Middlegame strategies for black

Black to play...What would be your strategy here as black. Do you think there are some tactics or it is a positional game from here. Thanks.

Considering that black has the following advantages:

1. More space on the kingside.

2. The majority of blacks pieces are pointing at the kingside.

I would play ...g4 attacking, and forcing the knight away.

If g4 the N can go to h4 making it difficult to open lines on the K-side. Better, I think would be Rdg8 and then start pushing the h-pawn

Nh5 ...will attack bishop...and damage structure...then pawn march

That’s maybe a bit unnecessary since you could play ...h5 straightaway and white can’t take the pawn on g5 anyway. 

Black looks to have a few reasonable options here; ...h5 immediately, ...g4 immediately followed by ...d5, or a semi waiting move such as ...Kb8 or ...Bb6. The position isn’t particularly sharp Right now but could become so. 

After 1...g4 2.Bh4 looks very tricky. And after 1...h5 2,Bxe6+ looks OK for White. I don't think it's all that easy for black

g4?? is probably the worst move among about 20 possible moves for  black! (except dropping pieces moves)

imo, h5, since it creates the threat of h4. Playing h3 (to play Bh2 on h4) is obviously sub-optimal for White, since it creates a "hook", that is, a target for Black to attack. (via g4, or some immediate sacs w/ Bxh3....although I haven't calculated it) 

g4 is the right idea, and you probably didn't know it was quite that bad until you plugged it into your engine. Sure giving up h4 is not nice, but the only way for white to punish it is Ne1.

Rdg8 would be pretty automatic for me here, preping h5-h4, and the engine says that's fine, but looking with an engine I see I didn't realize how delicate the position is for both sides.

As for the OP's specific question of if this is a strategic game or what the strategy is... that's pretty obvious. No one plays g5 then castles opposite sides without the very obvious intention of mutual mating attacks. Although black's d5 and white's d4 break complicate things. It's an interesting position, but I'd like to see the moves that lead to it. I get the impression there were some important improvements before this position happened.

@llamonde2, what do you think about h5 immediately?

EDIT - I mean, g5 hangs, but still, it opens the g-file immediately, and h4 is still a threat

My short analysis thought h5 was fine, but I thought I had time for Rdg8 so my thinking was why not.

But after that I looked with an engine, so whatever I say that goes beyond this short analysis doesn't matter anymore because it's not my ideas or lines it would just be parroting the engine 

As for something I can add, I'm more comfortable with the g5 stuff in the guccio piano when white plays h3. Without h3 I'm not as confident, although here white got his bishop stuck on g3 so I don't know. I get the impression black could have saved time by delaying castling and trying to press his attack a little more, but without seeing the game it's just a guess.

What I want to know is how the bishop ended up on b3. Also Be6 was probably an inaccuracy.

Thanks for the reply

I haven't looked at it w/ an engine yet, so....maybe i should rn

I agree w/ everything you said about playing Rdg8 over h5 immediately, and it's probably more sound, but I do still think that h5 is faster. (Then again, I have very weird (unsound) attacking ideas. Don't ask ) Maybe something like Bxe6 then b4 by White after Rdg8, idk, but White might get some good counterplay

One the basic ideas of d3 is that (big surprise) you're not playing d4

So Bg4 (correctly) seems dumb.

But when you're attacking like this (and there's B@g3 and N@f3) I've seen Bg4 in GM games... honestly I don't really understand it, but I know it happens. I remember this one pretty clearly because I had a disagreement with a guy about it after a tournament game. He said Bg4 was dumb in d3 lines

lol, looks like one of my first (and last!) attempts playing the Giocco Piano (Evans Gambit ftw!). White castled too early, played Bg5 pinning and then h6, g5

Okay, so after some engine analysis, apparently Rdg8 is better than h5. On h5, white has h4! which is counter-intuitive....why would you push your h4 pawn. It turns out though that Black doesn't actually have a way to continue the attack. WIthout this resource Black would be much better

I did not do engine analysis  when I did that post.  Obviously black should not allow white to block "h" file that will delay king side attack too much. White has quick counter Bxe6, b4, Nc4  to neutralize black's powerful bishop on c5.

When I let engines run, the top 3 choices are 

1. Rdg8

2. h5

3. Bb6

As the position is quite complicated with counter play from white, even Stockfish and Lco struggle to find the best move.  After 500 Million nodes , Lco 256x20-t40-1541 found Bg4!!

I manually entered that move into Stockfish( and take back ) and Stockfish keep sticking with that move!

The best principal variation for black is Bg4!! (h3,h5) and black sacrifice bishop and chase the king on "h" file.

Note:     : P.S , I analysed to find the best move "Bg4" for my personal curiosity, it is practically out of human brain power. 

Okay, so after you saying that Bg4 is best, it is actually pretty obvious why -- It pins f3, h5-h4 is coming. But perhaps most importantly, it prevents Bxe6 and thus b4. 

If we actually calculated this position (I'm pretty sure most people -- myself included -- just said the first move that popped into their mind), Bg4 doesn't seem "out of human brain power"....certainly it's difficult to see (I wouldn't have seen it, but a titled player probably would). You would just need to think prophylactically. Bxe6, b4 is white's main source of counterplay. d4 isn't possible, and White doesn't have another way to attack Black (and if White can't attack, then Black is better bc of the initiative). So how to prevent Bxe6?, well...you can move the Bishop or play d5. d5 seems a bit premature (and hangs e5), so moving the Bishop is best. G4 is the only square that is safe and prevents the trade of Bishops, so by process of elimination it's best.

On Bg4, h3, it's not really a variation that needs to be calculated much -- if at all. It just looks intuitively good. Also, most players would recognize that pattern from the exchange variation in the Ruy Lopez:

In your position, there is only single threat that you dont need to think too much branches/ variations.  For example, for 8 ply (4 moves ) tactic, you can think about two variations in each move=. 2^8= 64 positions is generally enough to ensure your tactics/ sacrifice is right.

But above OP position has so many threats and counters to consider. For example, 8^8= 16 millions+ positions.

Going back to OP's position, you can guess and play Bg4 sacrifice and take risks! However, I would  not take those risks and will simply play Rdg8!! and then h4 push.