Otdfbt

[FEN "r1bqkbnr/1ppp1ppp/p1B5/4p3/4P3/5N2/PPPP1PPP/RNBQK2R b - - 0 1"]

In the Ruy Lopez Exchange, black always takes back the bishop with 4...dxc6. In Timothy Taylor's 2011 book Slay the Spanish, he advocates 4...bxc6 instead.

I want to analyze his lines and reach my own opinion, but I've never played these lines before. I want to know more about the arguments for 4...dxc6 that he is arguing against first, so that I can be a bit more skeptical about the lines he gives.

White still can't win a pawn with 4...bxc6 5.Nxe5 because of 5...Qe7 6.d4 d6 7.Nf3 Qxe4+.

I can see that 4...dxc6 does more for development than 4...bxc6 does, and it leads to two pawn islands instead of three, but 4...bxc6 captures towards the center.

Is there more to it than that? What is concrete analysis that's usually used to argue that 4...dxc6 is so superior that it is virtually always played?

Good question. Indeed 4..bxc6 is extremely rare compared to the main line, and the reasons are multi-fold specially strategically, but foremost, from a concrete point of view it leads to a rather poor position for black after the immediate d4-exd4-Qxd4 and white's completely controlling the centre:

 [title "Important concrete reply"]
 [fen "r1bqkbnr/2pp1ppp/p1p5/4p3/4P3/5N2/PPPP1PPP/RNBQK2R w KQkq - 0 5"]

 1.d4 exd4 2.Qxd4

From a strategic point of view, 4..bxc6:

• leaves black with an isolated a6 pawn without any effort required from white.




• More damaging to the structure than dxc6 as black's left without any prospect of a queenside expansion since move four already! First repercussion of the latter is the fact that white has the assured prospect of blockading black's isolated a pawn and in doing so render it into a permanent weakness.




• Despite black having the bishop pair in principle, the bishops have very poor prospect in this structure, in particular the light squared one, e.g.: developping it along the long diagonal is no longer meaningful as the fianchetto option is gone and an indirect one with c6-c5-Bb7 will in turn block-in the f8 bishop. Moreover, to develop the bishop along the side diagonal c8-h3 is now at least delayed for an extra tempo, which means white could easily include h3 without any real drawbacks. Instead, with dxc6 both bishops are immediately activated and the bishop pair compensation is more tangible (specially with Bg4 being readily playable).




• The latter is really pointing at the fact that with bxc6 black is left with very poor piece activity, and the structure guarantees no real outpost for the knight, in fact not even the immediate Nf6 is playable after bxc6 d4 exd4 Qxd4 Nf6 the knight is pushed around to non-ideal posts such as b6 or e7 after e5-Nd5-c4.




• Unlike bxc6, dxc6 is in fact challenging the immediate d4 (which is one of white's main plans in this structure) as the queens trade becomes forced and having the bishop pair black definitely doesn't mind the endgame. Thus, as mentioned in the beginning, with 4...bxc6 black is even playing into white's hands by inviting d4.




• This line of bxc6-d4-exd4-Qxd4 is really critical as it perfectly highlights the major differences between the two sides: white has all the space, perfectly healthy pawn structure, ready to castle (while black is at least 2 tempi away), while black has no prospect to claim the centre, nor to liquidate the isolated a6 pawn. I highly encourage you to play through some of the ensuing lines after this move order, you'll quickly develop a feel for the kinds of struggles black is facing.

Now consider going through all above points but with the position after 4...dxc6:

 [title "Main line after dxc6"]
 [fen "r1bqkbnr/1pp2ppp/p1p5/4p3/4P3/5N2/PPPP1PPP/RNBQK2R w KQkq - 0 5"]

in doing so, hopefully you'll be able to convince yourself that all the aspects of dxc6 outweigh that of bxc6, both concretely and strategically. Last but not least, note that here the immediate c3 (for d4) is not possible due to Qd3, moreover, black is well in time to keep the fight over d4 alive (e.g. 5.O-O Bg4 pinning white's most important piece of they're to ever play d4).

At the high level, the key downsides to be emphasized for 4...bxc6 are: isolated a pawn, no prospect of expansion on the queenside, extremely limited minor piece activity, delayed development, and the lack of any ability to create a target in white's camp.

First of all, you are correct in your assessment, that taking towards the center normally is preferred. The reason for this is, that taking towards the center usually (!) allows for easier development and generally more active play. As a rule of thumb (remember - few absolutes in chess!) taking towards the center leads to sharper positions, taking away from it tends to favor calm games.

The Ruy Lopez often results in extremely rich middlegames, which for many players is poses some difficulties, as (especially with white) you need to find a plan in a position where you mostly won't get away with hurling all your pawns at the enemies king and checkmating. Mindlessly developing your pieces is easy - coordinating them on a full board is not.
I personally do not like the positions resulting from 4... bxc as it, while maybe inherently not being bad, gives white concrete targets to attack. Moreover the benefits for black (development, supporting d5) are a while away. Let's take a look at it.

1. Center
   
   Since you now have a pawn on c6, d5 is basically screaming to be played - attacking white's e-pawn. However after a rather quite d3 by white (the bishop is already out of the garage anyway) d5 doesn't really achieve a whole lot. If white were to take with 6. exd, you would have solved all your problems. However white is in no way obligated to do so (and shouldn't!). Taking on e4 yourself results in just a joke of a pawn structure.
   
   Even if the tension is preserved black's pawn on e5 is in trouble. Nxe5 poses a realistic threat now - and there is no good way of defending that e-pawn (much like when white plays c3 to early in a Ruy Lopez!). Bd6 would kill the bishop, Qe7 is awful in so many ways and f3... Well, never play f3!

(Position after 4 ... bxc 5. 0-0 d5 6. d3 dxe? 7. dxe | Not the best moves for white - but the resulting position is not looking good for black.)

Which brings us to...

2. ... the doubled c-pawns
   Which may come under fire. The c6 pawn is only defended by the d-pawn - and you are going to move that one away.




3. Development
   Black opened b7 for his bishop, but the diagonal is now blocked by his c6 pawn. Developing to b7 anyway doesn't feel good.
   Another idea would be to push the a6-pawn and put the bishop on a6 - this however weakens the pawn severely and the bishop is somewhat misplaced on a6.
   Black doesn't really gain anything here.

I sadly do not have more time for more analysis, but white can find easy ways to put blacks position under pressure and will probably be in the driver seat for the rest of the game. Black is required to play really precise to hold his position.

On a side note: 4. ... bxc 4. O-O has a 62% (!) win rate for white in the lichess masters database! The only commonly known grandmaster to play this is (Who else...? ^^) Bent Larsen. The game ended in a draw.