I don't understand this opening.

http://en.lichess.org/training/opening/20938

I found multiple correct answers from this position.

exd4, pawn takes pawn. Makes some sense it opens the position up a little, and relieves the tension.

e4 was also correct. It removes the tension in the center, and gains tempo on the knight, and that pawn grabs a bit of whites space.

I then tried h6. It grants a flight square to the king, and keeps the knight back. A sort of prophylactic move, but it really does have some merit.

So I got 3 of the 4 correct answers.

I then tried rook e8, and it was called "a good move, but you can do better" Now this move puts the rook on a file that may be opening up developing the rook stacking it with the queen. I can see logic in it.

I tried king h8. It gets the king out of the pin. They said it was a good move, but there is better.

I couldn't find another sensible looking good move at this point. I played a random move to get it wrong, and see what the computer suggests... (Apparently g6 is also good giving you a flight square, and that's correct...)

However one move marked as a correct answer I just don't understand!

a6 is a correct answer!?

I don't see any real benefit to a6 at this point in the game.

What could a6 do?
It could help keep a piece from b5. However the c pawn already does that, and furthermore no piece in the white camp is really looking to be going to that b5 square.

perhaps it helps prepare a b5 push? but... this would leave a backwards undefended c-pawn and I don't think that is a good plan.

I'm not saying a6 is a bad move. I don't see it as a good move working towards any plan though. I wouldn't call it a blunder or mistake, but I don't see the strategy behind it unless it is a very very deep calculated long term plan...

Can someone please explain to me why the move a6 is superior to a move such as king to h8 or rook to e8 that seemingly have more immediate benefits?

Also the move a6 makes a hole on b6 for an enemy knight to make outpost on potentially

Hm, I got a5 instead of a6; tried again — both a6 and a5 are acceptible :D
I could say a few words about a5, which later could be used for a4 (kicking the bishop; although now it's impossible, it needs some preparation) and maybe a3 after. It's some kind of expanding on the queenside as well. I don't think b6 square is that weak, as the knight, if it comes there, will most likely have no pawn support .
And a6 does seem weird, and the engine doesn't like it either. It's not blundering anything, but for me it's a tempo loss → strategical mistake. I took a look at engine lines; if a6 is played, it will play a5 later anyway.
But sure a5 is not the best move, black has to react quickly in the center.
I personally like Re8 best :D

Maybe a6 prepares for white's move d5? And after bxd5; black not only has a temporary strong center (soon to be challenged with Qxd5), black also continues to guard the b5 square and prepares a queen-side attack. Q-b5 is no longer possible restricting the development of both the knight (which is tethered to the e5 pawn) and the bishop (tethered to the defense of b7). The c8 bishop may later be fianchettoed. Only thing I can come up with :-/

So:

... a6
d5? bxd5
Qxd5 Rd8

So:

... a6
d5? bxd5
Qxd5 Rd8

Is that Rd8 on the end sound? Can't white start stacking the rooks to support the queen then? how does the knight move after Rd1 following your line?

Yeah there are probably flaws. I was just rambling to see if we could make any sense of this. :-D

so far i have got a5, a6, Rd8, Rb8, exd4, e4 all listed as best moves, there are apparently more than 4 answers, only four required for completion.

also while i can't make sense of a6, a5 does seem to force a4 as a response. locking a pawn on a light square.

i can now h6 and g6 for a total of 8 acceptable answers.