A complicated finish

[jts]

I won this game, and I think that (with a few glaring exceptions) I did okay in the middle part of the game, after B started a somewhat unreasonable invasion. I welcome comments on any part of the game, or even a full review, but three things are really bugging me:

(i) The upper left corner. What's the best way to invade it?
(ii) The end of the game, if B connects, can W seal B in without dying himself? (I'm assuming that if W plays L17 and let's B out he can survive, but I could be wrong about that too. All in all W was taking a dumb risk by starting that ko, I think.)
(iii) At 124, I feel like either d10 or e11 should capture (or at the very least, should capture if B responds to H6 by connecting), but I can't make it work. As things developed, the left side could have become hard for me had B not eliminated one his own liberties.

[delete]

Thanks!

Edit: Ah, gosh, I thought I had already edited this post. A big thank you to everyone who commented. (I'm not posting new replies until I stop being pleased by having 1111 posts )

@Loons:

You're undoubtedly right about , I can't decide whether I was on autopilot or just being a joseki zombie. I definitely agree that I was expecting black to build his own potential and force me to make an unpleasant invasion... but I didn't think that meant I should stop trying to develop my lower left corner. In the upper right, you may be right about O17. My thought process there was kind of stupid - I thought taking care of the cutting stone would be irksome for B, and it looked like I had lots of threats to connect under the right side in the event of fighting, so I thought that I would take the corner, he would set up a mutual development point around Q10, and I would take it. Perhaps I was over confident.

@Ed :

Yes, excellent point about 40-44. The way you put it down on the board makes it obvious how aji keshi that sequence is... once k4 is filled in, the m3 weakness loses a lot of it's bang.

@Mitsun:

Thanks for the feedback. With respect to :w124: : do you think that capturing B's cutting stones is enough for W here? I guess if I'm willing to play f2 in gote, and the e10 net works, I should definitely be willing to capture g5 in gote. ---- I'm not sure why I was focusing solely on the center-left group at the end. It seemed "obvious" to me during the game that if the center-left had too many liberties, the center-right had way too many liberties.

@Joaz:

Thank you for the review. On : You made me go back and look at this, because I very badly wanted to play g6; it turns out I had a double hallucination that (i) B could cut off d2 if I played g6, and (ii) d2 was actually worth cutting off. (I was visualizing e3-e2, but really what I want is e3-f2-e2-e1.) I definitely agree with G7 for , but for :w107:... hm, the more I think about it, the more palatable your G7 looks. Getting P11 sealed tightly in isn't really a problem, since the center is safe and B looks silly trying to make territory around P13.

[Loons]

A bit hard for me, but some thoughts:


6... I would have split or approached from the inside. Black is emphasising the left even more than normal.

10.. I would have played H4 instead; if/when black plays K3, you can use that to reduce the left/build the right. Otherwise white seems flat.

I am worried black's potential is much bigger than white's. Both black and white seem to persistently believe the reverse is true.

20.. Would R18 have been better for your result?

22.. I think I would play O17.

26.. Seems too close to thickness? If I were to play this, anyway it would be R7 or R8.

Nice kill.

124. Yeah, seems badly timed. I would play some more forcing moves around F9 before living.


To invade top left enclosure, C16 seems likely to live inside.

I don't see a good way for black around the end.

[EdLee]

Just the broken shape at :

[mitsun]

In answer to your specific questions --

At move 124, W can capture the two B cutting stones directly. If B saves them, then he is short on liberties, and the tight geta at E10 captures everything. However, the way W played was also fine. After capturing the large B center group, W has an easy win as long as his two slightly weak groups survive.

Killing the B left side was probably fun, but irrelevant to the game. I would have looked for a way to sacrifice the W left side stones to strengthen the center group and guarantee a win.

Why even contemplate invading the upper left corner? The only way W can lose is to die in the center, and almost the only way that can happen is for W to get distracted by another fight leading to a double attack. Playing D16 would never cross my mind at this point; I would probably play H12, more or less asking B to resign.

By the end of the game, W has gotten into trouble, but it looks like he still wins. If B connects (J13) and W connects (H15), then W has 6 outside liberties (K17 has three, since B has no way to stop W from getting that many) and 2 inside liberties. The left center B group has 6 outside liberties and an eye, so B wins that capturing race. However, the right center B group has only 7 liberties, so it looks like W can capture that group to save the center. (Once W starts filling liberties, B can make an eye, but that does not improve his liberty count.)

[Joaz Banbeck]

12: You are looking unbalanced: a bunch of low stones near the lower side and one lone stone slowly getting outnumbered on the upper side. I'd play something like N16. ( I have no clue what joseki is here )

14-25: My overall feeling is that this can't be good for you. You have at most 10 points in your corner, you are almost completely locked in ( ok, you can squirm down the right side, but that is a net loss ) and he has influence facing his high stones over in the upper left.
I'm not sure what the correct play was, but this can't have been it.

26: Might the top side be bigger?

28-38: Nice flow.

40: Locally, he may be too strong for this. You are starting to look thin. I might play H3. ( Make fist, then strike. )
Note that he has to patch M3 sometime.

42: Atari is not the answer to everything. You are too thin after he answers. Besides, you force him to fill a weak spot of his.

46: Good fighting spirit!

52: G6!!

60: Nice kill. Remind me never to give you a handicap again.

104: Do you really need this? If you count liberties, I think you don't. G7 solves a lot of life issues for you. You still have S12 if he tries to fight.

106: G7 still may be better. You sac the lower left, but gain sente. You are solid everyplace else, so you invade left and top for an easy win.

124: This would have never become an issue if you had played the geta back at move 52. I know it sounds like I'm not addressing your question, but it is far better to work on not getting into ugly situations like this than to work on surviving them.
You bypass multiple opportunities to connect this group, then ask for advice on how to save it once it has become isolated.

144: You are in a liberties race. Lock him in and start killing his liberties. A7 seems to work.

170-174 etc: Count. You are way ahead. Don't complicate. Just keep reducing with K17. Then reduce westward around G14. So what if he gets 35 points in the upper left? You have close to 100 points.
Invasions entail risks, whereas reductions are safer. Invasions may gain more sometimes, but you don't need more. Play safe and stroll to an easy win.