Queens Gambit is not a Gambit (complete guide)

Classical Beginner Analysis. Actually, it is. There is an opening called the Queen's Gambit Accepted. Black can hold on to the pawn in many variations but White usually gets some compensation in the form of an attack or central control. 

Could you please elaborate on some of these variations? I'm sure there are plenty of players that only play the (Semi)Slav or QGD because they've become afraid to take the pawn. I feel maybe they're not too sure how to know if White is not being sharp and you can get away with the extra pawn or you should give it up and complete development instead of tying yourself in knots defending the pawn.

I read so many wrong things here...

First, many wrong statements about queen's gambit accepted.

After 1.d4 d5 2.c4 d5xc4, white has 3 main moves : 3.Cf3, 3.e3, 3.e4. 3.Nc3 can be countered by 3...a6!

For many years, the main idea was to recover c4 with Nf3 first, then e3.

But black discovered a variation we could call the Svidler variation, based on 3...a6 followed by b5 :

I don't give analysis but just believe me, this variation just kills white's advantage.

This line became very popular at high level, it was the drawing weapon on 1.d4 d5 2.c4.

That's why fashion was 3.e3 and 3.e4.

I don't know the actual theorical status of 3.e3, but at least, white avoids this Svidler's line.

Same with 3.e4 with some fresh positions.

About 3.e3, I don't understand the variation given by OP ;

1.d4 d5 2.c4 d5xc4 3.e3 Be6 (indeed a critical move, instead of the classical e5) 4.Qa4 and now of course 4...c6 when white cannot recover c4.

Even 1.d4 d5 2.c4 d5xc4 3.e4 b5 is playable, as given by Glenn Flear in the book Dangerous Weapons - Queen's gambit.

It is not so simple for white to recover the pawn...

"But black discovered a variation we could call the Svidler variation"

Svidler was not even in the plan to be born when that 3...a6 move was played for the first time, 4.a4 is the answer to that, it has been played since about 1936 in thousands of games.

This is pathetic. It's called a gambit since White gives up  a pawn, even if it is temporary. 

I read so many wrong things here...

First, many wrong statements about queen's gambit accepted.

After 1.d4 d5 2.c4 d5xc4, white has 3 main moves : 3.Cf3, 3.e3, 3.e4. 3.Nc3 can be countered by 3...a6!

For many years, the main idea was to recover c4 with Nf3 first, then e3.

But black discovered a variation we could call the Svidler variation, based on 3...a6 followed by b5 :

I don't give analysis but just believe me, this variation just kills white's advantage.>>>

Aronian played badly in that game in post 28 but of course, in any case the QGA is fine for black. White can get a bit more space and some pressure but that's all. In return, black gets piece activity.

But I think Aronian, as well as playing badly, made a strategic mistake in not playing ab before bc. He was probably thinking that white should keep pieces on but that's illogical in that position. The two a-rooks cancel each other out so white may as well swap them off because white gets two advantages from that. One is more squares for his queen and the other is that in some variations black's bishop may want to swap diagonals, which takes an extra move from a8. People are playing this Svidler variation against me and I worked this out over the board the first time it was played. I usually get the better position .... maybe players of black are trying to copy GM moves down a line that isn't the best for white.

I have enjoyed the discussion of the QGA but you are all missing the original point. The point was not that I gave the best lines or that I gave a complete guide to the QGA. The point and the only point was that after 3. C4 which is by definition when the opening is titled "Queens Gambit" white has not sacrificed a pawn yet because tactically white can win the pawn back in every scenario. Despite many people saying the original claim is wrong no one has disproved the premise. If black does not take the pawn immediately then white can just take d5 if they so chose. 3. c4 does not gambit a pawn. Again it is probably the case that white should gambit the pawn for a better position but I never claimed that winning the pawn back was best just that it was optional. 

Yes you are right, after 1.d4 d5 2.c4 dxc4 3.e3 white can win the pawn back or get a won position if black insists in directly defending the pawn (there is not question about that), but black does not have to directly defend the pawn, black can play 3...e5 instead and equalize the game if white takes the pawn back right away

3.e4 is more testing than 3.Nf3, and, as discussed, 3.e4 e5 does not equalise immediately for black like you said originally; in fact ir is harder for black ti equalise than after 3.Nf3. This is well-known and most GMs are now playing 3.e4 as it poses black more problems.

Yes you are right, after 1.d4 d5 2.c4 dxc4 3.e3 white can win the pawn back or get a won position if black insists in directly defending the pawn (there is not question about that), but black does not have to directly defend the pawn, black can play 3...e5 instead and equalize the game if white takes the pawn back right away 

I just realized that you give your opinions without even bothering checking the positions with an engine first.  Way to go! buddy