Life In 19x19

A game to say goodbye to April and welcome May. Black got a liability on the board and I took full advantage of it. I feel like Kato the Killer for once. Given my weakness at attacking, this game is a welcome change, hope I can improve on it even more!

Already reviewed with Leela. Additional comments are very welcome!

I especially felt a bit cramped in the opening. Even though Leela has the game about level for a while, I felt at the disadvantage. Can a neutral observer give his take on the opening?

You don't need to play 96 to stop black from connecting.

Instead : , I would play something like L9 or M9. Then black need to make an eye in sente in the middle (*) or escape . That's unlikely

(*) : the eye on the side is not a real eye at the moment, you can destroy it

Tryss wrote:You don't need to play 96 to stop black from connecting.

Instead : , I would play something like L9 or M9. Then black need to make an eye in sente in the middle (*) or escape . That's unlikely

(*) : the eye on the side is not a real eye at the moment, you can destroy it

That's true. However I was willing to give black that one eye, because I knew there was no second eye for him unless I screwed up.

Maybe I don't need to play 96, you're right. However it's one of Leela's top 3 choices so it's not a bad move either, I think

Your playing level is now on par with your analysis level. Time to improve the latter
I corrected a few of the variations you included. If you compare mine to yours, you' ll see a few spots where your thinking can improve. It's mostly about "basic technique - level 2". If falsifying the eye by taking 2 corners, or denying a table shape, is good enough, there's no need for original moves for which you need to calculate more. Similar for spotting the "chapel" virtual group.

The positional judgment at 29 is not easy for me either. I'd be interested in what others think.

Ian Butler wrote:That's true. However I was willing to give black that one eye, because I knew there was no second eye for him unless I screwed up.

Well...this is still no reason to let him have an eye. Several thing come to mind:
- if you were sure that you could kill, why not play elsewhere then?
- in the game black finished gote on the side but what if he first plays in the center and tries to get a second eye by threatening to break out? Are you confident you can deny him an eye in the center in sente (I wouldn't even if it does seem hard for black). If black gets an eye (or a ko for an eye) in the center, he can then go back to the side and live thanks to 96
- I would interpret the fact that you played L9 to ensure your kill as showing that you made the same mistake as your opponent, i. e. not taking into account the possibility to play in the center first before settling things on the side.

Not trying too sound harsh (sorry if it comes this way), just trying to highlight how not removing the eye on the side with 96 is far more dangerous than you appear to think.

Congratulations on your kill.

Modern AI opening.

Not really necessary. But I suppose Leela approves?

The right side looks big.

Yeah, I think C-05 is good.

My feeling is H-03, to protect the White group and threaten to invade. What does White do if Black plays H-03 next?

Too deep.

Bravo!

Is P-06 Leela's play? What about O-04?

Only a small mistake?

@knotwilg
Thanks for the other variations. I'm glad I'm catching up to my level of analysis (and let's be frank, it's mostly because I don't blitz anymore, you were right). Exactly, my level of analysis needs to go up now. I think the Level Up Series I'm going through now is good for me basic sense of technique.
What is a chapel virtual group?

About the table shape. There's something odd about it. (let's NOT call it the Ian Butler Syndrome) Ever since I started playing Go and got acquainted with the table shape, I was like "Okay, easy enough to remember". To this day, I have almost NO sense of table shapes in the game. It's not that I can't recognize the shape (obviously, that's not so hard), but it never comes to my attention during a game. It's very odd and maybe I need some specific table-training to solve that.

@Shenoute
Okay I see your point. Either surround first and then let him have one eye (unless I can prevent) or destroy the eye and drive the fight into the middle, but not let him have an eye and only then surround (risky play). Something like that?
It's just that I wasn't entirely certain of preventing the eye. In the analysis it's obvious that I'm quite safe in doing so, but during the game I was a bit more anxious and wasn't thinking as clearly. I saw some (non-existing) ways for him to take advantage of it and so I'd wanted to cut him off more than I wanted to prevent the eye.

I kept playing in the middle because I've screwed up "sure" kills before and it felt the safest option. As long as he keeps playing to save his group, I could keep playing to make it dead. As soon as he'd tenuki, I'd play elsewhere, too. Does that make sense?

and L9 was for me just the killing blow. Maybe, yes, because it was a mistake to focus on the sides first, but it was a move I was always going to play. But I can see now that the move takes urgency over the side.

Only thing I don't understand is the part of making an eye in sente? How does one make an eye in sente in the middle? Meaning that he makes an eye and I have to react to it because it threatens a capture or something?

You can be harsh, as long as you're helping me (not that you are! being harsh, I mean. )

@Bill
I didn't know that modern AI opening Also I've checked out my opponent after the game and I discovered he ALWAYS plays black and ALWAYS opens with that 2 point jump from his 3-4 stone like in the game (most of the time his other stone is 3-4, too, but this varies). I've discovered on Walther's patern search that only a handful of pro games went like this, and black lost most of the time, so potentially it's inferior, but it might throw many people off and at DDK level, those are probably the best openings.

Leela sees no harm in 10. I did it because I felt it makes my group safer and it also prevents black from surrounding me more easily.

you are correct.

If black played H-03 first, my plan was B-02 or try to break out. But it was certainly a point I had my eyes on. Though black did get there first in the game. So playing there earlier might've been better, who knows.

yeah P-06 is Leela's move, and I don't like the result that knotwilg shows too much. O-04 is also on Leela's radar. Actually now after 100 000 nodes, in the analysis screen she gives white 56% to win for O-04 and 55% to win for P-06. So she marks P-06 as the best move. (while she thinks P-06 after less nodes searched) Bill, Leela could learn something from you

Not small. Leela doubles black's chances after this. Big mistake on my part, I guess. And here I was, thinking I made a nice net

Ian Butler wrote:@Shenoute
Okay I see your point. Either surround first and then let him have one eye (unless I can prevent) or destroy the eye and drive the fight into the middle, but not let him have an eye and only then surround (risky play). Something like that?

Yes, kind of. In my mind it's also related to a bigger picture than the position in this game, it's about being logical and playing moves that do something: if it's already dead then tenuki is better, if it's not dead then adding a move is correct, just make sure that the move is actually doing something. Adding a move that doesn't do anything to help secure the kill is...illogical. I insist(ed) on this, maybe too much (and I hope this doesn't make you think you played a bad game) because I realized late in my "career" as a go player that working on these kind of "psychological" mistakes (i. e. mistakes that are not related to lack of theoretical knowdledge) is quite important.

Ian Butler wrote: Only thing I don't understand is the part of making an eye in sente? How does one make an eye in sente in the middle? Meaning that he makes an eye and I have to react to it because it threatens a capture or something?

Sorry if this wasn't clear. By "in sente" I meant that by playing moves in the area of M9, black can threaten to escape. If you try and prevent him from doing so, he'll manage to play several moves that may create an eye there. The "in sente" part would come from the threat of escaping.
I added one some variations to show you what I have in mind. Of course I'm not saying it would work! It's just to give you an idea of the kind of potential there is for black in the middle and why white should still use 96 in an efficient way (which he should do even without this potential anyway ).

Ian Butler wrote: If black played H-03 first, my plan was B-02 or try to break out. But it was certainly a point I had my eyes on. Though black did get there first in the game. So playing there earlier might've been better, who knows.

If B H-03, W B-02, B D-02. Not so easy for White now.

yeah P-06 is Leela's move, and I don't like the result that knotwilg shows too much.

Knotwilg's variation is beautiful.

O-04 is also on Leela's radar. Actually now after 100 000 nodes, in the analysis screen she gives white 56% to win for O-04 and 55% to win for P-06. So she marks P-06 as the best move. (while she thinks P-06 after less nodes searched) Bill, Leela could learn something from you

Protoplasm power!

Not small. Leela doubles black's chances after this. Big mistake on my part, I guess. And here I was, thinking I made a nice net

Unless you see a better line of play, containing your opponent's weak group is almost always one of the best plays.

I agree with H 03. Thx bill