Travail Pursuit #58: Disturbing the peace

This image is used under Creative Commons terms and sourced from Remo Cassella's Flickr photostream

This image is used under Creative Commons terms and sourced from Remo Cassella’s Flickr photostream

Today I bring you part two of my four part series on the 9.Nf4 variation of the French Tarrasch, Closed Variation. In the first part published in June I looked at early deviations prior to 9.Nf4 and (if only briefly) the move that can be considered White’s main try against Black’s set up – 9.exf6. In this second post I’m going to start focusing on what can happen after 9.Nf4 but before 18.Bg5 which I consider to be a critical tabiya for the variation.

When I first started looking at this variation in 2007, it occurred to me that the passage of play between these two critical staging posts in the variation was so very forcing that it didn’t seem like there were an awful lot of ways to deviate from received wisdom at least until White got to move 14 where it appeared that 14.0-0 was a potential (if not optimal) continuation.

Eight years later and after a good deal of practical experience I can say that this is not the case at all. Black does have a number of ways to ‘disturb the peace’ before the game arrives at 18.Bg5. Granted some of these are objectively better than others but if White is not at the very least psychologically prepared for such possibilities then these ideas can inflict serious damage.

Black has played 9...fxe5!?

Black has played 9…fxe5!?

Surprise weapons for Black

Generally speaking I have found that most Black players bang out the continuation 9…Nxd4 10.Qh5+ Ke7 11.exf6+ Nxf6 12.Nxg6+ hxg6 13.Qxh8 Kf7 at high speed and almost by rote. I suspect that in some tome or other a great sage of opening play his decreed that the whole line with 9.Nf4 is dubious for White and that executing this sequence of moves is sufficient to give Black a very good game. It certainly isn’t as simple as that, as we’ll see in subsequent posts. However, it is also fair to say that whilst White is pretty restricted in his options between move 9 and move 18, Black does have some very interesting alternatives that he can try.

9… fxe5 was a very rude surprise indeed in the game Shapland vs Clegg, Calderdale League, 2012 (Game 1 in the viewer at the end of this post). Having already played this opponent twice in the Closed System the last thing I expected was for him to deviate so early in this game. The surprise paid off because, although I realised that continuing blindly with 10.Qh5+ and 11.Ng6+ etc would not work. I tried 10.Qh5+ and 11.Qg5+ instead but on this occasion Black bravely brought his king forward with 11…Kd6! Normally, this would spell disaster for Black but here my opponent discerned that an exception could be made and he won the game in fine style. I suspect that 10.Nxe6 and 11.0-0 is a better way for White to play but the position remains interesting.

11…gxf6 was also successful in the online thematic game Inter_Mezzo vs. frank0288, Thematic Tournament, 2013 (Game 2 in the viewer). In this game the opening certainly didn’t go so well for Black and I think I achieved a very good position almost immediately after Black deviated. However, the position remains complex and there’s plenty of scope for both sides to go wrong as we’ve seen several times already.

White has played 14.0-0

White has played 14.0-0

White plays 14.0-0 instead of 14.Qh4

As I mentioned above, when I first studied this line I took a fairly dim view of this move. I admit that this was predominantly because my own bible on the variation (Lev Psahkis’ book on the Tarrasch) made it clear that the alternative 14.Qh4 e5 15.Nf3!? was the move that had ‘breathed new life’ back into the whole Nf4 complex. That surely meant that 14.Qh4 had to be better didn’t it? Well, as always with opening theory, opinion shifts and judgements change. I now think that 14.0-0 is a perfectly acceptable try for White especially if he wants to take a less risky approach to the variation. The down side is that the move is less disruptive and possibly causes Black less problems. So, whilst White is less likely to lose, in my opinion, he is also less likely to win. I think that the four games given below support this view although of course, they do not represent by any means an exhaustive survey.

18.Qh3 was played in Sugden vs. Shapland, Calderdale Individual Championship, 2011 (Game 3 in the viewer). In this game we see Black aim to blunt White’s potential queen’s side pressure by playing 19…Bc5 and then 23…Bb6. The idea is to secure the vulnerable a-pawn thus freeing the Black rook for active duty. What is more the bishop also guards against any invasion by White on the c7 square and the bishop is completely invulnerable if White choses to exchange his bishop for the knight on f6. This seems like a critical defensive resource for Black in the 14.0-0 lines. White may have had a very brief window to cause some discomfort for Black later in this game but then, having missed his chance, he was on the back foot and the rest was brief and brutal.

18.Qh4 was my own choice in an online blitz game Inter_Mezzo vs. nosykes, Blitz Game, 2015 (Game 4 in the viewer). This move order is probably imprecise and White should probably first play 18.Bg5 although that does allow Bc5 in response. In this game we reached a similar position to the game below and we see why Black can’t go for an endgame with 19…Qh8? which allows White to win a pawn. This being a blitz game, I overlooked the opportunity and after a few more errors by both sides we then see an example of what can happen when White is allowed to invade with a rook on the c-file. In this instance the Black king also got lured forwards to his doom.

18. Bg5 seems to be White’s main choice as played in Webb vs. McDonald, London Chess Classic, 2014 (Game 5 in the viewer). Black responded with 18…Be7 in this game and White appeared to get the better of the position before offering an early draw. In this game we see White using his doubled b-pawns to his advantage by putting his rooks on the a and c-files and advancing his doubled b-pawn in order to give room for the queen to swing across to b3 or a3 in some lines. Had play continued there was also an idea of advancing the pawn to b5 in order to plant a White rook on c6 should Black advance his own b-pawn as he did in this game.

18… Bc5 was Black’s response in Winter vs. Potter, Sheffield Congress (Major), 2015 (Game 6 in the viewer). I must here thank Kevin Winter for sending me his game for this collection. 18…Bc5 seems more critical than 18…Be7 and again here Black has the defensive plan noted in the Sugden vs. Shapland game above. Possibly White missed one or two opportunities to initiate a rook lift of his queen’s rook to the king’s side but both players seemed reluctant to risk playing for the full point in this game.

Black has played 15...Bb4+

Black has played 15…Bb4+

Black plays 15…Bb4+

If White opts to play 14.Qh4 (intending 14…e5 15.Nf3!?) instead of 14.0-0 then he has to reckon on facing this move which, in my opinion, is critical to the whole 9.Nf4 complex and may well be the reason why more White players are reverting to 0-0. White probably has to settle for less than he might otherwise be hoping for from this variation. Certainly my only two experiences of playing against this line have been pretty chastening even if one of these games resulted in one of my very best wins. The move initiates a brief forcing sequence because, for tactical reasons, White can’t play 16.Bd2. Instead play continues with 16.Kf1 e4 17.Nxd4 exd3 and now White has to play accurately to keep his head above water.

Shapland vs. Bagley, Leeds League, 2011 (Game 7 in the viewer) was the first time I faced this line and I responded inaccurately with 18.Nb3? which landed me in very hot water indeed! This game is tremendously interesting and my opponent played it really well for a number of moves in which the position was extraordinarily complicated. Finally though he made a single mistake and I was able to initiate a devastating king hunt. It seemed like no justice for his excellent opening and middle game play but in this variation such changes of fortune are inherently part of the landscape.

Shapland vs. Clegg, Calderdale League, 2013 (Game 8 in the viewer) saw me getting it right on the second occasion playing 18.Be3. However, I subsequently went astray by playing too ambitiously and once again got into a losing position. On this occasion, despite having me at his mercy, Black suddenly began to play cautiously and I was able to wriggle free and was even a little better when we agreed to a draw.

I should like to mention at this point that I think White has another very reasonable option after the 16…Bb4+ sequence which is to play 18.Bg5. That is probably the move I would resort to if I were to face this variation again.

Time to try and summarise some general points to remember from these games:

  • 1. The sequence of play between 9.Nf4!? and move 14 appears to be pretty forcing with no room for either side to deviate without getting into a terrible mess. Objectively, that may be the case but practically, over the board, Black does have some alternative options that he can try to use to unsettle a well prepared opponent. White probably ought to have an idea how to handle both 9…fxe5 and also 11…gxf6 should they appear as they are certainly dangerous if you get caught unawares. There are better lines for Black though so unless 9.Nf4 has caught you completely by surprise I think I’d be inclined to avoid these variations.
    2. 14.0-0 is a less risky way for White to play this variation (as opposed to 14.Qh4) whilst still retaining some chances to put Black under pressure. Play is characterised by the opening of the a-file after White initiates an exchange of knights on b3. White can then put pressure on the Black queen’s side down both the a and c-files and by advancing b4 can either aim to swing his queen across to a3 or b3 or try a rook lift along the third rank to join the queen in a king’s side attack. A key defensive idea for Black here seems to be to play Bc5-Bb6 when it will be very hard for White to make progress on the queen’s side.
    3. After 14.Qh4 e5 15.Nf3!? Black can avoid the deepest levels of theoretical discussion by playing 15…Nxf3+ 16.gxf3 Bb4+. I suspect that this move might be the reason why more recent games in this variation have seen a higher frequency of 14.0-0 played. 16…Bb4+ is quite concrete and requires White to play accurately. However, with correct play it looks to me like White should be able to keep the position unbalanced yet equal, dynamic and interesting which is all you should really be hoping for if you play this line with White. It looks to me like White should be aiming for the sequence ending with 18.Bg5 which appears to maintain the tension and the possibilities.
    4. Some of the positional ideas that are critical to the deepest lines of this variation start to emerge in these games. White is trying always to find a point of entry into the Black position for his heavy pieces. This can be on the c-file, or sometimes via the h-file, a-file or even the g-file. Black of course is trying to keep these avenues closely guarded and White has to try to stretch Black’s defences to create a weakness. Black must also be careful about getting his king lured forwards. Almost always this leads to disaster.

As we can see from all of these games, whilst being a huge amount of fun to play, the position after 9.Nf4 is very concrete and complicated and places enormous pressures on both players. As always, knowing some of the main ideas is more important than memorising the moves by heart but there are some forcing sequences from which it is difficult to deviate and maintain realistic hopes of success.

Part three of this series will be the shortest as we reach a critical junction in the variation after 14.Qh4 where 14…e5 15.Nf3!? Nxf3+ 16.gxf3 Bf5 17.Bxf5 gxf5 18.Bg5 are played. Black has a really diverse number of options and we’ll look at a number of those saving the most frequently played (and best?) option for part four.


12 Responses to “Travail Pursuit #58: Disturbing the peace”

  1. Martin Carpenter

    Jul 29. 2015

    Very thorough! I’m seriously impressed you’re getting this many games this ‘deep’ into theory when there’s so many different ways to play after 3 Nd2.

    Oh, Black kings on d6 sheltering behind big pawn centers are very normal for the French defence :)

    Reply to this comment
  2. Dave

    Jul 29. 2015

    Hi Martin,

    It’s always been a surprise to me that I don’t get more games with 3…c5 over the board. I’ve very few compared to the closed system. Do players of the Black pieces (at club level at least) feel less comfortable playing these ‘open’ lines I wonder? Are the ‘closed’ lines more in keeping with the general style of play in other variations of the French? Maybe.

    Certainly although there are plenty of GM-level games played with 3…Nf6, at the very highest levels of the game I think you see 3…c5 played almost exclusively.

    And in terms of getting regularly into these deep lines of theory… I think this is one of those variations that, once Black has played 3…Nf6, there aren’t that many good opportunities to deviate before White has the chance to play 9.Nf4. To a certain extent, any player of the White pieces who is giving his opponent a rare opportunity to wallow into his pet variation is likely to be indulged to the fullest possible extent!

    Part three will hopefully not take as long as part 2!

    Reply to this comment
  3. Matt Webb

    Jul 29. 2015

    Great to see one of my games featured, particularly reminiscing our text exchange that morning as I worked my way through some very interesting lines!

    In case anyone followed the game annotations after the analysis 21… Bxb4? 22. Bxf6! you might be wondering about 22 … gxf6 when the aesthetically pleasing Rc7!! wins on the spot. It’s lovely little tricks like this which always appeal to me when doing any form of preparation.

    Reply to this comment
    • Dave

      Jul 29. 2015

      Thanks for mentioning that Matty! I thought that one was in my annotations but it must have dropped out at some point. I’ve just updated the game file to reflect this as it is a stunning tactic.

      Reply to this comment
  4. James Carpenter

    Jul 30. 2015

    Well, MPT plays 3…c5. I think it’s perceived as a bit of a drawing weapon, although at club level certainly you can easily just equalise and then outplay the opponent, it isn’t as if simplification is inevitable. Many notable chaos merchants have actually played similarly “quiet” openings.

    I think you’re a bit quick to dismiss Ng6+ in the Clegg game Dave, Bxe4 is at least unclear and with that king stuck in the center I’d fancy my chances :).

    Seriously impressive piece of work, as Martin says.

    Reply to this comment
    • Dave

      Jul 30. 2015

      Thanks James. I agree that sometimes 3…c5 is perceived as a drawing weapon and also that this is not really a realistic point of view when playing it at club level. There is actually plenty of wildness available to the players if they both want it though if when White captures on d5 Black recaptures with the e-pawn rather than the queen the game can become very quiet.

      I guess the main reason for the assessment of 11.Ng6+ in the game against Robert is that there are better ideas available and White ends up a pawn down compared to where he’d be in the lines where Black plays 9… Nxd4. That doesn’t feel like a brilliant deal to me even if the position after 11.Ng6+ may objectively be ok.

      Overall in this series I am trying where I can to say which options I think are currently ‘best’ for both sides. Of course the reality is that the position is so complex either player can get away with making sub-optimal moves if their opponent makes the last mistake. Take the game against Andy Bagley. He played that wonderfully for some time after I’d made some poor choices in the opening but it still didn’t take much of a slip from him to undo all his good work and get mated!

      Reply to this comment
  5. Martin Carpenter

    Jul 30. 2015

    I didn’t even mean 3.. c5 specifically. There’s just so *many* lines for black after 3 Nd2 :)

    My database (from TWICS) has 9k for Nf6, 6k for c5 (which is really two very different lines of course), 3k for Be7 (!), 3k for dxe4 (booo), 2k for Nc6, 1k for a6 and even nearly 250 for h6.

    Lest 3.. h6 be thought truly absurd, there’s an average black Elo of 2350 there….

    Goodness Nepomniachtchi seems to have played 3.. a5 3 times – 3/3 vs Sjugirov, Vachier Lagrave and Dominguez. Last two blitz games but still!

    To some degree this is symptomatic of the depth of the theory in 3..Nf6 – this line isn’t the worst of it!. Loads of fascinating lines but mostly just drawn with computer prep. It was already going that way 5-10 years ago. Inspired them to dream up stuff like Be7.

    Oh, the latest news from Berg is the 18.. Bc5 variation from that Webb- McDonald then 19.. g6 rather than the ‘crazy looking’ 19.. Kg6. Then a couple of pages worth of dense analysis.
    (The whole 3.. Nf6 repetoire is 250 pages!).

    Reply to this comment
    • Dave

      Aug 03. 2015

      There are a lot of sidelines Martin.
      It’s interesting that there aren’t that many black players at club level who go in for them. I’ve faced 3…dxe4 (booo!) on a good handful of occasions but not the others.

      At the top level there is a real trend to towards essaying playable sidelines and all the ones you’ve mentioned are playable. But at club level the French Defence seems to have a pretty hardcore following who are happy to play the mainlines in the hopes that they’ll know them better than their opponents (not an entirely unreasonable hope.)

      Reply to this comment
  6. Kevin winter

    Jul 31. 2015

    A good summary and worth playing over the games

    Reply to this comment
  7. Nick Sykes

    Aug 30. 2015

    Oh, I’ve only just noticed that one of our 5 minute Blitz games are on.

    It must be said that I have never played this variation OTB and you have much OTB time plus lots of online experience too.

    Hardly surprising I got smashed.

    Indeed 18.Qh4 is not the best after 18…Qd7 then 19.Bg5 can be answered by 19…Ne4 when I think Black is doing well!

    Reply to this comment


  1. Travail Pursuit #60: On the precipice - Yorkshire Chess | Yorkshire Chess - October 8, 2015

    […] I’ve played fewer games with this theme than I have the others. In the last post of this series (Travail Pursuit #58) I looked an important significant alternative for White in the form of 14.0-0 and then at ways for […]

  2. Travail Pursuit #61: Into the deep - Yorkshire Chess | Yorkshire Chess - November 19, 2015

    […] Disturbing the peace: in part 2 we examined some interesting deviations that can occur between 9.Nf4 and 14.Qh4 or 14.0-0. The theory books tend to suggest that the sequence following 9.Nf4 is forced, but I don’t think that is the case. […]

Leave a Reply