Queens gambit accepted - 3.e3 or 3.e4?


I prefer Nc3. I don't think that any move order of what you mentioned will make any huge difference. Playing Nc3 disallows black of protecting the pawn (unless making bad moves).

@tygxc
Lol 😆 you just can't stay away from comments. Your placing of moves is based on what?? 0.01 difference of advantage?? Like i said it's a matter of preference and all moves mentioned are good

@tygxc
Lol 😆 you just can't stay away from comments. Your placing of moves is based on what?? 0.01 difference of advantage?? Like i said it's a matter of preference and all moves mentioned are good
Yes, it is a matter of preference and all moves mentioned are good. But in fairness, I must admit that in playing the QGA myself I find that 3.Nf3,... by white does hinder black from playing e5 which the other 3 moves generally do not.

Just a matter of taste. 3.e3 allows e5 with IQP positions, 3.e4 allows b5 (the latest trend) with sharp, double-edged positions, and the most flexible move 3.Nf3 allows flexibility on Black's part as well.
I have mosty played 3.Nf3, but the 3.e3 positions are rational as well, and do not require a lot of memorizaion. Hardly the case after 3.e4.
According to computers, grandmasters, and normal players alike it is not overextending and it is a normal QGA theme. 3. Nc3 is probably inferior to 3. Nf3 and 3. e4 if not dubious overall.
Mine are GM Max Dlugy’s book on the QGA, engines which say Black easily equalizes after 3… a6, and lichess master database which has a 23% winrate for White and a 43% winrate for Black.

Mine are GM Max Dlugy’s book on the QGA, engines which say Black easily equalizes after 3… a6, and lichess master database which has a 23% winrate for White and a 43% winrate for Black.
Quite obviously you haven't read the Dlugy book.
He claims that after 3...a6 4.e3 b5 Black may equalize with precise play, but he has no winning chances, and of course 3.Nc3 (which he does not examine at all) a6 does not prevent 4.e3- actually 4.e3 works fine because there is no knight at f3:
1.d4 d5 2.c4 dxc4 3.Nc3 a6 (probably safest is 3...e5 4.e3 transposing to the 3.e3 line) 4.e3 b5 4.a4 b4 (positionally forced) 5.Bxc4! and Black cannot take the knight because of the stock Qf3 trick.

Here are three answers and they are not easy.
The first is based on a computer that already beats grandmasters so its rating should be better. The latest Stockfish version prefers 3.e4 but the one before that 3.Nf3
The second answer is based on analyzing the games of players of your level according to what they play. For example, for players rated 1800-2000, perhaps 3.e3 is best because there is the highest success rate for White. But here it requires a longer analysis and calculation. For example, after 3. e4 the best answer for Black is 3. ...e5 (or even 3. ...Nf6). But 1800-2000 level players most often play a very weak 3...e6 and White has a good game. So it is necessary to calculate what my opponents play most often and choose the opening accordingly.
The third option is to find your favorite player and see what he plays. For example, Caruana plays 3.Nf3

I think they are all about equally strong, with different plusses and minuses. I've played them all, and also faced them all as black since I have often played the QGA myself. I wouldn't have an issue with any of those three options as a part of a student's repertoire.
If I had to choose one I like the most, probably 3. e4 but that's just a personal choice. The difference between these three moves is pretty subjective, all roads lead to like +0.1 or +0.2 and ultimately black is fine if they have done their homework.
I agree that 3. Nc3 makes black's job a lot easier, and 3...a6 is the typical recommendation although there are other ways.