mentioned, and sometimes only recieves a footnote, but I think it deserves close attention. The point is 6.Bf4 is met by O-O!, and 7.Bxc7 Qe8! is probably better for black. So instead best is probably 6.Bg5, when black usually chooses between 6...Be7 (which I don't feel is correct), or 6...d6. Quite a few players, including Mr Monson, have been known to play 6.Bf4 and only on 6...d6 play 7.Bg5, as this prevents ...Be7, but if black is aware of the O-O possibility, I'm not sure whether this is good for white.

[MONSON]
I have long viewed 5...Bc5 to be suspect, but I do not disagree with Craig that it deserves "close attention."

White has a few options here, but one I like is 6.Bf4, which I only slightly prefer over 6.Bg5 since it forces black to decide between playing 6...d6 or 6...O-O.  Most common, of course, is 6...d6, after which white plays 7.Bg5 and prepares to wreck black's pawn structure.

However, the immediate 6.Bg5 is fully sound and quite dangerous for black, especially if he now plays 6...d6?!, which is even better for white.  Better is 6...Be7 after which white must decide between 7.Bf4, transposing to 5...Be7 6.Bf4 lines, or playing 7.Bxf6 Bxf6 8.Bb5, which is actually quite an interesting variation.

But to give an example of how white's play might be geared around black's attempt to hold the d4-pawn, I'd like to present the following game against an ICCF 2450 player.  I am not going to provide detailed annotations, but the astute player will notice that white makes no attempt to recapture the d4-pawn, rather using this to his advantage in keeping black's Bc5 as a virtual spectator, cut off from the action occuring on the kingside by his own pawns on d6 and d4!   Indeed, black will often be compelled to expunge this pawn voluntarily (...d3) in effort to get his useless bishop back into play.  Other themes are open lines (even at the cost of a further pawn sacrifice) and keeping black's king quarantined in the center.


Monson,B - Simmelink,J (2450) [C47]
Belgrade Gambit Corr. Thematic, 1998

1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.d4 exd4 5.Nd5 Bc5 6.Bf4 d6 7.Bg5 h6 8.Bxf6 gxf6 9.Qd2! Be6 10.Bb5 a6 11.Ba4 b5 12.Bb3 Ne5 13.Qf4 Ng6 14.Qg3 Rg8 15.0-0-0 Kf8 16.Nf4! Bxb3 17.axb3 Nxf4 18.Qxf4 Rxg2 19.Nh4 Rg5 20.Nf5 d5 21.Rhg1 Bd6 22.Qh4 Ke8 23.Nxh6! c5 24.exd5 Rxg1 25.Rxg1 f5 26.Qh5 Bf4+ 27.Kb1 Bxh6 28.Re1+! Be3 29.fxe3 d3 30.e4! Qf6 31.cxd3 Rd8 32.Rf1 Qd4 33.Qxf5 Qxd3+ 34.Ka2 Rd7 35.Rg1 Ke7 36.Rg7 Kd8 37.Qe5 Kc8 38.Rg8+ 1-0

The game was very complicated and at several stages it required subtle handling, but that is to be expected against strong opposition.  The reality is, black's position is much more difficult to play than white's.

As for 6...O-O!? being the "refutation," well, that is simply not so.  Yes, if white gets careless he can see the game turn on him with a counterattack (though it is usually a threat to the Bc7 that Black has in mind in these lines), but there are all kinds of problems black must deal with, not the least of which is the complete destruction of his pawn structure after 6...O-O 7.Bxc7 Qe8 8.Nxf6+ gxf6.  Note first of all that black is not up any material at this time and yet he has severe structural weaknesses in his pawn structure that will kill him in almost any endgame.  Moreover, his king rests on this same side of the board and thus will always be vulnerable to attack.  And even in the event black can manage to get an extra pawn (i.e., white can legitimately sacrifice the e-pawn on purpose in some lines), white can still work against these permanent weaknesses in black's position with virtually no risk at all to himself.

Black's entire strategy, therefore, seems to revolve around the idea of playing ...d6 and ...Qe7, trapping the bishop on c7.  But white has several answer to this, including just retreating the bishop and leaving the e-pawn for the taking, or ideas involving Bd3 and Qd2 (eyeing h6) and offering to let black just win the bishop (with consequences, of course).

This position, incidentally, is not the same as in the 5...Be7 6.Bf4 O-O!? line, in which black has much the same ideas but does not have to suffer the destruction of his pawn formation directly around his castled
king.

Bruce Monson

I've been looking at the 5...Nb4 6.Nxf6+ lines over the last few days, and I must admit that 7.Bc4! is a significant improvement. 9.Ng5 seems the best move in the position after 7...Bc5 8.O-O d6, though also, perhaps 9.Bg5 is deserving of attention...I would probably prefer Ng5 given that it's a gambit I'm playing, and there seems no clear continuation after 9.Bg5 Qg6 10.a3 Nc6... maybe 11.b4 is an idea, and 11.Qe2 Bg4 with the idea of a later e5 may be playable, but 9.Ng5 certainly seems better.
I should note here that 9.e5 dxe5 10.a3 seems to be the main continuation, but I think white is struggling somewhat in these lines, the plan is to bust open the centre with f4 as soon as can be arranged, but trawling through my databases I see that black seems to be defusing the white play quite easily at the moment, and Mr Monson himself has had plenty of success on the black side in correspondence tournaments. (Incidentally, 10.Bg5 may be a far better try, in the games I found with this move, white is scoring 50%, 10...Qf5 11.Re1 seems to give white possibly enough for his pawn)

Just a note on the Nxc7+ piece sac line (Monson gambit?!), after 13.h4, I was wondering if Bruce (or anyone else) has considered or seen 13...b5!?. when having a little look at this, I found it to be quite an interesting defence... the point is that if 14.Qxb5+, 14...Qb7 and b7 turns out to be a decent square for the queen, and white can obviously not exchange. Have you looked at this Bruce, and if so, have you found a reliable method against it?

Regards,
Craig

<because the pawn structures that arise are predictable>
There is hardly one opening with more predictable pawn structures than the Petrov. As TopNotch sees himself as a kind of guru, leading me ignorant amateur in the right direction, maybe he can tell me what to do after 1.e4 e5 2.Nf3 Nf6.

perhaps Belgrade fans should post some analysis rebuffing recently published antidote- both Emms in "Play The Open Games as Black" and Kaufman in "Chess Advantage"... give 5...Be7.

I can see the Belgrade being a surprise weapon, but if these 1.e4 e5 gambits had any real bite, we would see them played in high level OTB tournaments. Correspondence is a different matter - I think more players will play "critical lines" (5...Ne4; 5...Nb4) but if the solid 5...Be7 equalizes without much difficulty, it's what players will prepare OTB.

TOPNOTCH:
The problems arise when one becomes obsessed with these Offbeat Gambits, and gambit play in general to the exclusion of other important strategic Opening concepts that one must learn in order to improve as a chessplayer.

MONSON:
Once again, you're talking apples and oranges!  It does not follow that just because someone happens to specialize in a particular gambit that that person must therefore be ignoring other "important strategic Opening concepts." 

TOPNOTCH:
Yes, I know we all want to tranpose into or play Openings we love, even if deep down we know they are flawed.

MONSON:
So far you have not provided one iota of evidence demonstrating that the Belgrade Gambit is "flawed" in any way.  The Belgrade has withstood the rigors of world class correspondence grandmasters as well as world class OTB grandmasters.  Moreover, I doubt there is anyone in the world who is more up on current Belgrade Gambit theory than I am, and yet I know of no "bombs" for black that leave me with any "deep down" concerns about the gambit being "flawed."

Theory has shown, and continues to show, that white has good prospects against all of black's 5th move responses, and that includes the allegedly devastating 5...Be7. 

I checked my database, and according to the 940 games with 5...Be7 (that's probably about as many games as you have TOTAL for the BG in your database), white won 374 (40%) while black won 257 (27%) and 297 were drawn (33%), for an adjusted score of 56% for white against 44% for black! with an average ELO rating of 2349!

I'd say 56% against this devastating show-stopper in 5...Be7 is pretty damn good!  Perhaps this is why GMs such as Gutman, Bellon, Szmetan,  Peter Svidler (he's only like, what, among the top ten GMs in the world?!) and others (including a multitude of IMs and correspondence IMs and GMs) continue to include this gambit in their repertoir. 

Perhaps, too, this is why GM Lev Gutman wrote in his 1993 book d4 im Vierspringerspiel (d4 in the Four Knights): ...I believe that White's position has great potential and I intend to prove that, in all variations, Black must contend with the worse position. . . . for years the Belgrade Gambit has been considered unsatisfactory for White but now I am of the opinion that White has the better chances and Black has no easy defense."

Even Karpov was impressed with it.  In his 1988 book, The Open Game in Action he said: ...[N]ot everyone wishes to sacrifice a pawn (or even two) so early in the opening to obtain the initiative.  For that reason the [Belgrade] Gambit hasn't become very popular.  All the same this gambit leads to quite exciting and lively play.  I think those who favor stormy complications should include the gambit in their repertoire.

Svidler, in fact, played the BG through the 90's and as recently as 2002 used it against Morozevich.  Incidentally, Morozevich played 5...Be7, and far from being a ho-hum yawnfest, it turned into a tactical melee with pieces flying all over the place.  It was VERY hard-fought!  Ultimately the final result was a draw, but it was later discovered that Svidler missed a win.  Here's that game:

Svidler,P (2690) - Morozevich,A (2707) [C47]
St Petersburg-Moscow Moscow (1), 22.10.2002

1.e4 e5 2.Nf3 Nf6 3.Nc3 Nc6 4.d4 exd4 5.Nd5 Be7 6.Bf4 d6 7.Nxd4 0-0 8.Nb5 Nxd5 9.exd5 a6 10.dxc6 [10.Nxc7 Qxc7 11.dxc6 Re8 12.Be2 bxc6 13.0-0 Rb8 14.b3 +/= Monson; For 10.Nc3 see game ven der Weide-Ellenbroek below.] 10...axb5 11.Bxb5 Bg5 12.cxb7 Bxb7 13.Qg4 Qe7+ 14.Kd2 Bf6 15.Rhe1 Qd8 16.c3 Rb8 17.Kc2 Bc8 18.Qe2 c6 19.Bd3 Be6 20.Qd2 g6 21.Kc1 Qa5 22.Bxd6 Rxb2 23.Qxb2? (Svidler could have won with 23.Kxb2!) 23...Bxc3 24.Qa3 Qxa3+ 25.Bxa3 Ra8 26.Rxe6 Bxa1 27.Rxc6 Rxa3 28.Bc4 Bd4 29.f4 Ra5 30.Kd2 Rc5 31.Rxc5 Bxc5 32.a4 Kf8 33.Kd3 Bb6 34.Ke4 Ke7 35.f5 g5 36.Kd5 g4 37.Kc6 Bg1 38.h3 gxh3 39.gxh3 Kf6 40.Bd3 Kg5 41.a5 Kh4 42.a6 f6 43.Kb7 Kxh3 44.a7 Bxa7 45.Kxa7 Kg4 46.Kb6 Kf4 47.Kc5 Ke5 ½-½

The following game, won by BG expert Karel Van der Weide, occurred in the same variation as the previous game, though the position arose by transposition after . . . not 1.e4 but 1.Nf3!  It sure was lucky for Karel that his opponent played ...Nc6 and ...e5 or else he wouldn't have known what to do. . .

Van der Weide,K - Ellenbroek,T [C47]
NLD-ch sf Enschede (4), 1995

1.Nf3 Nc6 2.e4 e5 3.Nc3 Nf6 4.d4 exd4 5.Nd5 Be7 6.Bf4 d6 7.Nxd4 0-0 8.Nb5 Nxd5 9.exd5 a6 10.Nc3 Ne5 11.Be2 Bg5 12.Bg3 Ng6 13.0-0 Bf4 14.Qd4 Qg5 15.Rae1 Bf5 16.h4 (just another boring, drawish, position arising from 5...Be7) 16...Qh6 17.h5 Be5 18.Bxe5 Nxe5 19.f4 Ng4 20.Bxg4 Bxg4 21.f5 Bxh5 22.Re3 f6 23.Rh3 Qg5 24.Ne4 Qg4 25.Rg3 Qe2 26.Nxf6+ Rxf6 27.Qxf6 Qe3+ 28.Rxe3 gxf6 29.Re7 Rc8 30.Rf4 Bf7 31.Rc4 c5 32.Rxb7 Bxd5 33.Rg4+ Kh8 34.Rd7 Bxa2 35.Rxd6 Bb1 36.Rc4 Ba2 37.b3 a5 38.Ra6 Rd8 39.Rxa5 Bb1 40.Raxc5 1-0


TOPNOTCH:
The truth is you will never learn very much about the intricacies 1e4 e5 play if you restrict yourself to something like The Belgrade Gambit,

MONSON:
Again, apples and oranges!  Just because one happens to be an expert in the Belgrade Gambit DOES NOT mean that that is their ONLY focus, or that they neglect other aspects of the game.


TOPNOTCH:
why, because the pawn structures that arise are predictable and do not allow for much flexibility in strategy.

MONSON:
In my next post I will present numerous games involving the three main lines white generally plays against 5...Be7 (namely 6.Bf4, 6.Bc4 and 6.Bb5, all of which offer dynamic possibilities ranging between tactical skirmishes and positional nuances).  I'll let the other readers decide whether they are as benign as TOPNOTCH is under the illusion they are.  You'll notice that I didn't mention 6.Nxd4 (the line Craig originally provided a short line for, after which TopNotch chimed-in with his "this confirms what I already knew" diatribe).  That's because 6.Nxd4, while perfectly playable, is the least strong of white's options.

Just a short addition here.  Let me say that the Belgrade is actually more of a positional gambit than a typical kill-or-be-killed bloodbath.  Certainly, in nearly every major line there are variations that will send the game into bloodly tactical complications, but even when these are avoided white is usually left with the pleasant option of maintaining the tension or heading into a slightly better endgame.  Indeed, in addition to honing your tactical skills, the BG will also hone your endgame skills.  I will be providing examples to demonstrate this.

Regards,

Bruce Monson