See http://www.playchess.de/anno-java.c...

The problem with the Marshall, as I see it, is two-fold. First, there are the anti-Marshalls--and not just on move 8. White can anti on moves 7, 6, 5, and 4 as well with perfectly sound lines. It's just too hard to get into a Marshall.

Secondly, the burden of learning the analysis is all on Black. White does need to know one adequate defense, but he only has to learn that one. Black, on the other hand, must study and prepare for them all, since there are fewer sidelines for Black to specialize in.

As for some of the questions raised above: after 8.h3:

click for larger view

My experience is that 8...Bb7 is good for Black. You'd be amazed the number of opponents who will go into a particularly vicious type of Marshall with 9.c3 d5 10.exd5 Nxd5 11.Nxe5 Nxe5 12.Rxe5 Nf4!

As for the Qf3 line, I think what was meant was 15.Qf3 in this position:

click for larger view

Two lesser-played possibilities for White here are 15.Qd3 and 15.Qf3. It's been years, since I studied this carefully, but I think the gist is that Black should <not> attack the queen with his bishop; that is, play 15.Qd3 Bg4 or 15.Qf3 Bf5. Please don't ask me why, but that's what I remember.

There is another nasty trap in the latter line: 15.Qf3 Bf5 16.Bxd5 cxd5 17.QXd5 Rae8! 18.Be3?? Be4! I've caught a few with that one too, and it also works after 18.Rxe8 Rxe8.

I have to admit that these days I've been taking up 3...f5. There's not quite as many anti-Schliemann lines around.

click for larger view

And now <12...Bd6 13.Bxd5 Bxe5 14.Bxc6>.

Well, one could be particularly nasty and ask "Why settle for material equality (two pawns for the exchange) when you could be a pawn ahead?" But that's obviously not the point here. Black would be out of his comfort zone, and as long as the bishop can be maintained on c6 it will be difficult to play ...Re8 for e-file pressure. Also, I don't know if Black would especially want to trade light-squared bishops, as White's defense often involves the g2-g3 push when Black needs that bishop for attacking purposes.

Perhaps 12...Bd6 isn't Black's best response in the diagrammed position. 12...Be6 is worth a thought, but Black might end up a tempo down if White heads back to more normal Marshall lines. Or maybe Black would like to gambit another pawn after 12...Bd6 13.Bxd5 cxd5. A couple of fun possibilities are 14.Rxd5 Bb7 15.Rxd6 Qxd6 16.Qxb7 Rae8, or 14.Qxd5 Bxe5 15.Qxd8 (15.Qxe5 Re8) Bxh2+. Neither of these are forced, of course.

In short I don't see anything forcing for Black offhand, though our Friendly Neighborhood Mechanical Monsters may rule differently. So why isn't this idea played? From what I can see, it goes against a couple of time-tested principles of the Marshall.

First, White should beware of a premature Bxd5. Black's attack often involves the move ...f5, which self-pins the knight on d5 and greatly increases the bishop's power. Second, White's basic problem in the Marshall is the slow development of his queenside pieces. That's why 12.d4 is almost invariably played instead; White needs to get his pieces developed rather than leading with his queen and possibly exposing her to attack.

I've always thought that you should try your new ideas in established openings, as long as there is no immediate refutation. If the idea is good, all the better. If it's lousy, well, that should help you to better evaluate your ideas the next time one jumps into your head.

<Or maybe Black would like to gambit another pawn after 12...Bd6 13.Bxd5 cxd5. A couple of fun possibilities are 14.Rxd5 Bb7 15.Rxd6 Qxd6 16.Qxb7 Rae8, or 14.Qxd5 Bxe5 15.Qxd8 (15.Qxe5 Re8) Bxh2+. Neither of these are forced, of course.> On one hand, the type of player that plays the Marshall is the type that would trot out that kind of gambit faster than the next guy. But I have to say that I really don't like black's position after that. Similarly, I don't think playing 12...Be6 first is too threatening. White can just play 13.d4 ...Bd6, when I think the insertion of the moves d4 and Be6 favors white.

Also, I disagree that there isn't anything forcing. e.g. <1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.O-O Be7 6.Re1 b5 7.Bb3 O-O 8.c3 d5 9.exd5 Nxd5 10.Nxe5 Nxe5 11.Rxe5 c6 12.Qf3 ...Bd6 13.Bxd5 Bxe5 14.Bxc6>.

click for larger view

In your line, I think I would profer 14...Rb8 to 14...Ra7, which just walks into a pin. But the more I look at that position, the more I don't like it for Black. I don't know if it's losing by any means, but there just doesn't seem to be the type of play Black likes to have.

<1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.O-O Be7 6.Re1 b5 7.Bb3 O-O 8.c3 d5 9.exd5 Nxd5 10.Nxe5 Nxe5 11.Rxe5 c6 12.Qf3 Bd6 13.Bxd5 Bxe5 14.Bxc6 Rb8 15.d4 Bd6 16.Nd2 Qe7 (Qc7 is well met by Ne4!, and after 17...Bxh2 18.Kh1, white threatens 19.g3, and black can't put his bishop anywhere. 19...Bd6? 20.Nxd6 Qxd6 Bf4 ) 17.g3 Qe1+ 18.Kg2>

click for larger view

I'm beginning to think that Black should offer the second pawn with 12...Nf6:

click for larger view

What we have now is the original Marshall attack (11...Nf6) with the moves Qf3 and ...c6 thrown in. First of all, I don't know if White really wants to take that c-pawn. After 13.Qxc6 Rb8, Black is threatening to trap the queen with 14...Bb7, so White must retreat her and has no time for d2-d4. So let's say 14.Qf3 Bd6, reaching this position:

click for larger view

Now, if you try the 15.d4 Bxe5 16.dxe5 idea, Black is obviously better placed than before; the pawn on e5 is more exposed than it was on d4, and the White bishop is not on the commanding c6 post. Problem is, the rook doesn't have a great square to go to. Normally in the Marshall White would play 15.Re1, but that loses at once to 15...Bg4 16.Qe3 Re8!

There are several important things to consider before taking up the Marshall:

1) You won't get to play it a lot. White has many good ways to avoid it on moves 8, 7, 6, 5, 4, and there's always those people who don't play 3.Bb5 in the first place.

2) You have to be familiar with a lot of theory. White has a number of choices that Black must be familiar with, and what works in one line often doesn't work in another. This on top of all the extra theory you have to know for the non-Marshall lines.

3) At the top levels, I believe it is considered drawish. For the average player it is very double-edged and dangerous for both sides.

4) You can't go into it just looking to get the pawn back with a positional edge. Often, additional and heavy sacrifices are needed to keep the initiative.