Which is bigger, A or B?

Bill Spight · Post by **Bill Spight** » Thu Mar 16, 2017 4:48 pm

skydyr wrote:A few years ago at the US go congress, a friend and I were talking to Maeda sensei, and he told us that, apart from the endgame, sente is bad. Now, granted he was addressing this to a 5d player so it shouldn't be taken entirely literally, but his reasoning was basically that sente gains nothing since it's your privilege anyways, and that gote moves are the ones that confer real gains.

Well, that is true in the endgame, as well. So perhaps there is more to Maeda's thinking than got across.

OC, the best thing is to make gote plays with sente.

That happens a lot at the kyu level.

Drew · Post by **Drew** » Fri Mar 17, 2017 12:27 am

BlindGroup wrote:EDIT: Drew, apologies if this seems obvious. I was mislead by your "infant" rank . Looking at your other posts, I'm guessing my comments are not useful. However, I'm leaving them up in the hopes that they might be useful for newer players.

No apology necessary. You did an excellent job articulating my thoughts when I could not and I appreciate that!

My sincere thanks to everyone who has so enthusiastically participated in this thread. I hemmed and hawed over posting this question because it seemed so basic, but while I was able to see - once prompted - that there was something at B after mechanically playing A during the game, I was unable to conceptualize a useful implementation, so post it I did.

Thankfully my little question attracted just about every 19x19 luminary around for an energetic chat! :bow:

John Fairbairn · Post by **John Fairbairn** » Fri Mar 17, 2017 4:27 am

A few years ago at the US go congress, a friend and I were talking to Maeda sensei, and he told us that, apart from the endgame, sente is bad. Now, granted he was addressing this to a 5d player so it shouldn't be taken entirely literally, but his reasoning was basically that sente gains nothing since it's your privilege anyways, and that gote moves are the ones that confer real gains.

Yes. And one of the real gains of gote moves (by which we should really understand honte moves and the like) is that they leave behind sente moves - usually more than one. This is the basis of the classic book on "Gote no sente" by Sekiyama Riichi. This in turn is based on kendo, and it is worth viewing through that prism. If we try to visualise ourselves in a sword fight where the opponent comes rushing at us, I suspect most of us are likely to imagine ourselves deftly retreating (gote) rather than rushing back at him or running away (tenuki). That's because we like to see ourselves as not just brave but clever, so we would take time to think and imagine ourselves doing the serenely clever stuff. Our minds are more important than our swords.

But stones in go are not as sharp as swords and that dulls our responses. In go, amateurs tend to think the clever stuff is to resist the opponent's sente with kiai or to tenuki. That's often just grandstanding. Strong players will rather just accept gote and wait for the chance to strike back later on.

That's fine as far as it goes, but a truly strong player also has the ability to assume a gote position even when not faced with an immediate threat. In simplest terms in go this translates into making bases and honte.

Another point to mull over is that kiai is badly translated as "fighting spirit." This tends to conjure up Rambo-esque behaviour in western amateurs. The term just means matching your ki (qi) with your opponent's. That can cover fighting back immediately, of course, but it can also mean taking up a gote position so that your yin matches his yang. But in either case the point to be alert to is that this is not a static match. You are supposed to be trying to have the momentum move in your favour. That may mean tempting the sente opponent to overreach (in go that may include getting him to lose his future options or fill in liberties by playing his sentes too early). It may instead mean coiling up like a spring in a gote position. In go, as Sekiyama teaches, the important thing to take away about gote is to play it so that it likewise represents a coiled spring. If you play a gote that has no spring in it, you have possibly made an inefficient move somewhere along the line, and it is always worth tracing back to see where. It's surprisingly common to see that the inefficient backtracked move turns out to have been a sente when a gote would have been better.

(To forestall the usual questions, I don't think Sekiyama's book is in print. I think my copy is from the 1940s.)

dhu163 · Post by **dhu163** » Fri Mar 17, 2017 6:56 am

Kirby wrote:
It is a good point that professionals, compared to amateurs, may very well be better at exploiting aji, and also in showing that positions that are seen to be thick are not actually thick.

For this reason, yes, we must question our judgment as amateurs when we think something is thick. We should question our judgment when we think that there is little aji to exploit in a position. After all, we could be wrong- we are only amateurs.

That being said, this does not mean that our judgment is always wrong. Just because pros are good at exploiting aji, it does not mean that we always underestimate the aji that's present.

We cannot base decisions off of this.

...

I like this. We shouldn't disregard our own judgements when seeing stronger players' ideas.

The subject problem is a nice one. I've thought about it for several minutes, flickering from one side to the other. It seems very difficult to choose - taking A leaves extra aji in black's position with the dead W stone, and both A and B help with the aji with the dead black stone.

I've settled towards taking A myself - when black plays in the centre, i think there is a good chance W will get sente for the good point at the bottom (enclose corner + reduce the area) or some other big place (top left). A=20 points + life and death just feels too big. However note that taking A leaves black a nice ko threat.

___

@uberdude I don't see any reason not to kick compared to capturing directly in this position. As you say, B will not respond, but shorting the liberties is important, especially if you aim to crosscut. And yes, tenuki originally seems reasonable.

And your attach crosscut for B looks like a good idea, possibly overplay against the W wall, but at the point where you are worried for B, actually B should double hane, and I think B is fine.

@John Fairbain In your Chisaku example, I think it is more that the 3-3 in this position is just locally not such a powerful move, as B already has P17 and W already has R15. So if W took it , B could just hane and W would have to retreat again. Also, the N17 invasion remains, hurting both B groups, and N16 may well be bigger than R17. The value of R17 here is primarily in attacking W, not points, so if W is safe in the centre, especially with the N17 weakness, W isn't unhappy. (of course W isn't all that safe, so B's play is understandable) It is possible that B was worried W might get sente to invade around M16, but that would just be a fight, and P12 is the biggest point, so I think I agree with Shusai here.

dfan · Post by **dfan** » Fri Mar 17, 2017 6:44 pm

I had an experience like this in a tournament game last year.

Click Here To Show Diagram Code: [go]$$ $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . O O 1 . . . . . . . . . . . X X O . | $$ | a X O X . X . . . . . . . . X . X O . | $$ | . O X , . . . . . , . X . . . X O . . | $$ | . O X . . . . . . . . . . . . X O O . | $$ | . X X X . . . . . . . . . O X O X X . | $$ | . . O X . . . . . . . . . . O O O . . | $$ | . . O O X . . . . . . . . . . . . . . | $$ | . . O X b . . . . . . . . . . . . . . | $$ | . . O , . . . . . , . . . . . , . . . | $$ | . . . O . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . X . X . X . . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . O . O . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

I was White, and Black just played :b1:

, which we both assumed was sente, because of the threat of a. But Crazy Stone strongly prefers for White to give the group up and tenuki (e.g., at b), because Black has not only spent two moves to kill the upper left group, they're two moves that in retrospect are deep inside his own territory. My opponent and I were both very surprised afterwards to hear its opinion.

Uberdude · Post by **Uberdude** » Fri Mar 17, 2017 11:02 pm

dfan, how did you live? I hope it wasn't talking the stone but hane on 1st line under :b1:

as that's at least sente to sneak in the top side a bit next (which might not be sente on this board now but at least you have a follow up).

Kirby · Post by **Kirby** » Fri Mar 17, 2017 11:41 pm

Uberdude wrote:dfan, how did you live? I hope it wasn't talking the stone but hane on 1st line under as that's at least sente to sneak in the top side a bit next (which might not be sente on this board now but at least you have a follow up).

Depending on the ko threat situation (which is difficult for me to assess), the hane on the 1st line under :b1:

could theoretically be answered with a throw-in to the left of the hane, and if white captures, black could play to the right of the hane. White could then cut and cause a ko.

The ko is a little risky for black since white can get into the black territory on top, but if white loses the ko, white has to come back and live anyway - filling is not an option, since black can then kill.

A bit of a risky strategy for black, I admit.

Uberdude · Post by **Uberdude** » Sat Mar 18, 2017 12:06 am

Kirby wrote: the hane on the 1st line under could theoretically be answered with a throw-in to the left of the hane, and if white captures, black could play to the right of the hane. White could then cut and cause a ko.

White would probably tenuki as he's lived in sente, later Black would take the ko threatening to kill and white answers, and even later in endgame white retakes ko and black connects to avoid ko. So it almost ends up the same as no throw in.

dfan · Post by **dfan** » Sat Mar 18, 2017 5:46 am

Uberdude wrote:dfan, how did you live? I hope it wasn't talking the stone but hane on 1st line under as that's at least sente to sneak in the top side a bit next (which might not be sente on this board now but at least you have a follow up).

I took the stone

Kirby · Post by **Kirby** » Sat Mar 18, 2017 8:27 am

Uberdude wrote:
Kirby wrote: the hane on the 1st line under could theoretically be answered with a throw-in to the left of the hane, and if white captures, black could play to the right of the hane. White could then cut and cause a ko.

White would probably tenuki as he's lived in sente, later Black would take the ko threatening to kill and white answers, and even later in endgame white retakes ko and black connects to avoid ko. So it almost ends up the same as no throw in.

True, White gets a sente move if he tenukis, compared to just capturing. But in this case, black captures the original hane in sente, too.

It's probably better than just capturing to begin with like you suggest, but it seems a little weird to allow your stone to be captured in sente. Why not just descend directly, threatening endgame on the top later?

Click Here To Show Diagram Code: [go]$$ $$ +---------------------------------------+ $$ | . . W . . . . . . . . . . . . . . . . | $$ | . O O X . . . . . . . . . . . X X O . | $$ | a X O X . X . . . . . . . . X . X O . | $$ | . O X , . . . . . , . X . . . X O . . | $$ | . O X . . . . . . . . . . . . X O O . | $$ | . X X X . . . . . . . . . O X O X X . | $$ | . . O X . . . . . . . . . . O O O . . | $$ | . . O O X . . . . . . . . . . . . . . | $$ | . . O X b . . . . . . . . . . . . . . | $$ | . . O , . . . . . , . . . . . , . . . | $$ | . . . O . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . X . X . X . . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . O . O . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Maybe the value of sente is big enough that black wouldn't respond? In which case, the hane ends in gote and gets two more points?

BlindGroup · Post by **BlindGroup** » Sat Mar 18, 2017 9:17 am

dfan wrote:I was White, and Black just played , which we both assumed was sente, because of the threat of a. But Crazy Stone strongly prefers for White to give the group up and tenuki (e.g., at b), because Black has not only spent two moves to kill the upper left group, they're two moves that in retrospect are deep inside his own territory. My opponent and I were both very surprised afterwards to hear its opinion.

My reaction to this was different. What do you all think? I see white with two weak groups on the left, either of which can be settled in one move. So, which to prioritize?

Prioritize top group:
1. Sizable number of points.
2. More easily killed than middle group.

Prioritize middle group:
1. Sizable number of points.
2. More strategic value than top group because top group is surrounded.
3. Black can harass this group to build territory on the bottom by playing the marked stone below. The value of the territory on the bottom could be sizable.

So, this seems to point to b as the right move.

Click Here To Show Diagram Code: [go]$$ $$ +---------------------------------------+ $$ | . . O . . . . . . . . . . . . . . . . | $$ | . O O X . . . . . . . . . . . X X O . | $$ | . X O X . X . . . . . . . . X . X O . | $$ | . O X , . . . . . , . X . . . X O . . | $$ | . O X . . . . . . . . . . . . X O O . | $$ | . X X X . . . . . . . . . O X O X X . | $$ | . . O X . . . . . . . . . . O O O . . | $$ | . . O O X . . . . . . . . . . . . . . | $$ | . . O X . . . . . . . . . . . . . . . | $$ | . . O , B . . . . , . . . . . , . . . | $$ | . . . O . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . X . X . X . . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . O . O . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

gowan · Post by **gowan** » Sat Mar 18, 2017 10:22 am

John Fairbairn wrote:
A few years ago at the US go congress, a friend and I were talking to Maeda sensei, and he told us that, apart from the endgame, sente is bad. Now, granted he was addressing this to a 5d player so it shouldn't be taken entirely literally, but his reasoning was basically that sente gains nothing since it's your privilege anyways, and that gote moves are the ones that confer real gains.

Yes. And one of the real gains of gote moves (by which we should really understand honte moves and the like) is that they leave behind sente moves - usually more than one. This is the basis of the classic book on "Gote no sente" by Sekiyama Riichi. This in turn is based on kendo, and it is worth viewing through that prism. If we try to visualise ourselves in a sword fight where the opponent comes rushing at us, I suspect most of us are likely to imagine ourselves deftly retreating (gote) rather than rushing back at him or running away (tenuki). That's because we like to see ourselves as not just brave but clever, so we would take time to think and imagine ourselves doing the serenely clever stuff. Our minds are more important than our swords.

But stones in go are not as sharp as swords and that dulls our responses. In go, amateurs tend to think the clever stuff is to resist the opponent's sente with kiai or to tenuki. That's often just grandstanding. Strong players will rather just accept gote and wait for the chance to strike back later on.

That's fine as far as it goes, but a truly strong player also has the ability to assume a gote position even when not faced with an immediate threat. In simplest terms in go this translates into making bases and honte.

Another point to mull over is that kiai is badly translated as "fighting spirit." This tends to conjure up Rambo-esque behaviour in western amateurs. The term just means matching your ki (qi) with your opponent's. That can cover fighting back immediately, of course, but it can also mean taking up a gote position so that your yin matches his yang. But in either case the point to be alert to is that this is not a static match. You are supposed to be trying to have the momentum move in your favour. That may mean tempting the sente opponent to overreach (in go that may include getting him to lose his future options or fill in liberties by playing his sentes too early). It may instead mean coiling up like a spring in a gote position. In go, as Sekiyama teaches, the important thing to take away about gote is to play it so that it likewise represents a coiled spring. If you play a gote that has no spring in it, you have possibly made an inefficient move somewhere along the line, and it is always worth tracing back to see where. It's surprisingly common to see that the inefficient backtracked move turns out to have been a sente when a gote would have been better.

(To forestall the usual questions, I don't think Sekiyama's book is in print. I think my copy is from the 1940s.)

Many amateurs think that in the endgame the correct way to play is to take all the sente points possible consecutively until they have to give sente to the opponent. However, depending on the position of course, taking gote "prematurely" allows one to get better moves later, such as tedomari, when your opponent runs out of sente.

Bill Spight · Post by **Bill Spight** » Sat Mar 18, 2017 12:06 pm

Kirby wrote:
Uberdude wrote:
Kirby wrote: the hane on the 1st line under could theoretically be answered with a throw-in to the left of the hane, and if white captures, black could play to the right of the hane. White could then cut and cause a ko.

White would probably tenuki as he's lived in sente, later Black would take the ko threatening to kill and white answers, and even later in endgame white retakes ko and black connects to avoid ko. So it almost ends up the same as no throw in.

True, White gets a sente move if he tenukis, compared to just capturing. But in this case, black captures the original hane in sente, too.

It's probably better than just capturing to begin with like you suggest, but it seems a little weird to allow your stone to be captured in sente. Why not just descend directly, threatening endgame on the top later?

$$
$$ +---------------------------------------+
$$ | . . W . . . . . . . . . . . . . . . . |
$$ | . O O X . . . . . . . . . . . X X O . |
$$ | a X O X . X . . . . . . . . X . X O . |
$$ | . O X , . . . . . , . X . . . X O . . |
$$ | . O X . . . . . . . . . . . . X O O . |
$$ | . X X X . . . . . . . . . O X O X X . |
$$ | . . O X . . . . . . . . . . O O O . . |
$$ | . . O O X . . . . . . . . . . . . . . |
$$ | . . O X b . . . . . . . . . . . . . . |
$$ | . . O , . . . . . , . . . . . , . . . |
$$ | . . . O . . . . . . . . . . . . . . . |
$$ | . . X . . . . . . . . . . . . O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . X . X . X . . . |
$$ | . . . X . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . O . O . O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ +---------------------------------------+

Click Here To Show Diagram Code
[go]$$ $$ +---------------------------------------+ $$ | . . W . . . . . . . . . . . . . . . . | $$ | . O O X . . . . . . . . . . . X X O . | $$ | a X O X . X . . . . . . . . X . X O . | $$ | . O X , . . . . . , . X . . . X O . . | $$ | . O X . . . . . . . . . . . . X O O . | $$ | . X X X . . . . . . . . . O X O X X . | $$ | . . O X . . . . . . . . . . O O O . . | $$ | . . O O X . . . . . . . . . . . . . . | $$ | . . O X b . . . . . . . . . . . . . . | $$ | . . O , . . . . . , . . . . . , . . . | $$ | . . . O . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . X . X . X . . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . O . O . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Maybe the value of sente is big enough that black wouldn't respond? In which case, the hane ends in gote and gets two more points?

This sagari is an interesting play.

If the left side becomes no man's land it is 1/2 point better than simply taking the stone. (It gains 1 pt. on the top and loses 1/2 pt. on the left.) Also:

Click Here To Show Diagram Code: [go]$$ $$ +---------------------------------------+ $$ | . . W . . . . . . . . . . . . . . . . | $$ | . O O X . . . . . . . . . . . X X O . | $$ | . X O X . X . . . . . . . . X . X O . | $$ | . O X , . . . . . , . X . . . X O . . | $$ | . O X . . . . . . . . . . . . X O O . | $$ | 1 X X X . . . . . . . . . O X O X X . | $$ | . . O X . . . . . . . . . . O O O . . | $$ | . . O O X . . . . . . . . . . . . . . | $$ | . . O X b . . . . . . . . . . . . . . | $$ | . . O , . . . . . , . . . . . , . . . | $$ | . . . O . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . X . X . X . . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . O . O . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

does not threaten to kill the corner. This may or may not matter for the fight on the left side.

That aside, the sagari is worth about 7-8 pts. less than the hane.

Click Here To Show Diagram Code: [go]$$Wc $$ +---------------------------------------+ $$ | . . . W . 1 . . . . . . . . . . . . . | $$ | . O O X . . . . . . . . . . . X X O . | $$ | a X O X . X . . . . . . . . X . X O . | $$ | . O X , . . . . . , . X . . . X O . . | $$ | . O X . . . . . . . . . . . . X O O . | $$ | . X X X . . . . . . . . . O X O X X . | $$ | . . O X . . . . . . . . . . O O O . . | $$ | . . O O X . . . . . . . . . . . . . . | $$ | . . O X b . . . . . . . . . . . . . . | $$ | . . O , . . . . . , . . . . . , . . . | $$ | . . . O . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . X . X . X . . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . O . O . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

That's because of the sente jump to :w1:

. For reference:

Click Here To Show Diagram Code: [go]$$Wc Monkey jump $$ +---------------------------------------+ $$ | . . . 3 . 1 . . . . . . . . . . . . . | $$ | . O O 2 . . . . . . . . . . . X X O . | $$ | a X O X . X . . . . . . . . X . X O . | $$ | . O X , . . . . . , . X . . . X O . . | $$ | . O X . . . . . . . . . . . . X O O . | $$ | . X X X . . . . . . . . . O X O X X . | $$ | . . O X . . . . . . . . . . O O O . . | $$ | . . O O X . . . . . . . . . . . . . . | $$ | . . O X b . . . . . . . . . . . . . . | $$ | . . O , . . . . . , . . . . . , . . . | $$ | . . . O . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . X . X . X . . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . O . O . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

We could reach the same position via an earlier monkey jump. The proverb says that the monkey jump gains 8 points.

The throw-in doesn't really help Black in this regard.

Click Here To Show Diagram Code: [go]$$Wc $$ +---------------------------------------+ $$ | . W c W . 1 . . . . . . . . . . . . . | $$ | . O O X . . . . . . . . . . . X X O . | $$ | a B O X . X . . . . . . . . X . X O . | $$ | . O X , . . . . . , . X . . . X O . . | $$ | . O X . . . . . . . . . . . . X O O . | $$ | . X X X . . . . . . . . . O X O X X . | $$ | . . O X . . . . . . . . . . O O O . . | $$ | . . O O X . . . . . . . . . . . . . . | $$ | . . O X b . . . . . . . . . . . . . . | $$ | . . O , . . . . . , . . . . . , . . . | $$ | . . . O . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . X . X . X . . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . O . O . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Yes, White will have to capture :bc:

before filling at "c", but that is really not a problem.

Click Here To Show Diagram Code: [go]$$Wc Ko $$ +---------------------------------------+ $$ | . W . W B . . . . . . . . . . . . . . | $$ | . O O X a . . . . . . . . . . X X O . | $$ | . B O X . X . . . . . . . . X . X O . | $$ | . O X , . . . . . , . X . . . X O . . | $$ | . O X . . . . . . . . . . . . X O O . | $$ | . X X X . . . . . . . . . O X O X X . | $$ | . . O X . . . . . . . . . . O O O . . | $$ | . . O O X . . . . . . . . . . . . . . | $$ | . . O X . . . . . . . . . . . . . . . | $$ | . . O , . . . . . , . . . . . , . . . | $$ | . . . O . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . X . X . X . . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . O . O . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

In this kind of position White may well not wish to start the ko with a play at "a" because the corner group is at stake. Then, as Uberdude suggests, play is likely to go this way.

Click Here To Show Diagram Code: [go]$$Bc Sente ko $$ +---------------------------------------+ $$ | . W 1 W B . . . . . . . . . . . . . . | $$ | . O O X . . . . . . . . . . . X X O . | $$ | 2 B O X . X . . . . . . . . X . X O . | $$ | . O X , . . . . . , . X . . . X O . . | $$ | . O X . . . . . . . . . . . . X O O . | $$ | . X X X . . . . . . . . . O X O X X . | $$ | . . O X . . . . . . . . . . O O O . . | $$ | . . O O X . . . . . . . . . . . . . . | $$ | . . O X b . . . . . . . . . . . . . . | $$ | . . O , . . . . . , . . . . . , . . . | $$ | . . . O . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . X . X . X . . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . O . O . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Black takes the ko in sente, as White secures the group. Later on:

Click Here To Show Diagram Code: [go]$$Wc Sente ko $$ +---------------------------------------+ $$ | . W B 1 B . . . . . . . . . . . . . . | $$ | . O O X 2 . . . . . . . . . . X X O . | $$ | W . O X . X . . . . . . . . X . X O . | $$ | . O X , . . . . . , . X . . . X O . . | $$ | . O X . . . . . . . . . . . . X O O . | $$ | . X X X . . . . . . . . . O X O X X . | $$ | . . O X . . . . . . . . . . O O O . . | $$ | . . O O X . . . . . . . . . . . . . . | $$ | . . O X b . . . . . . . . . . . . . . | $$ | . . O , . . . . . , . . . . . , . . . | $$ | . . . O . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . X . X . X . . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . O . O . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

With the White group secure, :w1:

threatens to enlarge the ko with an atari at 2. Usually there is a ko fight, but the normal result is that Black ends up protecting with :b2:

, leaving a 1/3 point ko.

By allowing White to live with sente Black loses 8-9 points in the corner. Not to mention allowing White to go back and protect his group in the center left. That would be a horrendous error. :shock:

White's hane to live gains around 17 points, more than most plays in the opening. CrazyStone does not evaluate by points, so I wonder if its reckoning was based upon the hane or a lesser way to live. Perhaps the hane gives the White group a lower probability of living in pseudo-random playouts than the capture or the sagari. In that case, the program's evaluation may be off.

dfan · Post by **dfan** » Sat Mar 18, 2017 1:48 pm

My game position was more interesting than I realized! Thanks to everyone for their comments.

Kirby · Post by **Kirby** » Sun Mar 19, 2017 12:59 pm

I'd like to clarify something regarding this comment, which I don't get.

Click Here To Show Diagram Code: [go]$$Wc $$ +---------------------------------------+ $$ | . . W x . . . . . . . . . . . . . . . | $$ | . O O X . . . . . . . . . . . X X O . | $$ | . X O X . X . . . . . . . . X . X O . | $$ | . O X , . . . . . , . X . . . X O . . | $$ | . O X . . . . . . . . . . . . X O O . | $$ | . X X X . . . . . . . . . O X O X X . | $$ | . . O X . . . . . . . . . . O O O . . | $$ | . . O O X . . . . . . . . . . . . . . | $$ | . . O X . . . . . . . . . . . . . . . | $$ | . . O , . . . . . , . . . . . , . . . | $$ | . . . O . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . X . X . X . . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . O . O . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Here, white takes gote. But a play at 'x' is white's privilege - it's sente for white, but gote for black.

So later in the game, we can assume that white will get to push:

Click Here To Show Diagram Code: [go]$$Wc $$ +---------------------------------------+ $$ | . . O 1 3 4 . . . . . . . . . . . . . | $$ | . O O X 2 . . . . . . . . . . X X O . | $$ | . X O X . X . . . . . . . . X . X O . | $$ | . O X , . . . . . , . X . . . X O . . | $$ | . O X . . . . . . . . . . . . X O O . | $$ | . X X X . . . . . . . . . O X O X X . | $$ | . . O X . . . . . . . . . . O O O . . | $$ | . . O O X . . . . . . . . . . . . . . | $$ | . . O X . . . . . . . . . . . . . . . | $$ | . . O , . . . . . , . . . . . , . . . | $$ | . . . O . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . X . X . X . . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . O . O . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

So that will be the final position.

Contrast this with the hane:

Click Here To Show Diagram Code: [go]$$Wc $$ +---------------------------------------+ $$ | . . . 1 z y . . . . . . . . . . . . . | $$ | . O O X . . . . . . . . . . . X X O . | $$ | . X O X . X . . . . . . . . X . X O . | $$ | . O X , . . . . . , . X . . . X O . . | $$ | . O X . . . . . . . . . . . . X O O . | $$ | . X X X . . . . . . . . . O X O X X . | $$ | . . O X . . . . . . . . . . O O O . . | $$ | . . O O X . . . . . . . . . . . . . . | $$ | . . O X . . . . . . . . . . . . . . . | $$ | . . O , . . . . . , . . . . . , . . . | $$ | . . . O . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . X . X . X . . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . O . O . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

is playing at 'y' white's privilege? I'm not so sure. If things play out as suggested with the throw-in variation that was provided here, this was the final variation:

Click Here To Show Diagram Code: [go]$$Wc Sente ko $$ +---------------------------------------+ $$ | . W B 1 B w . . . . . . . . . . . . . | $$ | . O O X 2 . . . . . . . . . . X X O . | $$ | W . O X . X . . . . . . . . X . X O . | $$ | . O X , . . . . . , . X . . . X O . . | $$ | . O X . . . . . . . . . . . . X O O . | $$ | . X X X . . . . . . . . . O X O X X . | $$ | . . O X . . . . . . . . . . O O O . . | $$ | . . O O X . . . . . . . . . . . . . . | $$ | . . O X b . . . . . . . . . . . . . . | $$ | . . O , . . . . . , . . . . . , . . . | $$ | . . . O . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . X . X . X . . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . O . O . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Black has one more point at 'w' in this variation, and the option of later taking the ko.

I don't understand how we can assume the jump at 'y', above, since it's not clear to me that this is black's privilege, given the throw-in variation.

So I don't completely understand the comment that the sagari is 7-8 points less than the hane. It only makes sense to me if we assume the jump at 'y', but I don't see how this will happen given the throw-in variation.

Life In 19x19

Which is bigger, A or B?

Re: Which is bigger, A or B?

Re: Which is bigger, A or B?

Re: Which is bigger, A or B?

Re: Which is bigger, A or B?

Re: Which is bigger, A or B?

Re: Which is bigger, A or B?

Re: Which is bigger, A or B?

Re: Which is bigger, A or B?

Re: Which is bigger, A or B?

Re: Which is bigger, A or B?

Re: Which is bigger, A or B?

Re: Which is bigger, A or B?

Re: Which is bigger, A or B?

Re: Which is bigger, A or B?

Re: Which is bigger, A or B?