scandinavian sound sound

why is the scandinavian so sound for black?

it violates a lot chess rules.

queen out to early and neglets developments and moves the queen more than once.

white develops nearly all pieces in the opening. while black has a lot of pieces on the backrank.

I don't get. would a super gm be able to  punish blacks violation of chess principles?

The problem is, that black gets enough time to devolop pieces because all his moves are "forcing" so he is gaining devolopment and pushing white back too.

You said you shouldn't take your queen out too early. I agree with you but when you look at some lines the queen usually goes back or runs all over the board.

There are also many tactics allowing black to gain tempo plus letting the queen run safely. Novices can more than often mess up in this opening as black or even white when black plays aggresively. 

And i wouldn't call it super sound.

I agree that its losing time. I think it can be used as a surpise opening like in this case

Not the best game to use despite the result.  Caruana actually had an advantage for a good portion of the game, and Carlsen only won because he did a good job of taking advantage of his opponent's mistakes in the middlegame.  Remember, the result tells you nothing about what happened in the game itself.

The Scandinavian is okay, where Black tries to justify his early queen movements by setting up a solid position and forcing White to waste time going after the queen.  Not the best opening, but it's perfectly sound.

Oh, and all of this is assuming it's the 2...Qxd5 variation.  There's also the 2...Nf6 variation, which is not quite as good, but it's a completely different can of worms.

Play it if you want. It's playable and can even be fun. I used to pull it out of a hat every now & then years ago. Too many disasters to make it worth while though. Maybe you have to be a GM to play it decently.

Now this is what I calltwisted logic. Lynch-like plot.

Why is it sound? Because it works tactically...concrete variations are everything in chess. Principles are of course good but one shouldn't follow them blindly, concrete variations must be checked.
There is no principle that works in every chess position.
And it doesnt't really lose that many time. First of all, it is white who is basically losing a tempo first after exd5, moves the e pawn a second time. After Qxd5 Nc3 white gains a tempo that's true and probably another one after the queen retreats again. But white cannot play c4 because the knight is on c3 already. 
So it loses 2 tempi probably, (and black's position isn't very active but pretty solid) and then what? In the classical variation of the nimzo indian white also loses time with the moves Qc2 and Qxc3, and nobody seems to question it. 

The ...Qxd5 variations are OK for Black, provided that he is not terribly ambitious, and can defend  a healthy but rather cramped position for quite some time.

thanks for the input, I guess I want the principles to be a guideline to play chess well - but ok things aren't that simple.

I saw this video about the scandinavian https://www.youtube.com/watch?v=jp2S_ujxwxY and the GM that reviews the game has some ideas to punish the queen moves. The Qxd5 doesn't look that good later in the video.

edit: wrong video.

I dont't see why 2.Nf6 is bad. It could transpose to any line with Qd5. How is Nf6 bad?

Because there are lines in each of the 2...Nf6 lines that give Black trouble, especially in the fianchetto variations with an early c4-c5 for White:

The way i see it is that you don't capture the pawn right away.

3.Nc3 is no big deal for Black. 3.c4 and 3.Nf3 make things a little more "interesting." I think you have to be willing to sac that d-pawn if you want to play 2...Nf6. There's a nice lecture about it on the Melbourne Chess youtube channel.

Fine by me as White, because I think Black doesn't really have much compensation for the pawn due to White's heavy influence in the center after Nf3 c6 dxc6 Nxc6.  Feel free to prove otherwise, though.

I have yet to see a refutation to the Portuguese that I've been fully convinced by, so long as Black retakes on d5 against 3.Nf3. Though perhaps when Smeardon's book comes out I'll get to see what the true holes are.

Also note that 2...Nf6 just transposes after 3.d4 Qxd5 or 3.Nf3 Qxd5, so that move itself cannot be less sound than the mainline Scandi. White's remaining responses are all inferior. So it would seem the move has equal value.

Only a true masochist would take Black after 3.d4 Bg4 4.f3 Bf5 5.g4 Bg6 6.c4.

Smerdon himself has not repeated his Portuguese experiments since 2006, and currently the only strong player who's employing the Portuguese with some regularity is the Philippino GM Darwin Laylo, but his results are not enviable.

And of course if white does not care memorizing several moves, 3.Bb5+ is a perfectly good, and rather simple option.

3.d4 Bg4 4.f3 Bf5 5.g4 Bg6 6.c4... presumably intending Nc3 and Qb3 at some point, is an ambitious try but not without risk. White gets a pawn but also some weak squares, and Black's lead in development is not trivial. I've scoured those lines with an engine and White's advantage does not seem any greater than it is in quite a few respected mainlines. I'm not going to claim this is sound or equalizes by force. I'm really not sure, and I'm certainly not qualified to weigh in on such things. Black's position may have some kind of fundamental flaw, but proving so is not easy. A mistake from White after 2...Qxd5 gives Black equality. A mistake from white here gives Black a win.

Better still, 3.d4 Bg4 4.Nf3 Qxd5 5.Be2 Nc6 6.h3 - I believe this decline line has been a bigger problem for Portuguese players in terms of results than any accepted lines.

3.Bb5+ seems rather not so simple after 3...Bd7 4.Bc4 b5 5.Bb3 Bg4 6.f3 Bc8.

If someone wants to play for an advantage with minimal prep, there's always 3.Nf3 - those lines of the Portuguese do seem far-fetched. I think Black should retake.