Lopez Exchange with 5.h3

Fellow chessplayers,

I wonder what your collective opinion is on the move 5.h3 in the Ruy Lopez exchange variation. It's something I've thought about before, and I think it's worth a try.

I haven't done much analysis on it yet. The idea, of course, is to prevent the 5...Bg4 pin that occurs on 5.O-O

Here are some sample lines:

You allowed your opponent to eliminate the second exchange in the variation. It's the main idea, creating a pawn majority King side for White and giving a compromised majority to Black Queen side. That creates a winning pawn end game for White (all else equal).

You also wasted a tempo on a non-existent threat. Even if Black insists on giving up the bishop pair, it's not a problem. However, it was Black's compensation for the crappy pawn structure.

The text moves are better...back to the drawing board?

The problem is that 5.O-O Bg4!? is completely playable for black but certainly not tragic for white. as a consequence 5.h3?! is an overreaction, certainly playable but harmless since in most setups black wasn't going to  play Bg4 anyway. 

There are over 300 games with the 5.h3 variation in the ChessBase Big Database 2011.  It seems to be split with the same number of wins for White and Black.  One short game is this one.

5.h3 looks completely toothless. Black has an easy game by many ways, e.g. 5...c5!? (making d2-d4 difficult to achieve) and then proceeding by ...Bd6, ...Ne7 etc.

5...Ne7 may be even better. Black intends ...Ng6 and normal developing moves, and White wouldn't like to see a knight planted on f4 after he has castled short, because his unprovoked h2-h3 helps Black's play.

Black is simply better after 9...Qf6, your "great attack" is an illusion. A sole Queen does not constitute an attack.

But anyway, Black does not have to play ...f6, since the e5-pawn is not in any danger because of the h2-h3 tempo loss. Therefore 5...c5 or 5...Ne7 immediately makes more sense.

Your 5...Nf6 is certainly playable, but f6 isn't the best square for the knight in that type of position - actually it's Flohr's exchange variation with an extra move for white, h2-h3, which is not useless. White will continue with d3, Be3 etc without hurrying to castle, and now his h2-h3 move is rather handy.

The Flohr exchange is less harmless than its looks. Nakamura won with white against GM Hammer and drew against Magnus Carlsen, GM Papaioannou won GM Grivas... and so it goes.

5...f6 might look bad (it's almost always bad in open positions after all) but in fact is a very very common move in the exchange ruy. You have to consider that:

1) the white LSB is gone

2)after the common 5.O-O white play is all about pressurizing the e pawn.

With this in mind you can easily understand why the ugly looking f6 is usually not bad at all in this variation; it's a great move instead. It supports firmly the e5 pawn, and the weakened white diagonal is of no importance with the bishop already gone. In fact 5.O-O f6 is the absolute mainline of the exchange variation.

Of course in this particular line the move is still ok, but given that 5.h3 is a very anemic move black is justifyed in searching something more active such as Ne7.

...f6 is a solution when the "threat" Nxe5 is real. It's not the only solution, but it's standard.

But, if the e4-pawn lacks protection, then Nf3xe5 is not a real threat (it rather helps Black!), so Black can omit ...f6 and protect e5 in a more useful way. I can imagine ...Ne7-Ng6, ...c5 and ...Bd6, and ...Qd6. I prefer the former, as the knight is ideally posted on g6, and black can develop his pieces in an optimal way.

I really like your Ne7-g6 idea, black has been gifted some extra time and it makes perfect sense to refrain from f6. My point was to explain why f6 is not senseless in this sort of position, not to say that you always have to panick and defend e5 with every avaiable move :)

I guess that the reason which makes you say that eliminating both e pawns helps black is that white is not in possession of an effective pawn majority anymore(because he can't exchange his d pawn for black e pawn and have a kingside majority as in the normal lines)? 

Indeed, if you remove pawn e4 and e5 then white has no kingside pawn majority in the endgame. He just has to combat the two black powerful bishops, which become stronger and stronger, as the position simplifies.

Just for the record: ...Ne7 is serious option in the normal Ruy Exchange. But there Black has to counter 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Bxc6 dc6 5.0-0 Ne7!? 6.Nxe5(!) Qd4 7.Qh5 where black is a pawn down, and must seek positional concessions in a highly complex position.

The current status of theory (there is a fair amount of it) is that Black "may" not have enough for a pawn, but nothing is granted.

After such an innocuous move as 5.h3 though, 5...Ne7 makes perfect sense.

5. ... c5! = (6.Nxe5? not yet available) but having said that, for mainline 5.O-O, I prefer Bronstein's 5. ... Qd6! over 5. ... f6(Gligoric) but ocassionally adopt the sharp 5. ...  Ne7!?(Keres) or 5. ... Bg4. 5. h3?! is just losing a tempo, 6. d4 has to be played as soon as possible to create kingside majority, that is, the entire point of the lopez exchange. Perhaps I could claim that Black has already equalized after    5.h3 and have slight advantages due to possession of bishop pair and caught up in developement with white.

In the regular exchange Ruy, 5...f6 is standard, but after 6.d4 I do not feel 6..ed4 is best (although it has been played a lot). I'd rather go for 6...Bg4, which is a very tough (positionally speaking) position, where Black has good chances to play for the whole point. Marin's book on the open games gives a superb overview of that particular variation.

Thanks for the useful clarification, that is what i suspected but it's nice too have it confirmed by a strong player like you :)