Looking for interesting lines in the KG I noted the attached game and was quite amazed about Blacks 11...Nb4! It hangs the knight en prise, yet Black won after 12.Nxb4 and the computer assessment points at good play for Black after this sacrifice. However, I am in doubt if White with best play can actually force a Draw after 12.Nxc2+, i.e: 12...Kd8 13.Nxa8 Qe3+ 14.Kc3 Qc5 15.Kd2 Bg7 16.Be2! Rg8 17.c3 Bxc3 18.bxc3 Qe3+ 19.Ke1 Qg3 20.Kd2 etc. with perpetual check.
Any ideas how Black can avoid this and stay in the game with reasonable play?

Dear Dragonslayer,

thank you for your honest and enlightning answer!

After some analysis (and with the help of your ideas), I found some attractive possibilities for white:

After 18. ... Kd7 19.Bb5! (19.c3 followed by Bc4 is interesting, but black can activate the rook by Rae8, but 19.Bc4 fails to Nfd4: protecting e6) a6 20.Ne6 Qg6 21.Bd3! (the bishop retreats and pins the knight) Rae8 22.Qf4 Nce7 23.d5! and now the white queen can switch back to the queenside with an attack.

If instead 20. ... Qh6 the queen feels uncomfortable there: 21.Bc6:+ bc6: 22.Qa5 Rac8 23.h3!! (the knight on f5 was immune due to Qd2+ with mate, but white works to diffuse the threat). And the black pawn structure falls apart while white keeps the initiative.

I think your opponent was quite right to be afraid of 16. ... Kd7 17.Bc4 when Be6+ forces Kd8 anyway and white can open the kingside by h3 with a vicious attack!

Slightly off-topic but too lazy to locate the correct thread ......
Just read fellow ChessPubber Sverre Johnsen's blog (reactivated after some period of dormancy) that he is writing a manuscript on the KG (see http://sverreschesscorner.blogspot.com/2011/09/brief-return-to-chess-blogging.ht...)

I wonder if he can provide more info ......

OK.
Congrats with a fine game.

Ok, here are some of my findings after 3...g5 4.Nc3 Bg7
5.g3!? is interesting and the only way to avoid the piece sacrifices. It is neither better nor worse than the other lines, just different. I am not so worried about 5...g4 6.Nh4 d5.

5.d4 g4 6.Ne5 Bxe5! is bad for White.
5.d4 g4 6.Bc4 gxf3 7.Qxf3 is likely to transpose to the Rosentreter when White later plays Bxf4 (as observed by MNb) or the Pierce if Black plays ...Nc6. But Qxf4 is not out of the question in all lines.

After 5.d4 g4 6.Bxf4 gxf3 7.Qxf3 d6 8.Bc4 Be6 both 9.d5 and 9.Bxe6 are reasonable. But 8.0-0-0 is also interesting.

After 7...Nc6 8.Bc4 with transposition to the Pierce gambit is not forced. There are at least two other options that can be investigated.
Nocci-Ryan is a draw after 8...d6! (Line C) 9.Bxf7+. White has one better try for an advantage at move 9.

8...Nxd4? (line B) is a mistake and White is winning after 9.Bxf7+ Furhoff-Kallgren is an improvement over Motzko-Marco at move 12.

In fact also the McDonnell Gambit might just transpose:
1.e4 e5 2.f4 exf4 3.Nf3 g5 4.Nc3 Bg7 5.d4 g4 6.Bc4 gxf3 7.Qxf3 d6/Nc6 and has White better than 8.Bxf4 ? This is what actually happened in that Motzko-Marco game. So it seems just to be a pseudochoice.

PatzerNoster wrote on 08/22/11 at 17:38:15:

If you have faith in White's chance, by all means try it. Who knows? I am not convinced after 7.Nxd5 f3, but there is no reason to trust my judgment more than your own.
You might also take a look at 5.d4 g4 6.Ne5 Qh4+ 7.g3 fxg3 8.Qxg4 Qxg4 9.Nxg4 Nc6 10.Bf4.

After 6.Bxf4 gxf3 7.Qxf3 Nc6 8.Bc4 there are two more games:

Motzko-Marco, Cesky Tesin sim 1906
Polasek-Cernousek, CZEchT 2007.

PatzerNoster wrote on 08/22/11 at 14:19:05:

Yes, no. Why should a perfectly logical move like Bg7 harm Black?

Nocci,R (2311) - Ryan,P (2291) [C37]
WSTT/1/06/Semifinal 1 - King's Gambit,  ICCF, 25.05.2007
½-½



The Rosentreter has been discussed here on this forum a few years ago. The general verdict was that it's a forced draw indeed. Now your question, if I understand you correctly, if White can avoid a forced draw after 4.Nc3 Bg7 5.d4 g4. I'm not sure what the answer is, but I see no reason why Black should be worse. All of Black's moves look perfectly sound to me.

Furhoff-Kallgren, Stockholm Open 1994, saw this line. Furhoff is the pioneeer of White's opening strategy. There is also a thematic corr. game Nocci-Ryan, WSTT 2007. In both games Black played 7...Nc6, which just transposes. So does 7...d6.
To make things even a bit more confusing, play is also similar to the Pierce Gambit 1.e4 e5 2.Nc3 Nc6 3.f4 exf4 4.Nf3 g5 5.d4 g4 6.Bc4 gxf3 7.0-0 d6 8.Qxf3.

A theoretically quite irrelevant game:

MNb - Koller,G [C37]
em EM/H/135, 2004

Markovich wrote on 10/31/08 at 17:43:02:


I'm gunna take a wild guess at the meaning of this analogy and go with FLEETING .

Tops

ANDREW BRETT wrote on 10/30/08 at 17:13:53:

Fashion? Anyhow, the line 4.d4 g5 5.Nc3 and g2-g3 at the right moment offers decent compensation. Of course this does not mean that Black is worse.

FightingDragon wrote on 10/14/08 at 10:00:40:


Actually 10.Ne4 f5 11.Ng5 Nd7! is very good for Black (I discussed this with Bücker and tried everything: 11.Nxd6 cxd6 12.Nd3 seems best).
10.Ne2 has its problems too. Instead of Riemann's 10..Re8 Black has at least two better choices.

So it seems 9.Nc3 is under a cloud. However Bücker did a good job of putting 9.0-0 back on the playing field.

I don't think that' a fair comparison. In the main variation of the Kieseritsky black is an important tempo up (Nf6).

I think I know which line you mean:

1.e4 e5 2.f4 ef4: 3.Nf3 d6 4.d4 g5 5.h4 g4 6.Ng1 f5 7.Nc3 fe4: 8.Bf4: Nf6 9.Qd2?! d5! and we have Hector-Leko, 1995.

After that, it was discovered that white is fine after 9.d5!

But I always asked myself what happens after 8. ... d5!
After 9.Be5 Nf6 both 10.Bf6: Qf6: 11.Nd5: Qd6 -/+ and 10.Nd5: Qd5: 11.Bf6: Bb4+! 12.c3 0-0 13.Bg5 Bd6 =+ are better for black.

Alas I don't really see a better move than 9.Qd2, but maybe white is OK nevertheless.

Another interesting possibility is 7.Bf4: fe4: 8.c4 preventing d5, although it looks a little loosening.
Or 8.Qd2!? d5 (Nf6 9.h5 waiting for ... d5 10.c4!) 9.c4! which also looks interesting.

Reminds me of what Robert Byrne said long ago (apropos of why he stopped playing the KG):  "Black has one defense after another, all of them good."

I always thought the main reason for GMs giving up the KG was the main line Kieseritzky: 3.Nf3 g5 4.h4 g4 5.Ne5 Nf6, when the apparently sub-optimal 6.d4 has traditionally been the modern preference.  It should be dynamically equal if White chooses 6.Bc4 though.

As far as I'm aware the Fischer Defence with 3...d6 hasn't really caught on at master level in the way that 3...g5 or even the Modern (3...d5) has.

I recall I tried the 3...d6 4.d4 g5 5.h4 g4 6.Ng1 line as Black in a couple of casual games.  I really don't think the idea is as bad as it first appears- the f4 and g4 pawns are weak, the knight can redeploy on e2 to attack f4, and the black kingside is full of holes.  I reckon White has full compensation for the pawn in this line.

FightingDragon wrote on 10/14/08 at 10:00:40:


Because I dislike developing backward and have seen some games on respectable level where White did not do well. If my memory serves me well Leko had a nice win as Black. I am aware of Gallagher's opinion and also some Hebden wins, but both have given up the KG anyway, haven't they? Dissatisfaction with 6.Ng1 might have to do something with it - or perhaps not.
Sorry, I don't have the time at the moment to look closely at this variation and I assume you can find some relevant games yourself. I will be happy to be converted.

Excellent thread. Lots of new ideas.

• 1.e4 e5 2.f4 exf4 3.Nf3 Nc6 4.d4 g5 5.h4 g4 6.Ne5 (or 3...g5 4.h4 g4 5.Ne5 Nc6 6.d4) is a bit overrated (i.e. theory says +/- which a way too optimistic) White has other options. 4.Nc3 or 5.Nc3.
• 1.e4 e5 2.f4 ef4 3.Nf3 g5 4.Nc3 Nc6 5.g3 is not a TN. There was the game Littlewood-Kloster, Davos SUI 2004. Before that game I analyzed 5.g3:
  5...d6 6.d4 Bg7 gives White the play he wants. Better seems 6...g4 or even perhaps even better 5...g4 6.Nh4 Nd4!? blocking the d-pawn. I was unable to find a good continuation after this. Perhaps someone else is more lucky.
• 1.e4 e5 2.f4 ef4 3.Nf3 g5 4.d4 g4 5.Bf4 gf3 6.Qxf3 d6 7.Nc3 Nc6 8.Bc4 Nd4 9.Bf7+ Kf7 10.Qh5+ Kg7 11.0-0 Nf6 (=+ apparently, but) 12.Bh6+ Qg8 13.Qg5+ Kf7 14.Qh5+ Ke6 15.Qh3+ Ke7 16.Qh4 Nf5 -+ Fedorov (Informator 69) However 16.Nd5+ Ke8 17.Qd3 looks like an easy improvement and unclear in my view (don't trust the computer when it says =+).
  Here there is also' Carmelo Coco's 8.Bb5!?
  In the Quaade move-order 7.Bc4 is an important alternative for White. If White enters from the Fischer defence via 3...d6 4.Nc3 he is deprived of this option.

@Fightingdragon (no secrets)


1.e4 e5 2.f4 ef4: 3.Nf3 g5 4.d4 g4 5.Bf4: gf3: 6.Qf3: d6 7.Nc3 Nc6 8.Bc4 Nd4: 9.Bf7:+ Kf7: 10.Qh5+ Kg7 11.0-0 Be6 12.Rad1 Nf6 13.Bh6+ Kg8 14.Qg5+ Kf7 15.Qh5+ ... (Salmensuu-Alexandrov, Istambul, 2002)

11....Nf6 =/+(Fedorov) and there is no forced draw)

11.0-0-0 (Salmensuu's prefered this 2 years after the above game)
11......Qf6! 12.Nd5, Qg6 -/+ (Salmensuu-Nysti, Helsinki,2004)

======================================
MNb's remarks on 1.e4,e5 2.f4,exf 3.Nf3,d6 have
to do with move-orders and transpositions:

4.d4,g5 5.Nc3,g4 and black has no direct need for ...h6
4.Nc3,h6 (4...g5!) 5.d4 Transfers to Becker's Defence.
Though I doubt that white has an advantage here
[Mnb] probably thinks of the following variation he recommended in previous posts:

4.Nc3,h6
5.d4,g5
6.g3!?,fxg3 (after 6...g4 7.Nh4,f3 black has spent a tempo on the not so usefull ...h6)
7.hxg3,Bg7
8.Be3         (Mnb; More common is 8.Bc4,Nf6 though neither Morozovich nor Grishuk could attain advantage playing white)

A recent game, showing black has not really much to fear:
8.Be3,Bg4 9.Be2,Nd7 10.Qd3, c6 [unclear], Shulman-Formanek,Conneticut,2007

I agree with Mnb on the second part. The variation
3...g5 4.Nc3,g4 5.Ne5,Qh4+ 6.g3,fxg3 7.Qxg4,Qxg4 8.Nxg4,d5!?
was actually played in Pillsbury-Maroczy, Vienna, 1903 and
quickly led to a drawish endgame.
Even though there is not much news under the sun. Practically all suggestions, have been discussed previous in this section of the forum.
Subscribers to paid 1.e4,e5 section can verify this very quickly, as there is a small separate database for download there that contains the forum analysis and more.

Is nobody interested in the king's gambit anymore?

Did I post too many variations or is my suggestion entirely silly? 

After checking Vadim Zvjaginsevs games in the King's gambit I have some new ideas for white:

1.e4 e5 2.f4 ef4: 3.Nf3 g5 4.Nc3!?

now 4. ... g4 5.Ne5 is at least OK for white.

the game Zvjaginsev-Smirnov, 2008 went 4. ... Nc6 5.g3!? d6 6.d4 Bg7 7.d5 Ne5 8.gf4: gf4: 9.Bf4: Bg4 10.Bb5+ Kf8 11.Be2 with a balanced position.

what about 4. ... d6 5.g3!? (the new idea)? I couldn't find any games with that move. The idea is to transpose to the lines discussed in the Fischer defence thread (which to my mind give white nice compensation) without allowing the Rosentreter which I think is better for black (like after 5.d4 g4!).
After 5. ... g4 6.Nh4 f3 7.d4 play is very likely to transpose to these lines. One independent try is 7. ... Be7 8.Be3 (8.Nf5!?) Bh4: 9.gh4: Qh4: 10.Bf2 Qh6 when black is one tempo up compared to the Furhoff-Eriksson game from the Fischer defence. Nevertheless I think white has compensation after 11.Nd5 Kd7 12.Qd3 with the idea of 13.Be3 and 0-0-0 or even 12.Qd2!?

Black can also play 5. ... fg3: 6.hg3: Bg4 (Bg7 7.d4 h6 is the Becker defence) 7.d4 h6 (Bg7 8.Bc4 Nc6 9.Rf1 Qe7 10.Be3 h6 11.Qd2 Nf6 12.0-0-0 0-0-0 13.Rde1 and white has compensation, the question is if it's enough) 8.Bc4 Bg7 9.Rf1 and again the game has transposed to the Becker defence.

If this is a way to play the Fischer lines without allowing the Rosentreter, I think the King's gambit becomes interesting again!
I would like to hear your opinion! 

Stefan Bucker has posted an article on the main line of the Kieseritzy on Chess Cafe. An expanded version of this article will appear in the next issue of Kaissiber. In my opinion this patches up a line of the King's Gambit where white was struggling.

I have played the King's Gambit from both sides for many years both over the board and in correspondence. I am currently happy playing against every line from black. Black can equalize in some lines if he is very careful. However many of these positions are even but not drawish. Some of the positions that arise are very difficult and quite different to other openings.      

Here we go again, this time bashing the Allgaier.

1.e4 e5 2.f4 exf4 3.Nf3 g5 4.h4 g4 5.Ng5 h6 6.Nxf7 Kxf7
-7.Nc3 d5 8 d4 f3 9 Nxd5 Nf6 10 Nxf6 Qxf6 11 gxf3 Qxf3 12 Bc4+ and I think Ke8 is more than =+, eg 13.Qxf3 gxf3 14.Kf2 Bg4 15.h5 Nc6 16.c3 Na5 17.Bd3 c5.
-According to an analysis from Chabelsky 7.Bc4+ d5 8.Bxd5+
a)8...Ke8 9.d4 (but then why not 7.d4 immediately?) Nf6 10.Nc3 Nh5 11.0-0 c6 12.Bb3 Bg7 3.e5 Tf8 14.Ne4 Qxh4+ 15.Nd6+ Kd7 and Black has the extra piece plus the winning attack ...
b)8...Kg7 9.Bxb7 Bxb7 10.Qxg4+ Kf7 11.Qh5+ Ke7 12.Qe5+ and 100 years ago one knew better than current sicilon power: Kd7 13.Qxh8 (13.Qf5+ Kc6 14.Rh3 a5! and the black king escapes) Nf6 14.e5 Bxg2 (Fritz6 thinks White is better, but we should rather rely on the old masters) 15.Rg1 (15.Rh2 f3 16.Qxf6 Qxf6 17.exf6 Bc5, Gunsberg in 1895; Fritz6 gives 18.d4 Bxd4 19.Nd2 Nc6 20.f7 Ne5 but has changed its evaluation drastically) f3 16.Qxf6 Qxf6 17.exf6 Bc5 (Lewenfisj 1910). All this shows that Eric Schiller's books sometimes are useful; it's to be found in his "Who is afraid of the KG". Heartwarming that these old masters understood stuff like this better than 21st century computers.
-So I tried to revive 7.d4, similar to the HAG.
a)7.d4 f3 (this thematic move is probably even stronger than d5 8.Bxf4 Nf6 9.Nc3 Nxe4 or Be6) 8.Bc4+ d5 9.Bxd5+
a1)9...Ke8 10.gxf3 Nf6 11.Bb3 gxf3 12.Be3 Nc6 and I doubt White's compensation.
a2)9...Kg7 10.gxf3 Nf6 11.Nc3 Bb4 12.Bc4 c5! -+Le Goff-Bauget, corr 1993 is an improvement on the old 12...gxf3.

Sidenote, possibly not very relevant: C2 (see previous page). After 7...Nf6, Harding's main line is 8 d4 d5 9 Bxf4 Bb4 10 Bd3 and now Bxc3+ 11.bxc3 dxe4 12 Bc4+ Be6 and in no single line Fritz will drop his evaluation, as far as I can see

On the topic of how to answer 2 f4 exf4 3 Nf3, I'm not convinced that Black is better with best play after 3...g5.  Black can certainly equalise, but it's difficult to prove more than that.  The Hanstein Gambit, while unpromising for White, probably equalises, while in the Kieseritzky, see Stefan Bucker's analysis of 4 h4 g4 5 Ne5 d6 Nxg4 Nf6, and 5...Nf6 6 Bc4! (rather than the more ambitious, but insufficient, 6 d4 d6) d5 7 exd5.
http://www.chesscafe.com/text/kaiss34.pdf
http://www.chesscafe.com/text/kaiss35.pdf

Also, I wonder if Fritz's 7 Bc4+ d5 8 Bxd5+ Kg7 9 Bxb7 mentioned above is wholly sound in the Allgaier; if it is, even the Allgaier may be good enough for equality.

As Black (I sometimes meet 1 e4 with 1...e5) I quite like the Cunningham in response to the KG, and I also like the "nameless 3...Nc6 defence" followed by 4...g5- it prevents the Kieseritzky and in my experience often confuses White players who are expecting 3...g5.

MNb wrote on 07/29/08 at 01:57:37:



1.  After 8 d4 Bb4 I reckon White can try 9 Bc4+ d5 10 Bxd5+ Kg7 11 Bxf4, unclear, as Bb4 isn't usually one of Black's best moves in the dodgy 8 Bc4+ line.

2.  Having another run-through of 8 d4 d5 9 Bxf4 Nf6, I'm not sure that that stretch of the article is among Harding's best, as he glosses over Glazkov's 10 Bb5+ Bxc3+ 11 bxc3 Nf6 12 0-0 Kg7, when 13 exd5 is fine for White, and Rodricht's 14 Bxg4 in the 10 Be2 line, which also looks OK (though I feel Black has improvements in the 13...dxe4 line, I don't trust 10 Be2 and 13 c4 at all).

As well as 10 Bb5 which is about equal, 10 Bd3 is also better IMO than 10 Be2.  So maybe even that line may be playable for White.

3.  8 d4 d5 9 Bxf4 Be6 (the immediate 8...Be6 is illegal!) can be met by 10 Be2, undermining the g4 and d5-pawns, while 10 Bd3 isn't too bad either, e.g. 10...Nf6 11 exd5 Nxd5 12 0-0.

Anders wrote on 07/29/08 at 13:13:52:


Well, I usually play the Modern but I certainly would also consider the Cunningham.  For whatever it's worth.

I apologize that it's not in the spirit of this thread, but personally I would never consider 3...g5.  I understand that Black is better, but there are much simpler defenses that utterly frustrate White's quest for (a) the advantage and (b) the thrill of playing his heavily prepared, antique system ca. 1865.

Here's Harding's articles:
http://www.chesscafe.com/text/kibitz78.pdf
http://www.chesscafe.com/text/kibitz79.pdf

I've checked over Gallagher's NCO line with the assistance of Fritz 10, and think there may be improvements for White over it, but nothing particularly inspiring.  My analysis of the Allgaier from tonight's survey (which probably has many holes) runs:

A.  7 d4 f3 (7...d5!?) 8 Bc4+ d5 9 Bxd5+ Kg7 10 gxf3 Nf6 11 Bc4 (11 Nc3 Bb4 -/+ Gallagher, e.g. 12 Bc4 c5!) 11...c5 12 e5 Nc6 =+.

B.  7 Bc4+ d5 8 Bxd5+ Kg7 (8...Ke7 9 d4 gives White sufficient compensation) 9 Bxb7 is Fritz's suggestion (9 d4 f3! =+, transposes to line A) 9...Bxb7 (9...f3 10 Bxc8 f2+ 11 Kxf2 Qxc8 12 d4 with compensation) 10 Qxg4+ Kh7 (10...Kf7 11 Qh5+ Ke6 12 Qf5+ Kd6 13 d4 with compensation) 11 Qf5+ Kg7 12 Qg4+, with a repetition draw.

C.  7 Nc3

C1.  7...Bb4 8 Bc4+ d5 9 Nxd5 Be6 10 Qe2 with compensation.

C2.  After 7...Nf6, Harding's main line is 8 d4 d5 9 Bxf4 Bb4 10 Bd3 dxe4 11 Bc4+ Kg7 12 0-0 Nc6 13 d5, which is unclear.  This is one of those lines where Fritz initially thinks Black is much better, but push it down its suggested lines, and the evaluation drops substantially.

C3.  7...d5 is the one move that might be better than 7...Nc6.  8 d4 (8 Nxd5 Nf6 9 Nxf4 Qd4 10 d3 Bc5 -/+) 8...f3 9 Nxd5 Nf6 10 Nxf6 Qxf6 11 gxf3 Qxf3 12 Bc4+ Kg7 13 Qxf3 gxf3 =+.

C4.  7...Nc6 and now:

C4a.  8 Bc4+ d5 9 Bxd5+ Kg7 10 d4 Bd6 (10...f3 11 gxf3 Be7 12 Be3 Bxh4+ 13 Kd2 is unclear, 10...Nf6 11 Bxf4 with compensation) 11 Bxc6+ (11 e5 Nxe5 12 dxe5 Bxe5 leaves Black a pawn up and destroys White's centre) 11...bxc6 and I think Black may be slightly better, though the position is still murky.

C4b.  8 d4 d5 (8...f3 is probably best met by 9 Be3 with another unclear position, rather than 9 Bc4+ which tends to transpose to line C4a) 

C4b1.  9 Bxf4 Bb4 is good for Black, but I think White's two move-9 alternatives are worth another look: 
C4b2.  9 exd5 Qe7+ (9...Nce7 10 Bxf4 Ng6 11 Be5 Bg7 12 Qe2 unclear) 10 Kf2 Ne5 =+
C4b3.  9 Nxd5 Nf6 (9...f3 10 gxf3 Nge7 11 Bc4 Nxd5 is considered good for Black according to Harding, but White seems fine to me: 12 Bxd5+ Kg7 13 Be3 with compensation) 10 Bxf4 is extremely messy.

In conclusion, unless I've missed something (which I might well have done), White's best move in the Allgaier seems to be 7 Bc4+ with sufficient compensation, though Black can force a quick draw in that line.  7 Nc3 is best met by 7...d5 with an edge for Black, while the Hampe-Allgaier with 7...Nc6 looks unclear.  It also appears that MNb is right about 8 d4 with the intention of queenside castling, rather than 8 Bc4+, being White's best approach in the Hampe-Allgaier lines.

Tim Harding wrote an article on the Allgaier a while ago, in which he concluded that 7 Nc3 was the best move, but it was still doubtful as to whether White got enough against accurate defence.  Black's best may be to transpose to the Hamppe-Allgaier with 7...Nc6, when Harding suggests 8 Bc4+ as a largely unexplored possibility.

7 d4 and 7 Bc4+ are inaccurate in view of the line given in NCO.

One last addition: a transposition table for all the first 6 moves or so of the ...g5 systems. No evaluations on here, just move orders. Anything in bold represents the primary move order to achieve individual systems within the ...g5 strategy, italicized lines are transpositions. This is important because much of the future viability of the KG might lie in move orders avoiding unpleasant positions.




I also found these notes, which talk about some of the naming conventions and general strategy ideas:

General Guidelines (Black plays ...g4)

• The Kieseritzky is probably the most logical continuation
• Lines with 4 Bc4/d4/Nc3 either:
  
  • require White to sacrifice a piece
  • allow Black to play …Qh4+
• lines that prepare …g4 with …d6/Nc6 prohibit Nf3-e5
• White’s light-squared bishop usually belongs on c4, hitting f7 (though only after …g4 has been played because of 4 Bc4 Bg7!)

General Guidelines (Black sets up with Bg7/d6/h6)

• If White plays Bc4, it’s a Hanstein Gambit
• If White plays Bc4 and h4, it’s a subset of the Hanstein called the Philidor Gambit (or Greco-Philidor Attack)
• If White plays 4 Bc4 Bg7, 5 d4/Nc3 still allows Black the …g4 option
• If White doesn’t play Bc4
  
  • and Black plays …h6, it’s a Becker Defense
  • and Black plays …d6, it’s a Fischer Defense
  • and Black plays …Nc6 without …d6, it’s a Pierce Gambit Declined
• White’s light-squared bishop usually belongs on f1 (guarding h3 in some lines) or b5 (after …Nc6)

Other Tries
Outside of the Muzio, all these tries are considered suspect to varying degrees. I’ve seen a few comments about these on ChessPub but most work has gone into the Kieseritzky, Rosentreter, or Quaade. I’ll just give what I have in NCO and McDonald’s book. There’s a lot of transpositional possibilities between these, the Rosentreter, and the Pierce.

[1] Muzio Gambit (4 Bc4 g4 5 O-O) NCO gives 5…gxf3 6 Qxf3 Qf6 7 e5 Qxe5 8 Bxf7!? (8 d3 Bh6 9 Nc3 Ne7 10 Bd2 with comp.) 8…Kxf7 9 d4 Qxd4+ (9…Qf6 10 Bxf4 +/-) 10 Be3 Qf6 11 Bxf4 (Yoos-Kirton, Saskatoon 1994) with compensation. Michael Jensen claims 8 Bxf7 is bad (Fritz seems to agree, suggesting 11…Ne7 12 Nc3 Qc6 and 11…Nc6 12 Nc3 Bc5+). He and MNB recommend looking further at 7 d3, given by Bucker. The big problem with the Muzio is 4 Bc4 Bg7! and the Hanstein/Philidor complex.

[2] Allgaier Gambit (4 h4 g4 5 Ng5) NCO gives 5…h6 6 Nxf7 Kxf7 7 d4 f3! 8 Bc4+ d5 9 Bxd5 Kg7 10 gxf3 Nf6 11 Nc3 Bb4 -/+

[3] Ghulam-Kassim Gambit (4 Bc4 g4 5 d4) After 5…gxf3 6 Qxf3 d5! 7 Bxd5 Nf6 8 O-O c6! there is 9 Bxf7+ Kxf7 10 Qxf4 Bg7 11 e5 Rf8 12 exf6 Kg8! (Zak) and 9 Nc3 cxd5 10 exd5 Bg7 11 Bxf4 O-O 12 Bg5 Bg4! (Fritz improving on ECO) 13 Qf4 Nbd7 -/+

[4] McDonnell Gambit (4 Bc4 g4 5 Nc3) Now 5…gxf3 6 Qxf3 d5 7 Nxd5 Nc6 8 O-O Bd6 9 d4 Nxd4 10 Qh5 Be6 11 Bxf4 Bxf4 12 Nxf4 Nf3!! 13 Rxf3 Qd4+ 14 Kh1 Bxc4 15 c3 Qxe4 =+ (McDonald)

[5] Salvio Gambit (4 Bc4 g4 5 Ne5) McDonald gave 5…Qh4+ 6 Kf1 Nc6! 7 Nxf7 Bc5 8 Qe1 g3 9 Nxh8 Bf2 10 Qd1 Nf6 11 Be2 d6 12 c3 Bg4 13 h3 Ne5 14 d4 f3 15 Bxf3 Nxf3 16 gxf3 g2+ 17 Ke2 Bxf3+ (Goncarenko-Alekseev, Corr 1963) with 18 Kxf3 gxh1Q+ 19 Qxh1 Qxe4+ advantage to Black, but it seems play can be improved for both sides at times.

[6] Lolli Gambit (4 Bc4 g4 5 Bxf7+) has the distinction of being the line Neil McDonald was most pessimistic about in his already pessimistic (for White) KG book. He gives 5…Kxf7 6 Ne5+ Ke8 7 Qxg4 Nf6 8 Qxf4 d6 9 Nf3 Rg8 -/+

[7] Sorenson Gambit (4 d4 g4 5 Ne5) McDonald has some analysis on this (mostly quoting Zak, it seems), and it looks like White has a draw after 5…Qh4+ 6 g3 fxg3 7 Qxg4 f2+ 8 Qxh4 gxf1Q 9 Nc3. Charles Kennaugh comments “The main drawback to this line is that if black plays 7...g2+ 8.Qxh4 gxh1Q white only seems to have a draw which he should probably take immediately with 9.Nxf7 Kxf7 10.Qh5+ Kg7 11.Qg5+ etc. Black does himself no particular favours playing for a win with 9...Be7 10.Qh5 Qxe4+ 11.Kd1 Qxd4+ 12.Bd2 Qf6?! (12...Nf6?? 13.Nd6+ Kd8 14.Qe8+!; 12...Kf8! 13.Ne5 Qf2 14.Bh6+ Nxh6 15.Qxh6+ Ke8 16.Qh5+ Kf8=) 13.Nd6+! Kf8 14.Qe8+ Kg7 15.Bc3 Bxd6 16.Bc4.”

McDonald quotes ECO with 7…Qxg4! 8 Nxg4 d5 9 Ne3 dxe4 10 hxg3 Nc6 11 Bb5 Bd7 where White has some compensation (Black’s kingside structure) but not enough. (NCO agrees, giving the position =+) This line is pretty similar to the Quaade above, but inferior as the developed c3 knight is probably worth more than the move d2-d4.

Kennaugh gives 9.Ne5! f6 10.Bf4 Bg7 and now instead of the piece sacrifice 11.exd5?! dxe5 12.dxe5 Nd7 13.Bb5 a6 14.Bxd7+ Bxd7 which Glazkov gives as slightly better for black (though that assessment is untested) white can simply play 11.Nd3 with the idea 11...dxe4 12.Nc5 f5 13.Nc3! with a strong initiative for the sacrificed pawns.

References: McDonald, Bucker (esp for 7 d3 Muzio according to Jensen), NCO, Charles Kennaugh (link above) for the Sorenson Gambit.

Quaade Gambit (4 Nc3 g4 5 Ne5)

• [1] 5…Qh4+ 6 g3 fxg3 7 Qxg4 g2+? 8 Qxh4 gxh1Q 9 Qh5! +-
  
  • a) 9…Be7 and now Bird-Anon. 1886 features a nice smothered mate after 10 Nxf7 Nf6 11 Nd6+ Kd8 12 Qe2 Rxe8 13 Nf7#
  • b) 9…Nh6 10 d4 d6
    10…Bb4 11 Bxh6 Qxe4+ 12 Kd1 Qxd4+ 13 Ke2 wins; 10…d5 is similar to line given
    11 Bxh6 Be6
    11…dxe5 12 Qxe5+ Be6 13 Qxh8 is good for White
    12 O-O-O gives tremendous compensation for the exchange. (Analysis by Ben Hague)
• [2] 5…Qh4+ 6 g3 fxg3 7 Qxg4 Qxg4 8 Nxg4 d5 9 Ne3! (Craig Evan’s idea)
  
  • a) 9…d4 10 Nb5 Na6 11 Nxd4 (Evans-Canizares, IECC 2004) += or +/-
  • b) 9…c6 10 exd5 Nf6 11 hxg3
    Fritz likes 11 dxc6 Nxc6 12 Nb5
    11…Bd6 12 Ne2 Nxd5 13 b3
    13 d4 is better, keeping the queenside pawns together after knights are exchanged
    13…Nxe3 14 dxe3 = (Canizares-Evans, IECC 2004)
  • c) 9…dxe4 (Fritz’ favorite) 10 Ned5 Kd7 11 Nxe4 +=
• [3] 5…Nc6 (suggested by JEH) 6 Nxg4 Qh4+ 7 Nf2 Bc5 8 d4 Bxd4 9 Qf3 (Johansson: “looks roughly level at the first glance but there’s life in the position. After Qxf4 the queens may be coming off when pawns f7 and h7 may be weaker than the pawn e4 so Black may want to compensate by doubling the c-pawns with Bxc3 which in turn gives away the bishop pair.”)

Summary: Like the Rosentreter above, the Quaade is a move-order attempt to steer play into the modern Fischer Defense lines (3…d6 4 d4 g5 5 Nc3), but it also allows 4…g4. Unlike the Rosentreter, this line is not a knight sac as best is 5 Ne5!

Line [1] (5…Qh4+ with 7…g2+?) is clearly good for White; Line [2] (5…Qh4+ with 7…Qxg4) is also holding up well and should promise White at least an edge (even materialistic Fritz favors White in almost every line); Line [3] (5…Nc6!?) is probably Black’s best try, with rough equality but an interesting position (I’d prefer White).

The problem with the Quaade is not after 4…g4; rather 4…Nc6! transposing to either a Pierce Gambit (5 Bc4 g4 6 O-O) or Hamppe-Allgaier (5 h4 g4 6 Ng5!?).

References: Most analysis is homegrown here at ChessPub with Craig Evans responsible for the key idea (9 Ne3!). There’s a Quaade thread but most information is on a declined position also reached through the Fischer Defense – Furhoff Attack. Most info is hidden in the big threads (Death of KG, New Look at KG).

So last summer, I thought it'd be great to consolidate and summarize a lot of the homegrown KG analysis that can be found around these parts - ChessPub truly is a great resource for the KG even if no fully acceptable paths have been found (yet) after 3 Nf3.

After a little work, I got very ambitious and even suggested creating a playable e-book to hold all the analysis. I was going to try and merge analysis from just about every source I could get my hands on. Needless to say, that never ended up happening. BUT I do still have some of the material on the:

• Rosentreter (3 Nf3 g5 4 d4 g4 5 Bxf4!?)
• Quaade (3 Nf3 g5 4 Nc3 g4 5 Ne5!?)
• Furhoff Attack (3 Nf3 d6 4 d4 g5 5 Nc3 Bg7 g3) - a line in the Fischer Defense which I named after its strongest proponent
• Strongest lines for most of White's "misc" gambits after 3...g5 (excluding the Hanstein/Philador)
• A rough guide to forming a repertoire that involves any of the ...g5 lines
• A ...g5 systems transposition table
• General notes on naming convention, strategy, etc

While organizing some of my chess-related files, I found it this afternoon and figured I'd throw it up. There's nothing here that can't be found elsewhere on ChessPub, but it's a fairly nice compilation, if I do say so myself.

Long live the KG! (But it will have to be resuscitated by a stronger player than myself...)
David