My games to analyse

[goame]

Game 12
What should I improve?

[Knotwilg]

Move 72 is a pivotal point for me. If you could play again, what would be your preference?

Here are a few questions:

1) What's the status of Black's group on the left side? Alive? Killable? ALready dead?
If killable, is it worth it spending a move to do so or are there bigger plays?

2) How's White's moyo? Is it firm? Can it be invaded? Can it be torn apart?
Where would you reinforce it, if necessary?

3) How's Black's top? How does it compare to White's moyo, roughly?
Is it a good idea invading it? Or would you reduce it?

4) Bonus: how's Black's group in the lower right corner? Same questions as in 1.

At 108 you put yourself in a snap-back which is almost as basic as putting yourself into atari. There are two possible reasons for this
- you don't know about the snapback and then you need to invest in basic tesuji
- you were playing too fast and then you need to acquire some discipline in playing more slowly

Even without the snapback, Black would win the battle after ...

the cut at 101, which is the aji that White should have fixed at 72.

I like though that you kept playing after losing that group, which shows fighting spirit. You did quite well reducing the top but from 144 you start playing normal endgame while you are 30 points behind. I would give it a last shot to penetrate the top right and then resign if that doesn't work ... IF

the lower right corner wasn't waiting to be killed, which would reduce the gap with another 15 points. And this points to improvement of life & death

[gennan]

I'd like to note that move 65 was a tsumego mistake of a type that comes up often in SDK games.

S1 is the vital point of that group. With the game move, white could kill it with S1.

This black group was left hanging by both players until move 230, when white forced black to secure his group.

[goame]

Move 72 is a pivotal point for me. If you could play again, what would be your preference?

Here are a few questions:

1) What's the status of Black's group on the left side? Alive? Killable? ALready dead?
If killable, is it worth it spending a move to do so or are there bigger plays?

2) How's White's moyo? Is it firm? Can it be invaded? Can it be torn apart?
Where would you reinforce it, if necessary?

3) How's Black's top? How does it compare to White's moyo, roughly?
Is it a good idea invading it? Or would you reduce it?

4) Bonus: how's Black's group in the lower right corner? Same questions as in 1.

Move 72: I had the feeling that the position is completely lost for black. Since a long time.
at O7 was not a good move. I should have played at D8. Maybe F6 is even better. The Black's group should be save and F6 is invading White's moyo.
1) If White plays on D8 first the Black group is dead.
2) The lower half of White's moyo is save after G7 or G9 and the upper half not. It can be invaded at J11 or H11 or H10 or G10.
3) As Black I would play at J16 sooner or later. Maybe sooner. As White I would invade at the top before it becomes bigger than White's moyo.
4) Black's group in the lower right corner is alive but White can try a Ko fight later.

[goame]

I'd like to note that move 65 was a tsumego mistake of a type that comes up often in SDK games.

S1 is the vital point of that group. With the game move, white could kill it with S1.

This black group was left hanging by both players until move 230, when white forced black to secure his group.

Yes now I remember it. Thank you too very much.

[Knotwilg]

OK you were Black. That changes everything.

[tekesta]

COMLv15k is rather weak for your level, it seems. Tends to make silly errors during play. If you can beat it more than 50% of the time, you will want to move on to an engine that can beat you at least 50% of the time. On KGS I often play (and lose) against SwissBot3 on 19x19. I estimate its strength to be around 2d, so I can learn a lot from it. GadgetoBot 6k is a good choice of you want a rival that isn't crazy strong, but still capable of kicking your butt. GadgetoBot also gives handicap stones. If you can beat GadgetoBot easily, SwissBot3 can give you plenty of interesting games.

Though, I recommend playing against human rivals whenever possible. They don't always play crazy AI-like moves that can be difficult for DDK or low-level SDK players to counter effectively.