Here's a game on KGS against a 3k. I lost badly but most of that was a bad endgame, losing a big group. Though I'm pretty sure I was losing before that too. I self-reviewed (in the sgf) and made some variations and one thing I'm wondering is if they are any good because sometimes my variations are... dubious? Plus anything else I missed would be appreciated.

Hi mafidufa,

I reviewed up to 101 (although I made one or two comments later on, as well).

Your big problem was ignoring attack and defense. As soon as there are groups of stones on the board, the life and death of those stones is the crucial issue of the game until every single group is either completely alive or completely dead. This means when there are truly weak groups on the board, you attack them in preference to doing anything else; and even when you think a group is only weakish, you prefer to play moves that will leave you stronger when the time is ripe to attack it.

This happened twice in this game: you abandoned an unusually successful attack on Q7 (and subsequently, ignored the effects your moves elsewhere would have on the strength of Q7), and then subsequently abandoned the attack on M10 to make a measly few points in a place calculated to sabotage your attack.



Edit: glancing at Joaz's comments, I remembered that I forgot to show you why it's very difficult to connect with a jump on the second line. Jumps on the second line are terribly weak, and run into all sorts of terrifying problems with cutting points and shortage of liberties. So I added a few variations showing some ideas. These are not variations that it would make any sense for W to play out, by the way: if he needed to save his group later on, he might fight a depressing ko to connect, or use the aji to make an eye.

8: I think that W missed a move here. The easiest place to make life is in the corner. He should play R3, the standard way of defending a 5-4 stone...

9: ...so when you want to put pressure on him, and he does not go for the corner, you should. Play R4. You get corner territory, and you are prepared to undermine any attempt he makes to get eyes on the side.

10: Good move by him.

21: I'm not sure. They are both good moves. If all of your moves are as good as the worst of the two, you will be SDK.
Tenuki is a third possibility. I might be inclined to concede to him that he cannot be confined, and get territory with E4 or F4. If he wants to take R4 I'd let him have it - in gote. If he wants M6, I'd let him have it because I can play L5 and I'd have a partial grip on 5th rank territory.

33: I think that your opponent had a good observation. You seem to have made a decision that you must reply locally. So you struggle to choose between a bad shoulder hit and a bad armpit hit.
His stone at F16 is not threatening anything. You do not need to answer it - or not locally, at least.
You have sente, and there is a huge side unplayed. Try something around C10.

37: I would have gone the other way with R16. If you look at just the black stones, you have a slightly stronger extension to the south than to the west.
Also, R16 splits his stones and makes them fend for themselves.

51: You don't need this. You have a liberty at R15. L8 or L9 could be fun. C10 is still slow but big.

53: He encourages you to connect your stones, and he gets one point for it. Your connection is worth way more. Just play R10 and thank him.

It's dinnertime. Maybe more later

_________________
Help make L19 more organized. Make an index: https://lifein19x19.com/viewtopic.php?f=14&t=5207

thanks jtz and joaz for the comments. Interesting with the 2nd line jumps, its something I would have played and considered myself connected. Any comments on my E17 variations (move 121)?

and see if I can make SDK this year.

A few ideas, mostly on the upper right corner