Looking for advice in finding a better plan

I have played a game , where I did not find good ideas, could someone review ( I have already added comments ) and help as I felt quite stucked , what concept am I missing or any suggestion would be appriciated

9...Bg4? is a mistake. Your bishop is developed and active on f5. Do not move it a second time. You have not yet castled. Your king is not safe and your rooks are not connected. 9...O-O
14...Bc3? is a mistake. Your bishop is well placed and safe, but your rook not yet. Hence 14...Rfe8: move the rook to a file that is about to open. On c3 your bishop is in jeopardy.
19...axb5? is a mistake. You could protect your bishop with 19...cxb5.
28...Kg8? is a mistake. The king was better Kg7, where he protects all 3 pawns..

Sounds like I did not need to move with my both B multiple times, and develop my Rs on more central squares before starting to attack

Hi @Zelenko, here is my analyze.

9...Bg4?! loses a tempo, because you cannot maintain the pin and will end up trading off a strong bishop. Moreover you give up the bishop pair and move a piece twice in the opening. As you said, castling was better. A good line was for instance: 9...O-O 10.Ng3 Bg6. You still have the bishop pair, a strong bishop on g6: black are slightly better.

After 12.Qxf3, the position is as follows:
Black is better developed (White still has their bishop on c1).
There are no big weaknesses in the pawn structure: b3 and d3 on White side; d6 for Black. Nothing easily exploitable.
The d- and c-files are semi-open.
King safety: both are fine, but Black is better (White has h3 and the Ng3 defender is poorly placed).
White controls the center slightly better.
Piece placement: the black bishop is hanging and would be better on d6, but not immediately (it controls e1, for instance)

As a result, the position seems to me equal. How is it possible, since 9...Bg4 was not good? It's because the Ce2-Cg3 maneuver was very slow and lost many tempi for White.

Now what could be the White's plan? 1) Complete development (develop bishop and put a rook on the c column). 2) Push the e4 pawn (which is currently blocking their bishop). 3) No queenside expansion is currently possible but their pieces are well placed for an attack on the Black kingside.

As for the Black's plan: first take control of the center (Bd6, Rd8 or e8, Nd5). Be ready to strike the d4 pawn if White advances e4.

12...e5. I disagree with Stockfish (lol). It opens the diagonal for the White's bishop, which cannot be challenged since your light-squared bishop is gone. After 13.Rd1!, the opening of the columns will favor White. As you said, you should have develop your rooks on e,d or c columns before demolishing the center. If you still wanted to undermine immedialty the White center, 12...c5 would have be more precise, IMO.

14...Bc3?! As @tpr says, it's a useless move. The rooks need absolutely to be brought to the files that are about to open.

15...a6. Sadly, your plan (attacking Queen side) could not work. Pushing the b-pawn isn't possible without losing something and the lines will open for the White's bishop pair. The center will soon explode (look at the pawns d4-e5), controlling it with Re8 and Rd8 was necessary.

After 21...Red8, the position should be equal if it wasn't so tactical (but I'm perhaps wrong here, this position is too complicated for me). Stockfish finds clever ways for White to hold a pawn on the queen side. But if all the queenside pawns are exchanged, and likely a pair of rooks as well, there is no way for White or Black to gain a decisive advantage. And once further material will be exchanged, the game should end in a draw. That's what happens on the variant White chose: some inaccuracies and the game is now equal.

After 25...cxb3, you are not a pawn down, as you said in the comments. You even have a very advanced, dangerous passer on b3. White cannot capture it, as doing so would require bringing their king out (a very risky move given the presence of your queen and rook). The plan is holding this pawn and bringing it to b2. This is enough to ensure you at least a draw.

27...Kf8?? is of course a blunder; you end up a pawn down with a poor pawn structure and now, no remaining counterplay. Nevertheless, the position remains defensible, IMO. Since White have no distant passed pawn, their only winning plan is a kingside attack, yet it appears you can easily defend against it.

Thus believing you still can win from this position, or that you have to counterattack, is your truly fatal error. You need to regroup your pieces and build a fortress around your king, and White cannot really advance or threaten anything.

It was very interesting to analyze this game, thank you for sharing it!

Thx @Goran_Pastek on very deep and detail analize of my game, it will for sure help my future chess progress !