The Chess Mind

After a few days without seeing a response on James West's blog to my last anti-PCG post (or to other points in its predecessors), I stopped checking for a while. A browse on Sunday revealed that he hadn't been idle, however; he had returned to the scene of the crime! Meanwhile, by a happy coincidence, reader Marvin (Joe) Barker wrote in with a nice anti-PCG idea of his own.

For those who haven't followed the fun, here's where we are. The PCG (Philidor Counter-Gambit) starts with the moves 1.e4 e5 2.Nf3 d6 3.d4 f5(?!), and the dispute with West, who has played the PCG with an almost religious devotion for many years, begins after the further moves 4.Nc3 fxe4 5.Nxe4 d5?! (5...Nf6 seems to be stronger) 6.Neg5. We have examined 6...h6, 6...e4, and 6...exd4, but it's only the latter move we're going to look at today. Here I've offered two moves for White: 7.Nxd4 and 7.Bb5+(!?). West has something to say in Black's defense in both cases, and Barker has a contribution of his own here. Let's have a look.

(A) 7.Nxd4. One line I offered here continued 7...Qe7+ 8.Be2 h6 9.Ngf3 c6 10.O-O Qf6 11.Re1 Bd6 12.Ba6+ Kf7 13.Bd3 Ne7 14.Bd2 with the idea of following up with Bc3. White is at least clearly better here, and possibly winning. To avoid this Bd2-c3 idea, West proposes the clever 11...Bb4, and only after 12.c3 to play 12...Bd6. Now the c3 square is occupied, so that after 13.Ba6+ Kf7 14.Bd3 Ne7 White no longer has the Bd2-c3 idea.



Except that he does: 15.c4! Black's position looks better than it did in the 11...Bd6 analysis, but I still think White has a clear advantage here. For example: 15...Re8 16.Bd2 Na6 17.Bc3 Bb4 (neutralizing the c3-bishop, but now White's knight can use the e5 square) 18.Bh7 Bxc3 19.Ne5+ Kf8 20.bxc3 and neither 20...g6 nor 20...c5 reduces White's edge below +/-. (See the link below for more details.)

(B) 7.Bb5+ c6 8.Bd3 Bb4+ and now:



(B1) 9.Bd2 Bxd2+ 10.Qxd2 Nf6 11.O-O-O O-O 12.Nxd4 Re8 13.Rhe1 h6 14.Rxe8+ Qxe8 15.Re1

In West's earlier analysis of this position, he recommended 15...Qh5, and considered Black's position acceptable after 16.Nge6 Bxe6 17.Nxe6 Nd7. In my response, I pointed out that 16.Nge6 was an error and inferior to 16.Ngf3, but I didn't explain why. So let me rectify that omission: the reason Nge6 is bad is that it helps Black catch up in development, and the reason Ngf3 is good is that the knight is ready to leap into the more useful e5 square (and then follow up with Bg6).

This pertains to West's suggested improvement of 15...Qf8. His analysis continues 16.Nge6 Bxe6 17.Nxe6 Qd6 18.g4 Nbd7 19.g5 Re8 20.Bf5 hxg5 21.Qxg5 Re7 22.Rg1 Ne8 23.Nd4 Qf6 24.Qg4 Nf8, with (at least) equality, but after 16.Ngf3! Black is in trouble, whether he allows the knight to e5 or not. Again, see the link below for details.

(B2) 9.c3!? This gambit idea is Barker's, and it looks like a good one. White has a lead in development, so why not increase it? With his aggressively posted knight on g5, the well-placed bishop on d3, the open e-file, and the chance to post the second knight on e5, such a full-speed-ahead strategy deserves consideration.

Details here.

I received an email from a Benjamin Good, who recently played in the West Virginia Active Chess Championship. He doesn't make clear whether he had been a reader of this blog while I was posting on the anti-French line that starts 1.e4 e6 2.c4 d5 3.cxd5 exd5 4.Qb3, but it was clear that one of his opponents was! It turned out well for Good, who had Black in that game, but I think there are a couple of useful improvements for White, too. Have a look here.

Incidentally, I want to make clear that I'm not the inventor of this variation; that questionable honor goes to Stefan Bücker. (Though I am wholly to blame for occasionally playing this silly line over the years.) I can say, however, that all the analysis I've offered, pro and con, does emanate from me (with an occasional assist from my silicon friends). There was a profile of Bücker in New in Chess in the mid-1980s, and it listed a number of his pet lines. This variation was among them, but no further analysis was provided and there weren't any databases around to look for his games. So for better or worse, the analysis I've posted is my own.

The European Championships ended a week ago (won by Tkachiev in a playoff), but it's not too late to enjoy and learn from the contest. Two late-round games caught my eye and are annotated here. Both are sharp Sicilians, and show how razor-thin the margin is in that opening between being okay and losing.

The first game, Mastrovasilis-Timofeev, was a pitched battle in which the players both attacked each other's king, more or less simultaneously. That's not so unusual in itself in the Sicilian, where attacks on opposite sides of the board are quite common. What made this game interesting is that both sides castled long, and still managed to go hacking. Through much of the game, it looked like Black was in danger, but his position was resilient and it was White who finally went astray in the complications.

Game two was Volokitin's especially painful last-round loss to Jakovenko, who was part of the pre-playoff tie for first. If the previous game showed that Black's defensive resources are greater than one might think, this game offered a corollary: Black's margin for error is smaller than one might expect. Volokitin introduced a novelty on move 17, made an error on the very next move, and was probably lost as a result.

Sometimes the Sicilian is forgiving, and sometimes it's not. (Helpful, isn't it?)

Browsing New Jersey master Jim West's blog a couple of days ago, I noticed that he has responded to another of my analyses on one of his pet lines, the Philidor Counter-Gambit (1.e4 e5 2.Nf3 d6 3.d4 f5). The variation du jour continues 4.Nc3 fxe4 5.Nxe4 d5(?) (Stefan Bücker's 5...Nf6 6.Nxf6+ gxf6 looks more resilient, though still somewhat better for White, I think) 6.Neg5! exd4 7.Nxd4 Qe7+ 8.Be2 h6 9.Ngf3



and now we have a parting of the ways. In my earlier post, and in the game he considers in his post, the continuation was

(A) 9...c6 10.O-O Qf6 11.Re1 Bd6 12.Ba6+!

My analysis continued with 12...Kf8 13.Bd3 Ne7 14.c4, with a clear advantage for White, but West gives 12...Kf7 13.Bd3 Ne7 14.Bd2 c5 15.Nb5 Nbc6 16.Bc3 d4 17.Bc4+ Kf8 18.Nxd6 Qxd6



Black "will follow with moves like ...Qf6, ...g5, ...Kg7, and perhaps ...Rf8 as well, with a reasonable position", according to West.

I think his variation is pretty reasonable through Black's 18th move, but his conclusion is unduly optimistic. If Black gets to make those consolidating moves for free, then he'd probably be in good shape. But if we look at the position right now, we should be more impressed by White's bishop pair and Black's stunted development. White should now play 19.b4(!), threatening to rip apart the center for his bishop pair, and also (alternatively) to play b5 and then Ne5. White has a big, maybe winning advantage here, as I try to show in my analysis. (Link below.)

(B) West adds another move to the mix, 8...c5. He refers to a previous blog post, where in the notes to a June 1999 game with Jaan Ehlvest, he gives this variation: c5 10.Nb5 a6 11.Nc3 d4 12.Nd5 Qd6 13.Bc4 Ne7 14.Bf4 Qe6+



15.Qe2 Nxd5 16.Bxd5 Qxe2+ 17.Kxe2 Nd7 18.Rhe1 Nb6 19.Kf1+ Kd8 20.Bf7 Be7 21.Ne5 Rf8 22.Bg3 Bf5 23.Bg6 Bd6 24.Nf7+ Rxf7 25.Bxf7 Bxg3 26.hxg3 Bxc2 27.Rac1 d3 28.Re8+ Kd7 29.Rxa8 d2 30.Ra1 Nxa8 31.Bh5 Nb6 32.Bd1 Bd3+ "with compensation".

As Bent Larsen says (though I try to forget it when it applies to my own analysis): long line, wrong line. Look at the position around move 12-13. White has a significant, almost threatening lead in development, while Black's king is stuck in the center on an open file and vulnerable along the h5-e8 diagonal to boot. Can Black really be okay here?

My first reaction is "no way!" - but can I prove it? My first impressions might be reliable most of the time, but West has had years of practice with his "baby", and has undoubtedly developed a good sense of what Black can and can't get away with in this variation. On this occasion, however, I think his judgment is provably mistaken. Take a look at the previous diagram, which constitutes some heavy-handed foreshadowing. Those of you who experimented with "my" anti-French line won't be surprised by the right move here: 15.Kd2! The threats of 16.Nc7+ and 16.Re1 promise a massacre, so Black's only chance is to grab material with both hands and pray: 15...Nxd5 16.Re1 Nxf4 17.Bxe6 Nxe6 (17...Bxe6 is even worse - see the analysis linked below) 18.Ne5 h5 19.Qf3 and Black is going to be smashed - just what one should expect with an exposed king and a severe lack of development.

The analysis given above - and more - can be replayed here. Back to you, Mr. West!

For the Francophiles who have been terrified by the earlier installments of my anti-French Defense posts (in chronological order: here, here and here), rejoice: I give you the cure.

In fact, twelve cures. As I insisted from day one, the line, while dangerous, is objectively innocuous at best. Finding the antidote in the heat of battle is tough for most of us, especially when faced with something brand-new. So here, in the comfort of your own home, is the answer to your 1.e4 e6 2.c4 problems - and then some.



Response 1 - Head for the Sicilian with 1.e4 e6 2.c4 c5! One possible continuation, transposing to one of GM Rublevsky's pet lines, is 3.Nf3 Nc6 4.Nc3 Nf6 5.Be2 d5 6.exd5 exd5 7.d4 with a position that's playable for both sides.



Response 2 - Head for the English Defense with 1.e4 e6 2.c4 b6!

Response 3: 1.e4 e6 2.c4 c6. After 3.d4 d5, White can choose between 4.e5 (Black should be okay after 4...Ne7 and maybe 4...c5), 4.Nc3 Bb4! (4...dxe4 5.Nxe4 Bb4+ 6.Bd2 Qxd4 7.Bxb4 Qxe4+ is the very sharp Marshall Gambit in the Slav; it's playable for Black, but requires serious preparation), and 4.exd5 exd5 5.Nc3 Nf6.

Response 4: 1.e4 e6 2.c4 e5 - head for a double-king pawn position where the pawn on c4 is strategically double-edged. Interestingly, after the plausible 3.Nf3 Nc6 4.Nc3 g6 (4...Nf6 looks okay, too) 5.d4 exd4 6.Nxd4 Bg7 7.Be3 d6 8.Be2 Nge7 9.O-O O-O 10.Qd2 f5 11.Nxc6 Nxc6 12.exf5 Bxf5, we have transposed into a position from the Modern Defense, albeit one in which it's generally (but not always) Black to move. Even so, Black seems fine here too, but I don't think this is the best choice. (I'd also relatively unhappy about responses 2 and 3. All three options are playable, but the other replies are stronger.)



Response 5: 1.e4 e6 2.c4 d5 3.cxd5 exd5 4.Qb3 Bc5 - collecting on d5 leaves White with two weak d-pawns, and everything else leaves Black to capture or continue developing.

Response 6: 1.e4 e6 2.c4 d5 3.cxd5 exd5 4.Qb3 Bd6!? - Black now threatens to take on e4, as Bc4 Qe7 won't block in the bishop any longer. White should try 5.Qxd5 if the goal is to punish Black, but then 5...Nf6 followed by 6...O-O leaves Black with a big development advantage in return for the pawn.

Response 7: 1.e4 e6 2.c4 d5 3.cxd5 exd5 4.Qb3 Nf6 5.Nc3 Bc5 - this is similar to response 5, but with Black closer to castling it's just so much the better.

Response 8: 1.e4 e6 2.c4 d5 3.cxd4 exd4 4.Qb3 dxe4 5.Bc4 Bc5 laughs at White's threat - yes, you want the f-pawn? Take it! After 6.Bxf7+ Kf8, Black threatens ...Qf6, and if White retreats with 7.Bd5, Qh4 8.Qg3 Qxg3 9.hxg3 Nf6 gives Black an edge. Best is 6.Nc3, when one reasonable possibility is 6...Qh4 7.Bxf7+ Kf8 8.g3 Qe7 9.Bd5 c6 10.Bxe4 Nf6 11.d3 Be6, when Black has sufficient compensation for the pawn.

Response 9: 1.e4 e6 2.c4 d5 3.cxd5 exd5 4.Qb3 dxe4 5.Bc4 Qe7 6.Nc3 c6 7.d3 b5 8.Nxb5 cxb5 9.Bd5 Be6 10.Qxb5+ Bd7 11.Qb7 Qb4+ 12.Kf1 Qxb7 13.Bxb7 Bc6 14.Bxa8 Bxa8 15.Bf4 Nd7 16.dxe4 Bc5. Here the material situation is nominally in White's favor, but here the pawns cannot make their presence felt while the minors will start to coordinate and swarm very soon.



Response 10: 1.e4 e6 2.c4 d5 3.cxd5 exd5 4.Qb3 dxe4 5.Bc4 Qe7 6.Nc3 Nf6 7.d3 exd3+ 8.Kf1 Nbd7 9.Bg5 Nc5 10.Qa3 Be6 11.Nd5 (11.Re1 d2 wins) Bxd5 (11...Qd7 might be even better, but it's needlessly sharp) 12.Bxd5 O-O-O 13.Bf3 Kb8 and White has no attack to compensate for the two pawns or his bottled-up kingside.



Response 11: 1.e4 e6 2.c4 d5 3.cxd5 exd5 4.Qb3 dxe4 5.Bc4 Qe7 6.Nc3 Nf6 7.d3 exd3+ 8.Be3 Nbd7 9.O-O-O Ne5 10.Nf3 Nxc4 11.Qxc4 Qb4 12.Qxd3 Bd6 13.Ng5 Be6 14.Kb1 O-O-O 15.Nxe6 fxe6 leaves White down a pawn and without an attack.



Response 12: 1.e4 e6 2.c4 d5 3.cxd5 exd5 4.Qb3 dxe4 5.Bc4 Qd7 6.Nc3 Nc6 7.Nxe4 Na5 8.Qe3 (8.Qb5 Nxc4 9.Qxc4 Qe6 trades the queens, leaving Black with a better position and the two bishops) Nxc4 9.Nd6+ Kd8 10.Nxc4 Qd5 11.Ne5 Bd6 and Black, ironically enough, has a winning attack likely to be crowned by a rook move to the e-file, exploiting the king's losing the right to castle.



Problem solved!

(The analysis can be replayed here.)

Jim West has blogged a reply to my first anti-PCG post. As one would expect from someone who has spent a lot of time on that line, his analysis helps to move the discussion along. Unfortunately for the gambit's friends, the movement is probably to its grave, but time will tell.

Here are the highlights. After 1.e4 e5 2.Nf3 d6 3.d4 f5?! 4.Nc3 fxe4 5.Nxe4 d5, we've discussed two lines: 6.Ng3 and 6.Neg5!

(a) 6.Ng3. This is what I played against West, and although I was successful, it probably doesn't promise White very much. West, citing Kosten, gives 6...e4 7.Ne5 Nf6 8.f3 (Keres) and now instead of 8...Bd6 (played by West), 8...exf3 9.Qxf3 Bd6 10.Bd3 O-O 11.O-O Nc6 12.Nxc6 bxc6, when Black is fine. I agree.

(b) 6.Neg5! This is the move given by Bauer in his new book The Philidor Files (thanks to Everyman Chess for not naming it The Philidor Philes), and as far as I can tell, White has a clear advantage after this move. Bauer considers 6...e4 and 6...h6, and I threw in 6...exd4 as well, which happens to be West's preference.

(b1) 6...e4 7.Ne5 Nh6, and now the road divides. I presented Bauer's line, which continues 8.Nxh7 and concludes with a clear White advantage. West thinks Black can improve, but suggests 8.Nxe4 as even better. I think he's right about 8.Nxe4 - I'm not sure if it's objectively better, but it is simpler. His analysis of 8.Nxh7 seems flawed though, so both moves look strong.

(b2) 6...h6. As far as I can tell, West at least tacitly agrees that this move holds no hope for Black. Perhaps the following is key: 7.Nf7 Kxf7 8.Nxe5+ Ke7 (Bauer only considers 8...Ke6) 9.Ng6+ Kf6 10.Qf3 Bf5 and now not 11.Nxh8, as mentioned in Shibut's review of West's second PCG book, but 11.g4! as given in my first post.

(b3) 6...exd4. This is West's main line, and he claims that after 7.Nxd4 Qe7+ Black gets a good game. I've argued against this in my second anti-PCG post, which West hasn't yet gotten around to rebutting. He has, however, critiqued my suggestion of 7.Bb5+ c6 8.Bd3, but there too I believe I have an improvement on his analysis.

The details are here.

I've written a number of book reviews for Chess Today the past couple of years (and am continuing to do so), and with their permission, I'll now start publishing them on my website. Here's an early one, covering Khalifman's anti-Pirc and Modern volume. (Originally published in CT-1818.)

New Jersey master James West is a long-time fan of the Philidor Counter-Gambit (1.e4 e5 2.Nf3 d6 3.d4 f5), and in this recent post I wondered what he had in mind against the important line 4.Nc3 fxe4 5.Nxe4 d5 6.Neg5. Happily, West himself has written in to share his answer, or at least where it can be found: on his blog, in this post in particular.

His answer, in short, is 6...exd4 7.Nxd4 Qe7+, which looks like a safer approach than 6...h6 or 6...e4. But is it a panacea? I don't think so. Have a look here, where I recap my earlier analysis of 6...exd4 and add some more. My conclusion is that White is clearly better, and moreover in positions where Black isn't having any fun.

In my blurb for this week's ChessBase show, covering the game Robert Byrne - Yuri Balashov, I asserted that the English Attack against the Najdorf was invented/introduced into tournament play by Byrne long before "Nunn, Short, and Chandler were hammering people to bits with this plan".

To my surprise, John Nunn dropped ChessBase bigwig Frederic Friedel a note, which the latter passed along to me. Most of it quoted my blurb, and then Nunn wrote this:

The name English Attack refers to the line 1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be3 e6 7 f3 (or 7 Qd2), etc; in other words, the idea of meeting ...e6 by f3, Qd2, 0-0-0, g4, etc. Actually, I don't know of a single game in which Byrne played this way (he generally met ...e6 by lines involving Be2).

Very interesting! But is it true? And who decides these things? Here is a pair of quotations that might be of interest:

“The English Attack starts with the moves 1 e4 c5 2 Nf3 d6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 a6 6 Be3 [bolding in original] and is named after a group of then young English grandmasters (Short, Nunn and Chandler) who in the 1980s used the set-up Be3 + f3 against the Najdorf with great success….

“Black can also choose the traditional Najdorf plan 6…e5. This is also the most popular way of meeting the English Attack.”

This suggests that the English Attack isn’t limited to the Scheveningen lines with 6…e6, but refers to White’s set-up against the Najdorf, which can be met by Black in various ways. Of course, my source could be wrong, too, but if so I’m in excellent company: it’s from the introduction to Tapani Sammalvuo’s 2004 The English Attack, typeset by Nunn himself for his own publishing house (Gambit). Likewise, the Gambit work Mastering the Najdorf by Arizmendi and Moreno refer to Be3 + f3 vs. the Najdorf as the English Attack, and give that label the chapter in their book covering 6.Be3 e5. So maybe I'm wrong, but at least I'm in very good company - in company with Nunn's company!

Perhaps it's a case of semantic drift. Maybe "English Attack" originally referred to the anti-6...e6 Scheveningen system alone, but eventually and imperceptibly came to refer to the attacking formation with Be3, f3, Qd2, g4 and O-O-O against Sicilians with a Najdorf beginning, whether they continued with 6...e6 or 6...e5.

Let's assume the account in the previous paragraph is historically accurate. What now? Should we name Byrne's Variation solely after him, and reserve the English Attack for the ...e6 lines alone? That would give credit where credit is due, but will anyone follow suit? If Nunn hasn't corrected his own authors, to say nothing of the many other authors out there who call 6.Be3 e5 the English Attack, is there any hope of linguistic reform?

The Philidor Counter-Gambit, which is defined by the moves 1.e4 e5 2.Nf3 d6 3.d4 f5(?!), has long had a poor theoretical reputation, and I'm unaware of any recent work overturning that verdict. To take the most recent example, GM Christian Bauer's main line disposes of it as follows:

4.Nc3(!) fxe4 5.Nxe4 d5(?!) 6.Neg5! (He thinks 6.Ne5 and 6.Ng3 both offer some advantage, but not as much as 6.Neg5) 6...h6 (6...e4 7.Ne5 Nh6 8.Nxh7! etc.) 7.Nf7! Kxf7 8.Ne5+ ("All White needs now is accuracy in the conversion of his clear advantage.") 8...Ke6 9.Qg4+ Ke7 10.Ng6+ Ke8 11.Qe2+ Ne7 12.Nxh8 ("The h8-knight will escape, despite all of Black's efforts to prevent it.") 12...Bf5 13.g4 Be4 14.f3 Bh7 15.g5 hxg5 16.Bh3 Nbc6 17.c3 Qd6 18.Qe6 Qxe6+ 19.Bxe6 Nd8 20.Bg4 g6 ("Black is at last ready to 'pluck' the knight, but...") 21.h4 Bg7 22.hxg5 Bg8 23.Bf4 c5 24.dxc5 d4 25.cxd4 Bxd4 26.O-O-O Bxc5 27.Rhe1 Kf8 28.Bd6+-. (Christian Bauer, The Philidor Files (Everyman Chess 2006), p. 31.)

New Jersey NM James West has played this variation for a long time, however, and has even written a couple of books defending it. Reading this post on the Kenilworthian blog and following its links makes me wonder if West has addressed the 6.Neg5 line in his writings. Bauer's analysis may be new, but 6.Neg5 isn't - Mega2007 offers 10 games going back to 1979.

Do any West-savvy readers know what he thinks about this move?

(The analysis above, and quite a bit more besides, can be found here.)