I consider that 1. d4 is the best move on the board, the chief reason being that 1…d5 condemns Black to a long defensive struggle with very few winning chances. The analogous statement is not true of 1…e5 after 1. e4, a system where everything hangs on a tempo and Black often takes over the role of White after just one sloppy move by his opponent. By contrast after 1. d4 d5 2. c4, White by very simple means and with essentially no risk obtains a slight advantage and good prospects of converting it to a win. In the main lines of the Queen's Gambit Declined or the Slav, one or even two flaccid moves by White often leave Black still wondering where his counterplay is.

If Black wants to answer 1. d4 so as to preserve realistic play for the win, he needs to play 1…Nf6 and concede central space for the time being, or 1…f5, a dicey move that leaves Black's e-pawn weak. But I leave these subjects for another day and return to the position after 1. d4 d5 2. c4, where I have some suggestions about how to put together a winning repertoire for White. One of the very pleasant things about the Queen's Gambit is that White has more than one good way to play for a win against almost anything Black chooses. Therefore, I will be able to offer a considerable range of options to the player of the white pieces.

For example, if Black plays 2…dxc4, White has a reasonable expectation of advantage after 3. e4, which leads to a simple, open game and which I would recommend particularly to young and developing players. Personally, I believe 3. Nf3 is the strongest move, intending to meet 3…Nf6 with 4. e3 and straightforward play. I know some people who insist that e2-e3 should not be played after 1. d4 while White's queen bishop remains at home, but I think this viewpoint is silly. If Black after 2…dxc4 3. Nf3 Nf6 refuses to play …c5 followed eventually by …cxd4, then it is easy enough for White to get e3-e4 in, with excellent chances. White must, of course, be ready to play the role of the attacker with an isolated queen pawn. This is merely chess, and Tarrasch was right that if one never wants to have an IQP, he should give up the game. Thus, 4. e3 e6 5. Bxc4 c5 6. O-O a6 7. a4 Nc6 8. Qe2 cxd4 9. Rd1 Be7 10. exd4 O-O 11. Nc3 and White has excellent prospects.

The great Keres was a proponent of the White pieces in this position. How to handle this and many related and not-so-related IQP positions is necessary knowledge for chess players. I recommend Isolated Pawn by A. Mikalchishin, Ya. Srokowski and V. Braslavsky, Intelinvest 1994. This work, which is a compendium of useful examples of play, can be obtained from Chess Digest.

Another excellent system against 2…dxc4 is 3. Nf3 Nf6 4. Nc3. The gambit line that arises from 3…a6 4. e4 is highly favorable for White (though for play over-the-board, it requires a good deal of memorization). A minor catch is that Black can play 4…c6!, transposing into the Slav Defense -- not exactly a tasty meal for a compulsive gambiteer, particularly since 5. e4 is not considered very strong. But who wants to be a compulsive gambiteer, anyway?

This, then, brings us to the Slav, arguably Black's most stubborn defense to 1. d4. Highly sophisticated play is required for the winning attempt. Best by both is 1…d5 2. c4 c6 3. Nc3 Nf6 4. Nf3 (the order of White's last two moves can safely be reversed) 4…dxc4 5. a4 Bf5 (5...Bg4 and 5...Na6 offer White more winning chances) 6. Ne5 e6 (I don't have very high regard for 6...Nfd7 7. Nxc4 Qc7 8. g3, though it has been popular recently) 7. f3 Bb4 8. e4 Bxe4 9. fxe4 Nxe4 10. Bd2 Qxd4 11. Nxe4 Qxe4+ 12. Qe2 Bxd2+ 13. Kxd2 Qd5+ 14. Kc2 Na6 15. Nxc4 O-O-O (kingside castling is less reliable here) 16. Qe3!

Play is very difficult, but in my belief, White's piece is worth more than Black's pawns. Surprisingly enough, it is Black's king, not White's, that has trouble after such moves as 15…Nc5 16. Be2 Qxg2 17. Rhg1 Qxh2 18. Rxg7. Stronger is 15…c5, but even here, White has good play for the win, for example, after 16. Kb3. Promising though it is for White, this line should not be undertaken over-the-board without considerable prior research. And even in correspondence play, the positions are so difficult that some time should be spent examining them before play begins.

A more conservative but nevertheless very strong response to the Slav is 3. Nc3 Nf6 (3...e5 is just weak after 4. dxe4 d4 5. Ne4 and 3...dxc4 is untrustworthy) 4. e3!

This is the way the great Portisch played it, a giant whose games I would strongly recommend to any player of 1. d4. White's move looks unambitious, but it is a sophisticated attempt to sidestep the Slav, forcing play instead into the main lines of the Meran. Actually there is no "forcing" about it, but it just happens that 4…Bf5? 5. cxd5 cxd5 6. Qb3 is dreadful for Black, and 4…Bg4 5. f3 Bf5 6. cxd5 cxd5 7. Qb3 is not very comfortable either. The most solid is 4…g6, the Schlecter System, but there White has good prospects, for example, 5. Nf3 Bg7 6. Bd3 O-O 7.O-O Bg4 8. h3 Bxf3 9. Qxf3.

Black may not be losing, but his prospects in this position are quite dreary, I think.

This brings us, by logical progression, to the Meran System. After 1. d4 d5 2. c4 c6 3. Nc3 Nf6 4. e3 e6 5. Nf3 Nbd7 (or 4. Nf3 e6 5. e3 e6), White should play 6. Bd3. (I have no confidence in 6. Qc2 -- it is too slow.) I don't want to delve deeply into the theory of this difficult system, but only to say that White by straightforward moves such as e3-e4 and O-O has every prospect of advantage. White's compromise with e2-e3-e4 is matched by Black's with c7-c6-c5, and so White retains the advantage of the first move in a addition to the better pawn center and a position free of structural weaknesses. A critical line is 6…dxc4 7. Bxc4 b5 8. Bd3 a6 9. e4 c5 10. d5 (10. e5 also has something to recommend it) 10...c4 11. dxe6 fxe6 12. Bc2.

They say this position is equal, and I say, baloney! White retains his surplus tempo, he has the better center, and he has no structural weaknesses. Black, by contrast, has made irrevocable compromises. As Fischer once said of a certain Sicilian position, "This, I win in my dreams."

If you disagree, or if you are one of those dogmatists who insists that White's c1 bishop should never be imprisoned, however temporarily, then you can opt for 1. d4 d5 2. c4 c6 3. Nc3 Nf6 4. Nf3 e6 and now 5. Bg5 (you should be aware that 5…Nbd7 transposes, perforce, into the Cambridge Springs QGD -- see below). While 5…dxc4 6. e4 b5 7. e5 h6 8. Bh4 g5 is very double-edged, I believe the longest and sharpest of the two edges belongs to White. However, I would warn the reader away from 9. exf6 gxh4 10. Ne5 Qxf6 -- where Black's two bishops are a serious long-term problem. Instead, just 9. Nxg5 hxg5 10. Bxg5 with deep complications which, though requiring detailed preparation and study, are very promising for White. So promising, indeed, that many Black players have taken to answering 5. Bg5 with 5…h6. Now 6. Bxf6 leaves White with a pleasant a game but not very great prospects of scoring the full point, so recent attention has focused on 6. Bh4 dxc4 7. e4 g5 8. Bg3 b5 9. Be2.

This position is highly topical. White, in my view, has at least adequate compensation for the pawn, and very possibly more.

If Black tries the devlish "Triangle System," 1. d4 d5 2. c4 e6 3. Nc3 c6, White has three reasonable plays for advantage. Marshall's Gambit, 4. e4 dxe4 5. Nxe4 Bb4+ 6. Bd2 Qxd4 7. Bxb4 Qxe4+ 8. Be2! is highly promising. So is 4. Nf3, when Black can choose either the Meran with 4…Nf6 or the Noteboom with 4…dxc4. While the latter is one of the few viable winning attempts with the black pieces after 1…d5, I believe that White should keep some advantage with best play. Most precise for White, at least if one's repertoire includes White's side of the Meran, is 4. e3! when White has nothing better than 4…Nf6 5. Nf3 with a transposition to that system (not good for Black is 4...f5?! 5. g4!).

A system which, like the triangle, Black can always insist upon is the Tarrasch Defense: 1. d4 d5 2. c4 e6 3. Nc3 c5. This comes close to being an adequate defense to the Queen's Gambit, and it should hardly be dismissed merely because it saddles Black with an IQP. In most lines, Black's active pieces are adequate compensation. But after 4. cxd5 exd5 5. Nf3 Nc6 6. g3 Nf6 7. Bg2 Be7 8. O-O O-O 9. Bg5 cxd4 10. Nxd4 h6 11. Be3 Re8 12. Rc1! I believe that White has reasonble play for the win, while Black can hope only for the half point.

Certainly, when Karpov started to play this way as White, Kasparov gave up the Tarrasch. If anybody likes Black's chances in this position, then he should take up this defense, against which no line is better, and which is playable against virtually every closed system (for example, 1. c4 e6 2. Nf3 d5 3. g3 c5 and so forth). Yet, as Botvinnik pointed out, if the Tarrasch is good, then Queen's Gambit is no winning try. Like him and many others, I prefer to think that the Queen's Gambit is good, and hence, the Tarrasch must be bad!

We are left, then, with what Tarrasch called the "orthodox" Queen's Gambit Declined: 1. d4 d5 2. c4 e6 3. Nc3 Nf6. Against this, there are three major ways of playing:

1. 4. Bg5 followed by straightforward development and preparation for Black's ...dxc4, a move that is usually necessary to free Black's postition.
2. 4. cxd5 exd5 5. Bg5 followed often by a minority attack (b2-b4-b5 with rooks on b1 and c1), but sometimes by O-O-O and mutual king attacks.
3. 4. Nf3 followed soon by 5. Bf4 or, more rarely, by 5. Qc2 with the option of an early e2-e4.

Of these, (1) was considered best up until the 1950s, when the power of (2), the began to be recognized. System (3) was for a long time considered weaker, because Black can achieve ...c5 without serious compromise to his position. But the 1980s and 90s revealed that Black's game is not so easy even then, and today it is on at least an equal theoretical footing with the other two methods.

They say that White's advantage is insignificant after 15…dxe4 16 Bxe4 Rd8, but I'm not sure I believe them

Against Lasker's 7…Ne4, White should play 8. Bxe7 Qxe7 9. cxd5 Nxc3 10. bxc3 exd5 11. Qb3 Rd8 12. c4.

White has several good answers to the Orthodox, my favorite being Rubinstein's Variation: 4. Bg5 Be7 5. e3 Nbd7 6. Nf3 O-O 7. Qc2.

Against the Cambridge Springs, 4. Bg5 Nbd7 5. e3 (5. exd5, transposing into System (2), is a good option, since Black's fouth move impedes his c8-bishop) 5…c6 6. Nf3 Qa5, White's best play for advantage is 7. Nd2 dxc4 8. Bxf6 Nxf6 9. Nxc4 Qc7 10. Be2 Be7 11. O-O O-O 12. Rc1.

White has a space advantage and his opponent has difficulty finding counterplay.

In my own practice, I have shifted from 4. Bg5 toward 4. Nf3. Play is more complicated, and White has to be careful of handing over the initiative, but he has good winning chances. One little-appreciated point is that if Black soon plays ...Nbd7, White exchanges on d5 and plays his c1 bishop not to g5, but to f4. This form of the exchange variation, at least if White plays O-O-O, is better than the usual one. On f4 the bishop is less easily exchanged, and it stands out of the way of g2-g4-g5, a necessary sequence for White. Thus for example, 4…Nbd7 5. cxd5 exd5 6. Bf4 c6 7. Qc2 Be7 8. e3 O-O 9. Bd3 Re8 10. h3 Nf8 11. O-O-O.

This position significantly favors White, I believe, since White's pieces are much better disposed to back up his kingside pawn-storm than Black's are to support a response in kind on the queenside. The position is treated in the two games linked below.

Game 1: Morss - Raines US13P05

Game 2: Morss - Marples USCF-93RT21

For version 4 browsers: Java replay of all games

Next Month: My guest Gary Good on the Goering Gambit Declined with 4...Nf6!?