<The statistics the opening explorer gives doesn't support this statement (GM Fernandez gets easy equality with 6...dxc4) either. After 6...dxc4, white wins 33% and black 23%. Hardly equality I'd say.> You conveniently left out 44% draws (about as equal as you can get). Also you need to check historical results against current results. For the year 2003, for example, the explorer shows 7 black wins, 3 white wins and 16 draws. It is also the overwhelming choice as a response to 6. Bd3 in the opening explorer.

My sense is that the top players consider it good for a draw as evidenced by the short GM draws agreed to in Mamedyarov vs Zvjaginsev, 2003 and Radjabov vs Lautier, 2003

Ludicrous response, no offense. I really don't understand what you're saying. I might as well say that the Philidor equalizes (no it doesn't) in game X if black is a GM and white is an IM. Crazy. Black may win or draw, but that would only be because he is a better player, not because of anything from 6...dxc4 necessarily.

<You conveniently left out 44% draws (about as equal as you can get). >

That indicates that the defense is SOLID (sorry if it reads as though I'm shouting, because I'm not. Just doing this for emphasis.) Not equal. Black win % ~ White win given everything else is similar (ratings on both sides should be ~, which in a popular opening like the Semi-Slav seems feasible.)

<Also you need to check historical results against current results. For the year 2003, for example, the explorer shows 7 black wins, 3 white wins and 16 draws. It is also the overwhelming choice as a response to 6. Bd3 in the opening explorer.>

Actually, you just strenghtened my point more. You (1) made the statistical error of looking at a very small group. (2) You "conveniently" didn't look at the games from 2002. They reveal:

27 draws, 15 wins, 10 losses for white.

This amounts to:
Draws: 52%
White wins: 29%
Black wins: 19%

Many of the games are between GMs, which explains the above average number of draws. For better statistics, I would recommend you actually extend my search all the way to 1999.

< 6...dxc4 seems like a safe drawing weapon (good for easy equality). >

The inductive logic used in Chess analysis to determine "equality" is usually based on elements: (1) the current assessment of "theory" and (2) the results of games played by the best players.

In this case, I will concede that "theory" says the line leads to a "slight advantage for white" as assessed by Kasparov and Keene in BCO and other sources. Yet, interestingly in the two games Kasparov faced this line he lost one in Kasparov vs Kramnik, 1996 and drew the other in Kasparov vs Akopian, 1996 (fortunate to draw a piece down with perpetual check). So forgive me if I put more stock in results than "theory" here. My preference is to recommend what the top players are playing and not what they give as "theoretically best lines."

So, to try and clarify my position at this point, about the move 6...dxc4 in this line, my opinion is as follows:

(1) After 6. Bd3, theory holds that 6...dxc4 is the best move in the position, but that it leads to a "slight white advantage" with "best play." I agree it is the best move after 6. Bd3 but I believe 6...dxc4 equalizes with best play (based on the following).

(2) In the majority of master games after 6. Bd3 (position in this game), 6...dxc4 is by far most the popular response.

(3) In all 78 games in the ChessGames.Com database for the past two years the results show Black has good drawing chances and even winning chances.

If we average the results for Calendar Years 2002 and 2003 that gives of all 78 games in the database for the past 2 years with the following results:

43 draws: 55%

18 White wins: 23%

17 Black wins: 22%

About as even as it gets.

(4) My opinion is that "theory" is probably wrong about "white having a slight advantage" in this line, especially when the theoriticians can't beat it over the board.

(5) With players like Shirov and Kramnik playing it and getting good results (draws and wins) against top competition, it's probably a pretty sound move with good equalizing chances.

a. Kramnik has 5 wins 0 losses and 18 draws out of 23 games played with this line in the ChessGames.Com database.

b. Shirov has 7 wins, 3 losses (all prior to 1995) and 13 draws out of 23 games played with this line in the ChessGames.Com database.

c. Dreev has 16 wins 6 losses (none in the past two years) and 22 draws out of 44 games played in this line in the database.

Speaking of statistics, I've seen a lot of examples of higher rated players as black in the Semi-Slav... I don't think the sample in opening explorer's 2003 is very accurate. It's not very big either. I'm sure if we expand the sample size to more than just 2003 and 2002, to maybe 1999 or even further, we'll see a more accurate sample. You can see the significant difference between the one from 2003, an incomplete year with few games, and 2002.

Anyway, I have nothing wrong though of course with your opinion that black equalizes and that theory was wrong, I was just flabbergasted that you said it so bluntly the first time, as if it was obvious and undebatable.

With that said, I'd like to point out that for (5), you're listing the best practitioners of the Semi-Slav. Do they necessarily win because of 6...dxc4 or because of their play skill? The debate goes on! ;-)

<Does that mean I would play it as Black. Probably not, because I personally prefer more double-edged openings in response to 1. d4 such as the Benko Gambit, the King's Gambit or the Tarrasch. >

By the way, did a little research on Kasparov's approach to the QGA slav variation to examine his approach. He has played the variation with 6. cxd4 twice, resulting in quick draws in Karpov vs Kasparov, 1984 and J Nilssen vs Kasparov, 2001 (I wonder if Kaparov thinks the line is too drawish for his taste?).

Interestingly, Gary Kasparov still plays the QGA Slav, but seems to prefer varying early with such moves for example as <4... dxc4> as in Vallejo Pons vs Kasparov, 2003 or with <4...a4> as in Azmaiparashvili vs Kasparov, 2003 or with <5...a6> as in the well publicized deafeat in Huzman vs Kasparov, 2003 (Gary's instantly famous 21...Bc8?? Blunder spoiling an otherwise comfortable Black position). I mention those not to embarass Kasparov over this defeat, but because these moves are legitimate alternatives for Black to consider in this game's opening line (Gary Kasparov plays solid in the openings!).