Don't trust engines for opening evaluations, they can't see far enough into the middlegame to understand certain openings. It also calls both the Benko Gambit and the 1.e4 c6 2.d4 d5 3.e5 c5 Caro-Kann inaccuracies.
To fix this, it would be cool if Lichess could give Stockfish an opening book. Or have some qualification, like within the first 5 moves, if it is in the masters database in at least 100 games it can't be considered an inaccuracy.
I heard that even 1.d4 g6 gets an "inacc." (instead of a blunder! lol)
Just joking, I used to play 1. ... g6 - no big deal! But "inacc." rings true, I had the same experience with other opening moves.
"Or have some qualification, like within the first 5 moves, if it is in the masters database in at least 100 games it can't be considered an inaccuracy."
+1 for solving "inaccuracy" for openings in any manner not requiring a Stockfish change.
@sargon Delchev has actually co-written a book on the QGA (I recommend it, the lines given are fairly non-theoretical and easy to learn, and let you play for a win with black) in which the game you posted is featured. I play that line myself with black.
Yeah, my opponent told me that in Vienna. I used to be a e4 player most time of my life that's why I wasn't quite sure what to to. I found f3-f4 otb and had a similar structure (with h2-h3 though, played that earlier in the game). Black had definitely compensation, my game ended in a draw.
