how can we analyze a lost game ?

hi,
i've started learning chess on my own, going on and off for years now. I really would like to improve. I've also read that's it's very important to analyze your losses.
So here is my game:


Do you take notes after you've played a game?
Which ones?

I really tried to play the best moves i thought about.
It was my first game with a small timer to think about 30s/1mn for each move. Because i've noticed that i play too fast. But it's even hard to take time to think because the opponent just answer directly.

Back to my game, for example:
on move 12: i didn't know what to look for in this position
i noticed the g5 pawn it seemed important to keep it alive while aiming at the h7 pawn at the same time, but the computer doesn't seem to like my idea.
on move 15 i wanted to take moves to replace my bishop on the diagonal b3 to g8 (but the bishop got trapped by black pawns)
on move 18 Qd2 was to keep the bishop in the corner of the king to maintain the black king in a small place
It was "small plans" but the computer tells me "wrong ones" :p

On moves 15 to 31 the computer analysis tells me i'm losing more and more (even before blundering) but i don't really understand why.

Any help on how you analyze your losses / how you improve after a lost game?

You are supposed to find your mistakes and find a better move/plan. This can be applied to every move.
12: What stands out for me is the mass of pawns on the king side. I would start pushing these pawns; f4, g4, and g5; to cramp (take away space from) black and, maybe later, open files to attack, especially for the bishop.
15: Not only does his pawns restrict your bishop, the pawn moves allow his bishop more freedom. Once again I like the idea of expanding on the king side with f4. If the knight retreats, c4 (that a2-g8 diagonal) becomes available for your bishop and you can open more lines with f5.
18. You have the two bishop advantage, and you should preserve that long-term advantage by retreating the attacked bishop.
Notice that you pieces seem to be retreating almost every move from 15-23. After f3, notice how weak the dark squares look around your king. This allows him to attack with three pieces (queen, knight, and bishop) against your lone king.
Although computers are almost perfect at tactical mistakes (like the mate-in-three at move 32), they still lack in some positional aspects. You need to use your own judgement when using the computer's suggestion on positional/strategic circumstances.

the only way to justify the doubled pawn on f5 is to play aggressive with your kingeside pawns.
if u dont feel comfortable being so aggressive and leave your king weak then dont play 10.Nf5

Thanks a lot for your analysis & tips!

So i've to look deeper for positional/strategic understanding and computers won't help me much on that side :)
While playing i didn't realize i was getting a doubled pawn on f5 & that it meant i would have to go for a pawn attack on the king-side.
I better understand know why i read about players talking about using computers "last" to have a look at tactical mistakes.

You played ok, but you should defenetly bring out your rooks earlier.

Move 12: Bd3 is passive. What does the bishop do there better than on e2? It stares at its own pawn at f5. Better develop the other bishop: Bf4 and the rook: Re1.

Move 15: what does the bishop do at b5? It just forces black to play good moves: ...c6, ...d5, gaining more control of the center.

18 Qd2? gives up the pair of bishops. A bishop is generally worth more than a knight. So you had to retreat Bh6: 18 Be3.

21 Ne2? is wrong: the knight stands well on c3, guarding central square e4. You had to contest the e-file by playing a rook to e1.

23 f3? weakens the black squares around your king. Especially since you have traded away your dark square bishop that hurts you.

32 Bxd5+? misses the stronger 32 Nxd5, when you threaten a discovered check.

Thanks tpr too!
Human explanations is much clearer than the "computer" "??" or "?" Will try to apply your advice for future games.

There seems quite a lot of concepts to be grasped in chess.
Piece activity or position (move 12 / 21), Bishops vs Knights (move 18), Weakened squares (move 23), Tactics...
I read about them in books but applying them in games is quite another story. Does it come with practice ?

It comes from practice and from analysing your lost games. It is also important to think long enough about your moves, especially at the beginning. Look at your white bishop: 3 Bc4 7 Be2 12 Bd3 15 Bb5 16 Ba4 17 Bb3. So one out of 3 of your moves were with the bishop. After that it did not do much except the erroneous 32 Bxd5+. The pawns at c6 and d5 held it at bay. If you wanted the bishop at b3, then you could have played it there on move 7. Every move should have a long term purpose. Many moves have also long term consequences. The final 38...Ng3# follows from the weakening 23 f3, which follows from the wrong 21 Ne2.

A suggestion I forgot to mention: although finding mistakes in a lost game is easier, you should still analyze your wins and draws to find better moves/plans.

with the opening explorer u can see what other human players played in your position, u should definitely experiment with it.
opening explorer goes up to 25 moves.