How do you play the endgame?

emeraldemon · #1

Do you try to calculate values for moves? Do you look at sequences of exchanges (tedomari, etc.)? I've come to realize that I usually play endgame by very roughly guessing which move is bigger or which move is sente, going by instinct as much as anything. I'm not sure that's the best way to play.

Just to have something concrete, here's a game of mine at move 150:

Click Here To Show Diagram Code: [go]$$Bcm1 Black to play endgame $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | O O . . . . . O . O . . . X O . . . . | $$ | X O O . . O . X O . X . . X O . O . . | $$ | X X . O . . O O X X . X . . X X . O . | $$ | . . X . . . . X X O X . . . . . . . . | $$ | . . X . O . O . . O O X . . X O O . . | $$ | . X . . . . . . O . O X . . X X O . . | $$ | . X O O . . . . O . O X . . X O . O . | $$ | . . X . O . . . . O X . . . . X O . . | $$ | . . X O . . . O . O X . . . . X O . . | $$ | . . . X O . O X X O O X X . . . X X . | $$ | . . . X . . . . . O X X O X X . X O . | $$ | . . . . . . X . X O . . O O . X O . O | $$ | X . X . . . . . . O O O . . . X O O . | $$ | O X X . . . . . . . X O . . . X X O . | $$ | . O X O . . . X . X . X . . X , X O . | $$ | . O O . . O . X . . . . . X . X O X O | $$ | . . . O . O X . . . . . . . X O O X . | $$ | . . . . O X X . . . . . . . . . . O . | $$ ---------------------------------------[/go]

I ended up losing this game by 3.5 points. Of course there were other, bigger mistakes earlier, but I'd really like to know how you approach a board like this. What seems big? How accurately do you count the value of moves at this stage?

Shaddy · #2

As I'm playing and reading, I keep track separately of moves that do stuff and moves that are worth endgame points. I remember most of the moves worth endgame points, so there's not a lot of reading left to be done in the endgame, the only thing left for me to read and count usually is mutual damage and ko stuff. The rest is already counted (or I know how many points it's worth) or roughly counted and cataloged away in some approximate order. But I don't do stuff like keep track of the last big move at some temperature, and I'm bad about keeping ko threats around, and I can't solve those elaborate endgame puzzles with long corridors.

edit: I only really count in tournaments. I keep track of endgame moves all the time, because they sometimes turn into working invasions or stuff like that, but if it's not an important game I just guess how much stuff is worth.

Uberdude · #3

On OGS in close games I will calculate the size of the various yose moves (and record them in the private board notes; I used to save the sgf before that feature) to play the best endgame. Tedomari is indeed very important. I don't use miai counting as that's useless for tedomari (it's validity is based on the assumption there is a near-continuum of slightly-decreasing-valued plays which is rarely the case in practice). So my analysis is of the form "a4 = 3 pt gote, if ignored further 7 pt gote follow up which if ignored has a >15 so I can't be bothered to calculate more precisely so lets call it sente followup". I win many close OGS games in yose, such as http://www.online-go.com/games/board.php?boardID=322249.

In real-life tournaments I'm usually in overtime from the middlegame onwards so just blunder my way through to the end with guesstimating.

Toge · #4

Here's my estimate:
The top right corner is by far the most important area now, as both attacking and defending there will have big impact on territory. Area around B looks to be at competition point, but there's a risk it will become dame. There's sente moves for white at positions C but I find it hard to estimate how much preventing these is worth. It would be best to calculate the next sente after black defense at C points. C16 leaves cutting point at white position and might be particularly good. Black could probably push at point like D and get white to answer, but again from white's perspective it's hard to say how much these are worth. I don't trust my counting enough to play the kind of active yose Kageyama describes in Fundamentals.

Click Here To Show Diagram Code: [go]$$Bcm1 Black to play endgame $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | O O . . . . . O . O . . . X O . a . . | $$ | X O O . . O . X O . X . . X O . O . . | $$ | X X c O . . O O X X . X . . X X . O . | $$ | . . X . . . d X X O X . . . . . . . . | $$ | . . X . O . O . . O O X . . X O O . . | $$ | . X . . . . . . O . O X . . X X O . . | $$ | . X O O . . . . O . O X . . X O . O . | $$ | . . X . O . . . . O X . . . . X O . . | $$ | . . X O . . . O . O X . . . . X O . . | $$ | . . . X O . O X X O O X X . . . X X . | $$ | . . . X . . . . . O X X O X X . X O . | $$ | . . . . . . X . X O . . O O . X O . O | $$ | X . X . . . . . . O O O . . . X O O . | $$ | O X X . . b . . . . X O . . . X X O . | $$ | . O X O . . . X . X . X . . X , X O . | $$ | . O O . . O c X . . . . . X . X O X O | $$ | . . . O . O X . . . . . . . X O O X . | $$ | . . . . O X X . . . . . . . . . . O . | $$ ---------------------------------------[/go]

emeraldemon · #5

Could you explain a bit more about the aji in the upper right? Can black play R18?

hyperpape · #6

At my level, I find that simply focusing on keeping sente (i.e. not playing obvious gote) makes it possible to gain some advantage in the endgame. On OGS, I vary in how much time I spend, and sometimes play by my gut feeling, but also do a moderate bit of counting. In real time games, I almost always play by intuition.

I don't know that I'm good at the endgame, but I care more than most of the players I play against. But I still could be trying much more consistently than I do.

jts · #7

hyperpape wrote:

At my level, I find that simply focusing on keeping sente (i.e. not playing obvious gote) makes it possible to gain some advantage in the endgame.

The one problem with this is that if there might be an endgame kos, you want to keep all of those 1pt sente-with-30 pt follow up for use as ko threats. Also, once all of your sente moves are gone the possibility for nasty aji often disappears too.

Toge · #8

hyperpape · #9

jts wrote:

hyperpape wrote:

At my level, I find that simply focusing on keeping sente (i.e. not playing obvious gote) makes it possible to gain some advantage in the endgame.

The one problem with this is that if there might be an endgame kos, you want to keep all of those 1pt sente-with-30 pt follow up for use as ko threats. Also, once all of your sente moves are gone the possibility for nasty aji often disappears too.

Yes. Looking for valuable kos does matter. But there are still games where there are not large kos left on the board.

Anyway, my point was not that you should always play your one point sente, etc, but that avoiding unnecessary gote is already a substantial advantage.

Uberdude · **#10**

hyperpape wrote:

jts wrote:

hyperpape wrote:

At my level, I find that simply focusing on keeping sente (i.e. not playing obvious gote) makes it possible to gain some advantage in the endgame.

The one problem with this is that if there might be an endgame kos, you want to keep all of those 1pt sente-with-30 pt follow up for use as ko threats. Also, once all of your sente moves are gone the possibility for nasty aji often disappears too.

Yes. Looking for valuable kos does matter. But there are still games where there are not large kos left on the board.

Anyway, my point was not that you should always play your one point sente, etc, but that avoiding unnecessary gote is already a substantial advantage.

I have noticed that one of the mistakes many of my opponents make on OGS is allowing me to play (good) reverse sentes, i.e. they failed to play their sente yose early enough. Reverse sente and tedomari often go hand-in-hand.

tapir · **#11**

As "sente is worth nothing" you don't have to calculate them

You already count them as if played anyway. But distinguishing sente from gote moves properly is likely the most crucial ability. (And spotting and playing double sente moves.) This sounds trivial but isn't. Losing sente is much bigger than mistaking a 5 pt. gote for a 4 pt. gote - which may or may not lose a single point, but not more.

In this spirit: I wonder about R16. Isn't simply Q17 and P19 in sente better? After White defends at Q18 and R18 R16 will still be sente for Black, if White defends differently at S18 Black will instead have another sente move at Q19.

Both G8 and R12 are larger than they look imho. (Ask Bill for a miai count

)

Uberdude · **#12**

But nothing is absolutely sente. Working out how sente moves are is a key part of playing a strong endgame. If the follow up of a move if ignored is to kill a 50 point group we tend to think of that as absolute sente, but if there is another move of your opponent's whose follow up is to kill a 60 point group that is more sente. As a real example of this, consider this move from the game I posted.

Click Here To Show Diagram Code: [go]$$Wcm130 Not so simple sente $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . a . . . . . . . . . . . . . | $$ | . X X X X X O O . . . . . . . O O . . | $$ | . X O O O X X X O O O . O . O X . 1 . | $$ | O O O X O O X . X X . . . . . . X 2 . | $$ | O X X X X O X . . . . . . . . X O . . | $$ | . O O X . X O O . . O O . . . . X . . | $$ | . . . X . X . . O . O X O O W C . . . | $$ | . . O O X . X O X X O X X X B X . . . | $$ | . O . O X . O O . , X . . . . , . . . | $$ | . . . . X . O . . X . . . . . . . . . | $$ | . O X . . X . . . . . . . . . . . . . | $$ | . X O . . . . O . X . . . . . . . . . | $$ | . . O . . X X . . . . . . X . . . . . | $$ | . . . . . X . X . . . O . . X . . X X | $$ | . . . O O O X . . X . X . X O X X X O | $$ | . . . . . X O c . . X O X O . O O O O | $$ | . . . . . O . . . X O b . O O . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

I played s16 and he answered. To the casual observer you might just think this is a simple sente move. It is not. In order to play that move at that time I calculated the size of my follow up if ignored, as there are 2 other huge moves on the board, g18 and m2 (something like 12 and 10 gote iirc) as well as a pretty big reverse sente at h3. Black had made a small mistake earlier which meant this was my sente when it should not have been, he blocked p12 at p11 rather than q12. Whilst this looks better as it doesn't leave the weakness of the p11 push and cut, that aji is going nowhere, but as a downside it makes s16 more sente for me as if I pull out r14 he has a cutting point he needs to defend at q13. So what this essentially means is if he tenukis to one of those big gotes (maybe with h3 first) my follow up is sente so I can then come back and get the other of the two big gotes and hence get tedomari. If he had q12 I would spend two gotes at s16 and followup and he would get both g18 and m2. Things are also complicated black can also try to play a sente rather than gote sequence around g18.

ez4u · **#13**

Uberdude wrote:

But nothing is absolutely sente. Working out how sente moves are is a key part of playing a strong endgame. If the follow up of a move if ignored is to kill a 50 point group we tend to think of that as absolute sente, but if there is another move of your opponent's whose follow up is to kill a 60 point group that is more sente. As a real example of this, consider this move from the game I posted.

Click Here To Show Diagram Code: [go]$$Wcm130 Not so simple sente $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . a . . . . . . . . . . . . . | $$ | . X X X X X O O . . . . . . . O O . . | $$ | . X O O O X X X O O O . O . O X . 1 . | $$ | O O O X O O X . X X . . . . . . X 2 . | $$ | O X X X X O X . . . . . . . . X O . . | $$ | . O O X . X O O . . O O . . . . X . . | $$ | . . . X . X . . O . O X O O W C . . . | $$ | . . O O X . X O X X O X X X B X . . . | $$ | . O . O X . O O . , X . . . . , . . . | $$ | . . . . X . O . . X . . . . . . . . . | $$ | . O X . . X . . . . . . . . . . . . . | $$ | . X O . . . . O . X . . . . . . . . . | $$ | . . O . . X X . . . . . . X . . . . . | $$ | . . . . . X . X . . . O . . X . . X X | $$ | . . . O O O X . . X . X . X O X X X O | $$ | . . . . . X O c . . X O X O . O O O O | $$ | . . . . . O . . . X O b . O O . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

I played s16 and he answered. To the casual observer you might just think this is a simple sente move. It is not. In order to play that move at that time I calculated the size of my follow up if ignored, as there are 2 other huge moves on the board, g18 and m2 (something like 12 and 10 gote iirc) as well as a pretty big reverse sente at h3. Black had made a small mistake earlier which meant this was my sente when it should not have been, he blocked p12 at p11 rather than q12. Whilst this looks better as it doesn't leave the weakness of the p11 push and cut, that aji is going nowhere, but as a downside it makes s16 more sente for me as if I pull out r14 he has a cutting point he needs to defend at q13. So what this essentially means is if he tenukis to one of those big gotes (maybe with h3 first) my follow up is sente so I can then come back and get the other of the two big gotes and hence get tedomari. If he had q12 I would spend two gotes at s16 and followup and he would get both g18 and m2. Things are also complicated black can also try to play a sente rather than gote sequence around g18.

Sorry, I can't read it out completely myself. How would you have answered this :b31:

?

Click Here To Show Diagram Code: [go]$$Wcm130 Not so simple sente $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . a . . . . . . . . . . . . . | $$ | . X X X X X O O . . . . . . . O O . . | $$ | . X O O O X X X O O O . O . O X . 1 . | $$ | O O O X O O X . X X . . . . . . X . . | $$ | O X X X X O X . . . . . . . . X O . . | $$ | . O O X . X O O . . O O . . . . X . . | $$ | . . . X . X . . O . O X O O W C . . . | $$ | . . O O X . X O X X O X X X B X . . . | $$ | . O . O X . O O . , X . . . . , . . . | $$ | . . . . X . O . . X . . . . . . . . . | $$ | . O X . . X . . . . . . . . . . . . . | $$ | . X O . . . . O . X . . . . . . . . . | $$ | . . O . . X X . . . . . . X . . . . . | $$ | . . 2 . . X . X . . . O . . X . . X X | $$ | . . . O O O X . . X . X . X O X X X O | $$ | . . . . . X O c . . X O X O . O O O O | $$ | . . . . . O . . . X O b . O O . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

Uberdude · **#14**

ez4u wrote:

How would you have answered this :b31:

?

With language not appropriate on this forum ;-)

To be honest, I don't think I considered this move. I can kill b9 thanks to d10 (which was a funny move if you look at the game record, but I should have played it earlier than I did), and I assumed I can kill 3-3, but you are right there is some bad aji here and black may be able to find some clever sequence to put these two weaknesses together and get something to work. It seems I can answer with the hane underneath at b5 to keep everything safe but that does give black some sente reduction. Compared to the game white lost 2 points (and black spent an extra gote here in the game to to that). But white does get g7 in sente now which is useful for j6.

Click Here To Show Diagram Code: [go]$$Wcm130 Submit $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . X X X X X O O . . . . . . . O O . . | $$ | . X O O O X X X O O O . O . O X . 1 . | $$ | O O O X O O X . X X . . . . . . X . . | $$ | O X X X X O X . . . . . . . . X O . . | $$ | . O O X . X O O . . O O . . . . X . . | $$ | . . . X . X . . O . O X O O O . . . . | $$ | . . O O X . X O X X O X X X X X . . . | $$ | . O . O X . O O . , X . . . . , . . . | $$ | . 7 6 b X . O . . X . . . . . . . . . | $$ | . O X a . X . . . . . . . . . . . . . | $$ | . X O 9 0 . z O . X . . . . . . . . . | $$ | . c O 8 . X X . . . . . . X . . . . . | $$ | . 3 2 4 . X . X . . . O . . X . . X X | $$ | . . 5 O O O X . . X . X . X O X X X O | $$ | . . . . . X O . . . X O X O . O O O O | $$ | . . . . . O . . . X O . . O O . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

If white cuts and black goes into the corner it seems like a dead L group (can also get this in different order with 2nd line hane first and black into corner).

Click Here To Show Diagram Code: [go]$$Wcm130 Cut and black dies $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . X X X X X O O . . . . . . . O O . . | $$ | . X O O O X X X O O O . O . O X . 1 . | $$ | O O O X O O X . X X . . . . . . X . . | $$ | O X X X X O X . . . . . . . . X O . . | $$ | . O O X . X O O . . O O . . . . X . . | $$ | . . . X . X . . O . O X O O O . . . . | $$ | . . O O X . X O X X O X X X X X . . . | $$ | . O . O X . O O . , X . . . . , . . . | $$ | . . . . X . O . . X . . . . . . . . . | $$ | . O X . . X . . . . . . . . . . . . . | $$ | . X O . . . . O . X . . . . . . . . . | $$ | . . O . . X X . . . . . . X . . . . . | $$ | . 5 2 3 . X . X . . . O . . X . . X X | $$ | . . 4 O O O X . . X . X . X O X X X O | $$ | . . . . . X O . . . X O X O . O O O O | $$ | . . . . . O . . . X O . . O O . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

However, it seems black can do this:

Click Here To Show Diagram Code: [go]$$Wcm130 Cut and white dies $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . X X X X X O O . . . . . . . O O . . | $$ | . X O O O X X X O O O . O . O X . 1 . | $$ | O O O X O O X . X X . . . . . . X . . | $$ | O X X X X O X . . . . . . . . X O . . | $$ | . O O X . X O O . . O O . . . . X . . | $$ | . . . X . X . . O . O X O O O . . . . | $$ | . . O O X . X O X X O X X X X X . . . | $$ | . O . O X . O O . , X . . . . , . . . | $$ | . 4 6 7 X . O . . X . . . . . . . . . | $$ | . O X 5 . X . . . . . . . . . . . . . | $$ | . X O . . . . O . X . . . . . . . . . | $$ | . 8 O . . X X . . . . . . X . . . . . | $$ | . 9 2 3 . X . X . . . O . . X . . X X | $$ | . . 0 O O O X . . X . X . X O X X X O | $$ | . . . . . X O . . . X O X O . O O O O | $$ | . . . . . O . . . X O . . O O . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

Click Here To Show Diagram Code: [go]$$Wcm140 Cut and white dies part 2 (46 connects at 43) $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . X X X X X O O . . . . . . . O O . . | $$ | 6 X O O O X X X O O O . O . O X . O . | $$ | O O O X O O X . X X . . . . . . X . . | $$ | O X X X X O X . . . . . . . . X O . . | $$ | 4 O O X . X O O . . O O . . . . X . . | $$ | 5 3 2 X . X . . O . O X O O O . . . . | $$ | . 9 O O X . X O X X O X X X X X . . . | $$ | 8 O . O X . O O . , X . . . . , . . . | $$ | . X X O X . O . . X . . . . . . . . . | $$ | . O X O . X . . . . . . . . . . . . . | $$ | . X O . . . . O . X . . . . . . . . . | $$ | . X O . . X X . . . . . . X . . . . . | $$ | . O X O . X . X . . . O . . X . . X X | $$ | . 1 X O O O X . . X . X . X O X X X O | $$ | . . . . . X O . . . X O X O . O O O O | $$ | . . . . . O . . . X O . . O O . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

Click Here To Show Diagram Code: [go]$$Bcm149 Cut and white dies part 3 $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . X X X X X O O . . . . . . . O O . . | $$ | X X O O O X X X O O O . O . O X . O . | $$ | O O O X O O X . X X . . . . . . X . . | $$ | O X X X X O X . . . . . . . . X O . . | $$ | O O O X . X O O . . O O . . . . X . . | $$ | O O X X . X . . O . O X O O O . . . . | $$ | . O O O X . X O X X O X X X X X . . . | $$ | X O . O X . O O . , X . . . . , . . . | $$ | . X X O X . O . . X . . . . . . . . . | $$ | . O X O . X . . . . . . . . . . . . . | $$ | . X O 1 3 . . O . X . . . . . . . . . | $$ | . X O 2 . X X . . . . . . X . . . . . | $$ | . O X O . X . X . . . O . . X . . X X | $$ | . O X O O O X . . X . X . X O X X X O | $$ | . . . . . X O . . . X O X O . O O O O | $$ | . . . . . O . . . X O . . O O . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

Anyway, I was behind and so couldn't afford to spend a move here even if something works. I was in fact rather surprised that the game was even close as my result on the left side was rather miserable and then black effortlessly converted the centre and right into territory (as I played d10 too late I lost my flexibility to reduce the centre). Pretty soon black played c9 in sente which did help me fix some of my bad aji here. In fact the weakness here actually enticed black to play the game losing tesuji (my instinct was to play honte at e6 instead of o1 but I counted o1 was best). His clever move at 169 worked, but was simply not the largest move on the board.

Click Here To Show Diagram Code: [go]$$Bcm169 Game losing tesuji $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . X O O . . . . . . . . . . . . | $$ | . X X X X X O O . . . . . . . O O O . | $$ | X X O O O X X X O O O . O . O X X O . | $$ | O O O X O O X X X X X O . O X . X X . | $$ | O X X X X O X O . . . . O X . X O . . | $$ | O O O X . X O O . . O O . . . . X . . | $$ | . O X X . X X O O . O X O O O O X . . | $$ | . . O O X . X O X X O X X X X X . . . | $$ | . O . O X . O O O X X . . . . , . . . | $$ | . O X . X . O . . X . . . . . . . . . | $$ | . O X . X X . . . . . . . . . . . . . | $$ | . X O . O . . O . X . . . . . . . . . | $$ | . . O . . X X . O X . . . X . . . . . | $$ | . . . 1 . X . X O . . O . . X . . X X | $$ | . . . O O O X . X X . X . X O X X X O | $$ | . . . . . X O X . . X O X O . O O O O | $$ | . . . . . O O O X X O . X O O . . . . | $$ | . . . . . . . . . . . . . O . . . . . | $$ ---------------------------------------[/go]

emeraldemon · **#15**

This is very interesting, I think I'm going to try keeping endgame notes for DGS games and see if I learn something. Uberdude, do you think playing that way on OGS has improved your realtime, byo-yomi endgame?

Uberdude · **#16**

emeraldemon wrote:

Uberdude, do you think playing that way on OGS has improved your realtime, byo-yomi endgame?

No.

tapir · **#17**

Uberdude wrote:

emeraldemon wrote:

Uberdude, do you think playing that way on OGS has improved your realtime, byo-yomi endgame?

No.

This endgame seems to be a major part between 3 dan EGF and 7 dan OGS. Scary.

cyclops · **#18**

I believe it was Shaddy who explained that he made mental notes during a game of the endgame points and their values. My side question is whether it is allowed to make real notes during a real game. Because I am suffering from some memory leakage :sad:

. In the same sense I would like to write down my territorial estimates.

Uberdude · **#19**

tapir wrote:

This endgame seems to be a major part between 3 dan EGF and 7 dan OGS. Scary.

It's part, but the elimination of blunders and byo-yomi idiocy is a bigger factor.

cyclops wrote:

My side question is whether it is allowed to make real notes during a real game.

I'd consider that cheating, though I did notice one opponent in a recent tournament getting out a notebook and jotting things down (presumably his count, I too often count and then forget after counting the other colour) and I didn't complain.

Uberdude · **#20**

In case people are interested, here are my yose notes from around move 160 in http://www.online-go.com/games/board.php?boardID=338815, another game which I won in yose.

Quote:

N19 vs back down o19 = 6
T2 = 4.66
L1 = 3 sente
a7 = ?
b4 = 4.25, wb4 plus b5 connect = 7 (vs b 1 gote and a3 sentes later) for 2 gotes. 7/3 = 2.33 miai.
c14 = 5.66 vs w b15 gote. If w c14 sente then diff is 4 pts

I started precise calculations on move 116 when white peeped at o15, a clever move I had missed as if I connect p18 is severe. Earlier moves like r10 and j6 were played by feeling rather than precise calculation, though if I had done so maybe I would have realised quite how huge n18/m19 was (but I didn't think I could omit r10; I was very dissatisfied with my result on the right side). Losing p14 in sente was painful, but at as consolation it made white o12 wasted. The m18 ko made the yose very complicated, it was one of those cases where playing big gote moves elsewhere are threats because if you end the ko the opponent gets 2 big moves which were basically miai. White's s15 surprised me, I thought he might n2 which is about 9 points, but that move created many more threats for him. Trying to caclulate the size of the m18 ko was too hard for me, particularly after white had got s15 because if white wins the p18 ko my entire group can die. Maybe Bill can help ;-)

.

How do you play the endgame?

Who is online