MikeKyle v. TelegraphGo (Blind)

[TelegraphGo]

14-9
O9

Knight's move.

I give up reading this. Another thread going on right now is about ordering the variations you read, to read efficiently. That's a lot harder when you can't see the stones

I've read a hundred dumb variations and maybe 2 or 3 relevant ones. I'm definitely losing the thread here.

Regardless, this seems to be the natural move - when I read just a couple moves deep, it becomes apparent that other moves become inefficient quickly for several different reasons. One basic assessment - if I play the move one line below, then he eventually gets 16-7 in sente, and that's one corner pinched on a potential eye at 15-8. The move I'm playing appears to offer a little more flexibility. I can't find a perfectly clean way for white to continue, but I get a feeling that I'm missing something dumb. The blindfold is really becoming a serious problem here. There are some variations where it becomes a capturing race between my bottom group and his bottom group. I can't quite figure out yet whether I can snag a seki - it's a such a mess. In some ways this is an improvement to my position earlier, where I was strictly worse. If I get out of this, the territory should be looking a lot better.

Anyway, let's see what he does. I'm probably dead if he plays it correctly, but it ain't gonna be too simple (I hope).

[MikeKyle]

16(Q)-7

This doesn't seem like it could be the breakout tesuji option so I think black is trying to trigger a favourable capturing race. Alternatively I did leave my top right group alone so maybe black will harras it, get some nice endgame and then figure out if the game is playable.

Also possible of course that I'm incorrect about the placement of one or two stones, the wall is fine and black is just playing good endgame.

[TelegraphGo]

15-6
P6

Connect.

The only other option is 17-11 right away, offering a trade of my bottom right stones for his central stones. If that trade happens, then he would have some 50ish points in the bottom right, and 85-90ish overall. I would have to make around 70 points on the top to make the game playable. I think I can make around 60, so I'm on the edge of being able to keep playing, surprisingly.

But I still can't quite see how he kills me. Saving the bottom right group is bigger than giving it up. I'm going out in a blaze of glory! Here we go!

[MikeKyle]

15(P)-7

Looks more and more like the capturing race option but black seems very busy to stop my liberties increasing.

[TelegraphGo]

Bwahaha, you've revealed my hidden move! I went back to make sure, it seems I played it back on page 2.
I'm guessing this was a typo?

[MikeKyle]

Oops

No typo, visualisation failure! Just forgot that one was there. Now you mention it I remember the shape at the beginning of the game. I guess I lost it when we took that break over Christmas and didn't properly internalise it when we resumed.

My moves from 5-3 must have seemed pretty strange.

Although it's a tough game for me now, if you'll allow me to replay that failed move then I may try playing on a little longer?

[TelegraphGo]

Yeah, let's keep playing. No need to end the fun

[MikeKyle]

Good stuff. I'll have a go at getting us somewhere close to a reasonable endgame before the ghost stones attack again!

8(H)-7

Oof! Tough game!
I think maybe before I played 10-8 it's possible that the bots would say I had a lead? (they would probably be a lot more confident about dealing with the centre potential, against a stronger opponent I would expect it was comfortable enough for black)
I think 10-8 was slack and then each move since has been loosing both points and aji.
Black's bottom left group is connected through and can live anytime by extending into the corner. However since it's not 2 eyed I think I need to lean hard on it in an attempt to try to do more at once. Expecting a squeeze here but I'm not confident black won't respond differently.

[TelegraphGo]

17-11
R11

Poke.
Knight's move.

Well. I wanted to play 8-6, and capture this stone in a ladder. But I actually think I might be killing myself by doing that - when he squeezes my three stones I would lose some critical aji in the center. I'm actually using my two dead stones at 11-5, 11-6 to make my last forcing move for the second eye in some variations. So I think I need to make this exchange first, and then I think I should have enough aji on the side to make my second eye. After that I can come back to play 8-6 and save my three stones.

Of course, this is offering a trade where he can ignore me to take my center three stones. He could then continue to try to attack my bottom left group. But that just seems like it won't be too profitable for white, when my followup cut on the right is not only capturing 20ish points and securing my group but also threatening atari on his right side ponnuki.

I'm somewhat suspicious of his evaluation of this game. It doesn't feel like he's aiming for a squeeze on my three stones to go back and attack my bottom right. It feels more like he's trying to tenuki and he thinks this cut is the proper followup in the center. I don't think that that's true, but maybe it is just a technique to aim for the squeeze? Maybe I have another ghost stone floating around, whatever it was that made him play 10-8.

[MikeKyle]

16(Q)-13

Its great advice that after a mojor blunder you should stop and regroup and beware making another. My last move was a bit slack. The right hand side isn't done yet.

On the plus side black has forced me to play a move I kind of waned to play.

[TelegraphGo]

13-11
N11

Atari.
Connect.
This move was frustrating. Let me talk through my reading. Here's the threat:

What if he starts at 15-8, I block at 14-8, and he protects at 16-9? I'd have to protect the cut with a move at 13-7, but then he can try to cut me off from my middle stones by playing at 13-10. I really want to be able to make two eyes or connect from that position. If I push at 13-11, he might push at 14-10, letting me continue at 14-11, and then connect back his two stone with something like 13-9. I'd have two cuts to worry about, 15-9 and 15-11. 15-9 pretty much has to be defended instantly with by playing at 15-9, which happens to be atari. Then he can cut at 15-11. Remember that this cut is supported by his old knight's move at 15-12, so I can't ladder it or anything. It looks like the best variation for both sides is just for me to capture the stone, and him to make a ko for my two eyes. So unless I can find better moves to play, I'm currently only alive in ko.

The move I was incorrectly eyeing up behind my last comment is 17-9 first in response to 13-10. The problem is that he can play 17-8 and demand that we still play a ko.

But why am I only now suddenly worried about this, when before I couldn't find a way for him to kill me? Well, what if I play 12-9 in response to his 13-10, trying to break out on the bottom? He pushes at 13-9, I turn with 13-8, and he saves his stones to stop my from connecting with either 13-11 or 12-10 (or maybe 11-9). Then I should atari his 12-8 stone at 12-7, threatening to connect to my three stones. The key here is that I can atari his two stones at 11-6, 12-6 because I still have my two dead stones at 11-5, and 12-5. So this way works! This is the variation in response to his 16-9 tiger's mouth attack that I had prepared assuming he was trying to kill me outright.

The problem is that it stops working when he squeezes my three stones in the center, because he gets the move at 11-7 in sente.

Do I think that he sees that the cut on the left that he's playing is currently threatening a ko on my bottom right group? Frankly, no. I'm not sure that my reading is correct either, since I could easily have ghosted a stone in one of my variations by accident. But I should play as if I'm correct, and he knows too. It's incredibly frustrating to have to admit that your opponent just made a really good move, especially when you think it probably wasn't completely read out. It's just how it goes sometimes, I suppose.

So how to defend? It's really awkward, since no move nearby is close to as big as the left. I guess I'm playing the central atari, since the take will be my sente and I'll have a lot less central aji to worry about. But the game is definitely good for him again.

[MikeKyle]

16(Q)-9

Yep. I was just looking at how that elephant eye looked a bit shifty for black. I think my last move fixed a weakness to make it work so black had to do something like that to fix up.
(not a great time to write up example sequences now but may add later)

It's tempting to play again in the centre so that I'm safe and maybe blacks lower left group can be mildly poked about but I have the opportunity to take more than my earlier counted 24 points here and to fix up so I can't be harrased further.

[TelegraphGo]

15-8
P8

Push.
Connect.

Between this and 14-8, I prefer this one, since it exposes an extra weakness and is good local shape. 15-10 itself is not heavy and I shouldn't treat it as if it is (not that he can push up immediately anyways).

I guess this might be considered a good exchange for white, but he really ought to get back to his 8-7 stone. It's not as good to make exchanges when there's a threat looming, because your exchanges lose the ability to threaten to take a big gote. Also, I just defended this area, so in general I'm thicker than I used to be here - if he wanted Q9 he should have played it before I defended. Even up into the mid-dans people make this mistake, where their eye follows their opponents last move and they unintentionally justify it. When your opponent adequately defends themselves, you should generally look to play elsewhere, and not follow it up.

[MikeKyle]

16(Q)-10

I should have taken that 15-8 point before pushing through here. I'm not paying enough attention to this game at the moment. I guess that might have harmed my centre group slightly though

[TelegraphGo]

8-6
H6

Cut.
Double atari.

Um. I just said he couldn't take gote. Well, looks like he gets the right side, and I get everything I wanted. I'll take it! I briefly considered making an exchange on the right, but nothing looks good enough to do before the important thing.
I double checked the ladder just to make sure I wasn't hallucinating

I've got no idea what he's going to do next.