How should black play for a win against 1. d4?

The slav and semi-slav complex seems to be a reasonable try to win, but it is all bypassed by the exchange variation of the slav, which gives a symmetrical pawn structure.

For the QGD, Sam Shankland says that the opening is mainly playing for a draw, not a win. White seems to often hold a very small edge, and even if black equalizes, it is unclear how he is going to have winning chances. Pfren has recommended the Lasker variation in the past, but black exchanges his two active/developed pieces in exchange for safety, so it is hard to see good winning chances there. 

The QGA can lead to symmetrical pawn structures if white decides not to accept an isolated pawn, and draws can go as high as 50% in this line.

Instead of the queen's gambit, white can also try the London System. Even if black equalizes against this, white can play this system very solidly.

The Dutch defense is a try to imbalance the game from move 1, but Hikaru slams 1... f5 as not being as good as 1... d5 in his chess personality video.

Both the Benoni and the King's Indian have not been seen as great choices in recent years, with white supposedly having an edge in both of them. The Benko gambit has also had theoretical problems. Recently, Carlsen almost lost a game in the Benko gambit, saving a draw after his opponent blundered as white.

http://www.chessgames.com/perl/chessgame?gid=1926308

The Nimzo-indian complex is a try, but white can bail out with 3. g3, when black might as well transpose to a Benoni or a QGD. The Nimzo is also difficult to play as a club player; without strong opening and positional understanding, black can find himself with less space in a worse position.

The Grunfeld is very highly regarded, but it is also very difficult to play. White can also avoid it with 3. Nf3 or 3. g3. Is the Grunfeld approach really the only option left that both gives Black winning chances and is difficult for white to get an advantage from?

So how can black play for a win against 1. d4, against players who are lower rated and happy to draw against a better player? Help me, chess.com. You are my only hope.

Easy solution: play the Triangle. You like the Semi-Slav but don't like the Exchange Variation? This move order avoids it:

Additionally you can't play the Triangle against 1.c4 and 1.Nf3, so you'll need to play something different:

Unfortunately, here, white has the option of playing the dangerous Marshall Gambit.

Here, white's superior development, combined with black's weaknesses along the dark-squares without a dark-squared bishop, black's difficulty in developing his light-squared bishop, and black's lack of king safety more than compensate for 1 pawn. In my opinion, white has an advantage here, and it will require very strong defense and heavy book knowledge in order for black to fight for a draw, let alone a win.

Right, well you seem to be very picky with your openings

You didn't mention the KID - is it because you don't like it? It's not an easy opening to play though but it does very much hold its own at the 2700+ level.

Also I wouldn't dismiss the Nimzo. You seemed to be complaining about 3.Nf3 and 3.g3, but these aren't something to be worried about as you play 3...d5 and then opt for a dynamic setup.

I just play the QGD (and QGA sometimes) but I don't mind as I tend to enjoy "boring" and "drawish" openings, also the Petroff. The OP probably wouldn't enjoy the QGD as it doesn't give black clear winning ideas, rather equalising ideas where you win by just outplaying your opponent.

Any of the many standard openings would be fine. Just play better moves than your opponent and you'll win a high percentage of your games.

Get a solid repertiore against 1. d4. Focus on the kind of plans that you get from the opening. Get your tactics in order, then your endgame, then tactics, middlegame, tactics.

And fix your ego. Kill your fears with studying and sparring the opening lines that you fear and, more importantly, get rid of the belief that the outcome of a game is decided in the opening.

Excellent advice! Thanks!

I am going to learn how to play the King's Indian as black.

Indeed, inspirational! I'm going to take it a step further and learn the King's Indian Attack as white!

KID as well as NID, QID, Bogo served me well.

No matter how often such things are asked: you have to learn your openings well. Tons of theory and even more praxis, years of practicing.

Which opening doesn’t matter after all.

You shouldn't expect any winning advantage in the first 10 moves of opening. Playing with black against strong player,  sometimes I dont even have a winning chance up to move 30, and sometimes struggle to get draw from start to end.

But every chess player does mistakes ( including  3500 rated Stockfish running on super compters do mistakes), and keep looking with eagle eye to get an opportunity to punch him whenever he does mistake.

Frankly, at your level (and mine) avoiding symmetry is not as important as understanding the middlegames you're getting.

Of course if it's a must win game, you should generally avoid symmetry, but as far as general repertoire choices don't worry so much about the slav or KID or whatever. Your best winning chances will be from the middlegames you know best, and to learn them play over a lot of GM games, and play the openings that lead to those middlegames often in OTB tournament chess to gain experience. That's what will make you a dangerous player in _____ opening.

And just as an example of the basic questions you'd ask when looking at GM games...

What area of the board did white seek play on? (queenside, center, kingside). Same question for black.

How did the players seek play in that area? (pawns or pieces) Which pieces had the big roles? Which pawn breaks were played?

If you answer these questions for 3 or 4 games, you may get 3 or 4 different answers. Not so helpful. But if you keep doing this you'll start to notice patterns. If you do this for 50 or 100 games you'll have a pretty good idea of what those middlegames are all about. And this shouldn't be a long process. Spend 5-10 minutes a game, less than 1 hour a day, and you can look at 100 games really easily in just a few weeks.

Well let me show you what I mean, GhostHorse0. This is a game that I played a few months ago OTB. At this point in time, I was approaching 1600 USCF- my rating is now (overrated) at above 1700 USCF. My opponent is at his rating floor of 1500.

I won this game, but I consider myself lucky to have been able to. As the lower-rated player, he had every incentive to draw against me. By move 18, we have an almost identical position, except for one key difference- both my rooks are on the queenside, whereas his kingside rook is sitting at f1. He would not move this rook until move 28. If he had moved both his rooks over to the queenside by this point, I would be quite hard-pressed to find a win against him.

The computer assigns a score of 0.00 to the position by black's ninth move. In theory, that means that the opening was a success for me as black. In practice, I am very fortunate to have outplayed my opponent in an almost symmetrical position. I like Bobby Fischer's attitude of playing for a win with black as well as white, and I do not want to draw against weaker players.

As always, analysis of my game is appreciated. 

If you want to play for a win with black like Fischer then you'll have to be willing to accept bad positions... at least positions the engine will tell you white is practically winning out of the opening. You just have to ignore that eval and study GM games to find the main ideas (in bad positions sometimes the engine is suggesting really awful moves unless you mess with e.g. the contempt setting).

---

As for drawing against weaker players, no one likes to do that, but you don't have to go to extreme lengths to avoid these draws unless you're a top GM. You can play practically any opening for a win. But if you like the idea of imbalanced fighting games, then sure, play a KID or something, that's fine... but do it because you like it, not out of necessity

See post 17 for context.

one of the ways as Black that I am starting to adopt is to play the Tarrasch Variation of the Queen's Gambit Declined. Black can adopt this defense through a multiple set of openings by White. Play can go 1.d4 d5 2. c4 e6 3. Nc3 c5 and White can play either 4. cxd5 or 4. Nf3 as well as other replies. Black usually develops his Knights in alphabetical order and the dark squared Bishop to e7 and castles short. I read through a book by Eric Schiller titled A complete defense to Queen's pawn openings. It is definitely a way to go if Black is fine with defending an isolated Queen pawn.

Yeah, the exchange slav is pretty dry.

A few tips

1) Other than rook imbalances like you mention, you can develop your knights in an imbalanced way too  If they put theirs on the 3rd, you can go Nf6 and Nd7 for example. This is a way to imbalance a game where the structure is symmetrical (although a bit subtle I admit, still, strong players mention these things, I didn't come up with it myself).

2) When beating lower rated players, 0.00 in the middlegame is not something to worry about, but having no mechanism to play for a win is. It helps to have a mindset that "this game might take 100 moves to win, and I'm ok with that." If you keep applying pressure (via your winning mechanism, which can change as the game goes through its phases) they'll eventually show you why they're rated ____

3) 0.00 out of the opening is not necessarily what you want. For example I like the Benko gambit, and sometimes the engine really complains loudly, giving white +1.00 out of the opening. What the engine thinks on move 10 isn't so important, what's important is whether you like and understand the middlegame in front of you.

IMO the tarrasch is a great choice for a lot of reasons. Educational (IQPs and hanging pawns), imbalanced (when you want it to be), very hard to crack (when you play it a la Kramnik, semi-tarrasch), and you initiate it early (on move 3) so there aren't a lot of ways for white to avoid it.