Today, I tried to teach a just-read-the rules beginner from Israel (maybe another Hikaru no Go kid) the concept of eyes (and the concept of false eye) with some beginner level life and death problems on the KGS teaching ladder for an hour. It was tough ... much more tough work than I expected.
I might not do that again, never again. It was hard job for me.
But!!! Based on today's experience, I got some rough idea about how to teach 30k.
- If he or she has read an introductory go tutorial of Japanese rules, make them learn Chinese rules (Area counting). (Japanese counting might be risky. It can make a beginner have the wrong idea that Go is a game of territory.) The AGA's video tutorial may be good enough for it.
- The concept of atari
- From eyes, over false eye to making two eyes.
- Connection and cut.
- Two basic tactical skills to capture the opponent's cutting stone: net and ladder.
- Basics of capturing race to train them to keep track of liberties of a group of stones.
- Basic skills to make eyes (a base) - corner, 3th line vs 4th line in corner and side.
- Trade-off between speed of development and stability of stones.
After this? no idea what to do ...
Anyway... It's much more important to teach how to make a strong local position (with more liberties, and eyes) in a local fight than how to surronud empty area to make territory. After all, Go is a game of ''power of stones''.
- territory moyo.
- secured territory
- secured base (eyes)
In my view, the third line is the line of secured eyes is a more correct translation.
The following is a discussion that I had with a 13k student yesterday. I've posted most of it using copy-and-paste, with minor edits. I hope it will be helpful for some ddk-level players.
Belldandy [13k]: hmm... general question...
minue622 [7d]: ok.
Belldandy [13k]: let's say here, if white doesn't connect at a, but tenukis, and the tenuki isn't urgent... should I try to disconnect at a?
minue622 [7d]: yes.
Belldandy [13k]: how do I do that? G16 G17 G15?
minue622 [7d]: G16.
After black plays a, cutting white, both the upper and left- side white stones become weak (they don't have enough eyes). So the cutting move a is urgent.
minue622 [7d]: G16. After black plays (cutting white), both the upper and the left-side white stones become weak (they don't have enough eyes). So the cutting move is urgent.
OK, check the stability of each white stone. The first thing to check is whether each of them has stable eyes. The upper white group has about 0-2 points of eyes. This is not enough to make it secure. On the left side, white has almost no eyes. You may wonder how many eye points are necessary to make these stones safe.
Belldandy [13k]: 6 at least?
minue622 [7d]: yes. for example, lets look at this basic joseki. As you know, the joseki ends with white L3.
The white L3 move is the end of the local joseki because by L3 (marked), both players have enough eye space. Each of them has roughly 6-7 pts of eyes and also has advancing routes to the side and center.
minue622 [7d]: Another example.
The move is the end of a local sequence, because by , both groups have enough eye space to be safe.
Each of these groups has two advancing routes to the center and the side. By having these advancing routes, they can increase their liberties, escape, and connect to friendly stones.
6-7 points of eye space, plus some routes to a side and the center, meet the conditions for a group's safety. This is a general guide that you can use to judge the safety of stones.
Belldandy [13k]: I remember some dan once said that if you have two weak groups, at least one is bound to die. I barely ever manage to kill anything.
minue622 [7d]: You don't need to kill anything.
Killing is just one of many possible ways to profit by attacking. In fact, killing is not the main purpose of making your opponent's stones weak. Killing the opponent's stones is very hard.
Simply think this way: To kill just one stone, you need to invest at least four of your stones.
We make the opponent's stones weak because it is a very effective way to drive the opponent's stones into a narrow area.
minue622 [7d]: OK, lets start from the fundamentals.
Very simply speaking, stones can do jobs of two kinds.
1. A stone can surround empty space, like the black stones in the bottom right corner.
2. Stones can be used to surround the opponent's stones, weak stones, like the center black stones.
Generally speaking, surrounding the opponent's weak stones is more important (urgent) than surrounding empty space.
Here is an example.
Assume this position occurs in a four stone handicap game.
As I've said, stones can do jobs of two kinds.
First, the three marked black stones can be used to surround the right-side empty area, to make territory directly.
This is an example of such a play to surround an empty area in order to make territory directly.
Here, black tried to use his stones to surround the right-side empty area in order to make territory there.
Belldandy [13k]: black seems secured. So does white, sorta.
minue622 [7d]: so far ok?
Belldandy [13k]: yes.
But this play is a common mistake. Count the stones in the intitial position.
Black had three stones, and it was Black's turn to move, and White had two stones in the lower-right area. Black has more stones than White, so Black has a stronger position and the white stones are weaker than the black ones.
This is the right way for b to play. The sequence itself may be hard to understand fully at your present rank, but see and follow the main flow of the stones.
Black tries to surround the weak white stone Q8 (marked) by playing .
White resists with .
Black blocks wWhite's advance to the center with . Black is trying to drive White into a narrow area (the right side), blocking White's advance to a wide area (the center).
and are to drive the white stones in the direction of the corner (narrow area), blocking their route to the side.
Belldandy [13k]: this looks great for black.
minue622 [7d]: now, black surrounded the entire white group, letting it live small. In other words. black drove white into a narrow area (the right side), blocking all of White's advancing routes.
Why this is good for black?
Suppose that black could do the same thing in all the other areas of the board, as in the left-side example. Then the rest of the board would be Black's territory automatically, without Black's having to make any moves to surround empty space in order to make territory directly.
If the opponent's stones are weak and thin, it is much easier to do this job.
We see that white is weak and thin on the left side. So we try to attack (by playing ).
White escapes to the center, and Black keeps putting pressure on the white stones and plays .
Black has made nice potential territory on the bottom by attacking.
We can think of Go as a game of driving the opponent's stones into a narrow area.
For you to drive the opponent's stones into a narrow area, those stones should be weak. (If they are strong, safe stones, your opponent can simply ignore your attack.) This is why making eyes for your stones and taking away your opponent's eyes are so important.
Assume this position, where white has a very solid and thick eye shape.
Now, even if Vlack tries to attack, White just ignores Black's threats, advancing to a wide area (the bottom side).
So, in Go, we hate to see our opponent get thick and strong positions and stable eyes (a base). If possible, we never miss a chance to make the opponent's stones weak.
The cutting move is urgent.
Compared to his tesuji knowledge, quite unbalanced. Silly me to find it too late. Should have noticed it much earlier.
Anyway, the cure for it is very simple. 100-150 ddk L&D problems a week for the next two months will give him about KGS sdk L&D reading ability.
Yesterday, played three rengo games with Kleene on Cyberoro. We won all the games. (One of them we won against two 1k players, the other we won against a 3d and a 1k.)
Btw, he surprised me a little by playing an advanced-level haengma move.
is Kleene's move.
In the marked position of black and white stones, is one of the implied local follow-ups. Usually used for settling or compromising a local position when Black is in a defensive position.
White's style is crushed. Impossible for black to get a result better than this one.
How wonderful! Pipon pipon!!.
Drau has done a fun job in his blog ( http://tomdraug.blogspot.com/).
Especially, there is a cute tsumego display - Pipon Pipon. How wonderful！ See you again.
The following is a discussion with Eris about two local positions on KGS. I did a copy-and-paste and edited a little.
minue622 [7d]: w turn
erislover [10k]: well hmm
erislover [10k]: seems like many things can happen but b should hane underneath I think
erislover [10k]: but this is not good. (referring to lower right sequence) I don't know, nothing seems compelling to me (upper left sequence)
minue622 [7d]: then show your best guess
erislover [10k]: for b or for w?
minue622 [7d]: for both. Go is a game of two players. so reading is to find best sequence for both.
After seeing his two sequences,
minue622 [7d]: it's your guess? (upper left one)
erislover [10k]: no seems bad for w
Eris showed lower right sequence(-). It was wrong. So I asked (refering to upper left position):
minue622 [7d]: if u were b, where would you play?
erislover [10k]: oh b would not connect here
minue622 [7d]: u right. b would not connect. but take ponnuki
Finally, I showed the solution.
minue622 [7d]: w is thick.
erislover [10k]: b is almost totally contained but has sente to develop C10 so... I dunno
minue622 [7d]: this result (locally) favors white a little. so, in normal situation, pros don't agree about this sequence. This result is not joseki, and favors white.
But don't think that this advantage is decisive for white.
erislover [10k]: no, not decisive, but... on the other hand, hmm. white would probably make this approach fairly early since undecided corner. so, giving w a lot of influence very early.
minue622 [7d]: this course is hardly played. but when the value of the left side is abnormally big, pros take this play for black.
erislover [10k]: I can think of situations where it would be nice to have both C10 and sente
minue622 [7d]: but as i said, locally speaking, this result favors white a little, but for sure ..definitely
erislover [10k]: yes in general this result only marginally better than outright 3-3 invasion of 4-4 point.
End of First position.
Problem is black's right attack sequence on upper right white.
Watching the shape of marked white stones, obvious to see which move is best to reduce liberty of white stones. a is a shape point to prevent white to make a bamboo joint. It's just common sense. So, one should be able to spot that move instantly.
So, first move is . White plays to escape from Black's attack.
Black's Next move is . is used to surround white stones effectively as shown in follow up sequence. By forcing white to capture black stone, black surrounds White stones, getting thick center influence.
Normally, it's much more efficient to use our stone is to make it work to surround opponent's stones, driving them into narrow area than to use our stones for surrounding just mere "empty area".
Especially, if our stone is working to surround weak or unstable opp stones, one can not imagine better than that.
After -, White is still unstable (and heavy). So, Black keeps putting pressure on white playing . is aiming at next surrounding move black a. White plays to avoid Black's enclosure, trying to advance to center.
Seeing bottom side position, the white stone is moving from left(a) to right direction(b). One should try to block opponent's stone's further advance. So, hane move - is a obvious and natural move.
1. In upper right corner, (or Black a) is not the right direction.
White's corner has already a base (eyes) safely. So strong stones. can not put any threat on corner White's stability. Not good to approach to strong stones.
At most, is useful just to add territory, surounding right side emtpy area. So, quite a small move.
2. In the lower left corner, or black a is not good. The marked black stone was played to crush White's shape (reducing its liberty). Now, the marked black stone has done its job already, so it is light. Saving or connecting a light stone which has done its job already is heavy.
White will play , making eyes in side quickly. Not a good story for black...
Problem position happens when black plays tenuki against White corner approach.
He said that almost all of his 16k opponents ignore and tenuki, playing a big move elsewhere, but more than a half of 15k players know urgency of this kind of positions and reply with agastin .
It was a bit hard to believe, and I said, "drau, just 1k rank difference between 15k and 16k...so..."
But he said that strength gap between kgs 15k and 16k is much bigger than gap between 15k and 14k, or 14k and 13k. If it is really true, fun fact...
From my teaching game for kleene (KLu).
White played hane against attachment. is a obvious move here, as our common Go proverb suggests (hane against attachment)
Problem is black's reasonable middle game sequence after .
In this position, against is not good way of moving stones. Ater , white stones feel uncomfortable.
Too small white base, black's big corner. One more bad thing - black is pressing down white's head of 2 stones (,).
White is moving in too heavy way.
White should play counter hane . Now, White has one cutting weakness at a, but black also has its cutting weakness at b. One of implications of this move is a trade sequence in below diagram.
Each side captured one opp's stone. But this trade is not fair, but better for white (normally).
The value of marked black stone (which white captured by ) is bigger (in normal situations). In this result, White got thick position.
In lower right corner, white is going to tie a black tiger with rotten ropes. Connection of white stones too thin. so, can not succeed if black follows right sequence.
black plays attachment. is only move. Black can make its finishing blow, with support from its thick position (marked blacks). Follow-ups after is straight-forward and simple.
Now, white is down. can not get reasonable position in anyway as long as black doesn't make big blunders.
(If can not find right follow up variations after by yourself, look up any Go book of handicap games. Every Go book of handi game explains this pos (probably...) ).
Time to be back my teaching game.
Black should play counter hane , cutting off White's waist.
Thick connection of Marked black stones helps this to work well.
after , can fight against white in favorable position, at least in even position and gets satisfactory results.
Sequence in below diagram is just one of many expected variations.
The following are the problems that we will see and solve. Problems 1-10 are intermediate level (Kgs 5k~1d), and the rest are advanced ones.
I took three problems (problem A,B and D in advanced section) from chinese classical Go literature - "Guan ja bo" (관자보-官子譜 : korean pronunciation of the title of this book, its literal meaning is "The management of stones )
1 - Problems of intermediate level (1-10)
Problem 1) : Upper right corner - B's turn to play.
This position can happen when black plays tenuki against w's 3th line corner approach. Using the belly attachment move (too much of a hint?), B can make miai and, capture either or
Problem 2) : Lower right corner - W's turn to play.
For white, is a good move,(meak). Black played , seeing the miai of 'a' and 'b'. Now problem is how to defend against this black's attack for White.
Problem 3) : Lower left corner - B's turn to play.
This position is from Black's 5-4 move joseki. is just its overplay. Black can save its 2 stones, capturing or stones. We assume that ladders are always good for black.
Problem 4) : Upper right corner - B's turn. B can kill marked 2 white stones in corner.(This problem is a very easy one)
Problem 5) : Lower left corner - B's turn. Move 'a' can not work. So, something brillant(?) move is requried for black to kill corner white. Ko fight is the best for both .
Problem 6) : Upper right corner - B's turn.
This position can happens when white plays double 3th line approaches to black hoshi. In this variation, was a really big mistake(so_called "thank you move"). White should have not played atari move. Due to White's big mistake(), black could play tenuki().
Now, white is trying to get rid of black's base with ,.
Black's goal is to keep its corner base, killing these White's 2 stones(,).
Problem 7) : Lower left corner - W's turn to move.
This is acutally the same with problem problem(5). white should make 2 eyes in corner, preventing black's expected tesuji,
Problem 8) : Upper right corner - W's turn. Save marked its 2 white stones in corner. (Hint: Read out what will happen if white cut at 'a'.) 
Problem 9) : Lower left corner - W's turn. Save corner white. This position can happen when W invades at 3-3. Usually, is not a standard move (this implies that this black move is not good, generally...)
Anyway, Black seems to hope to kill and , playing . But For white, there is a simple way to make 2 eyes in corner, preventing B's belly attachment move.
(Hint: Actually, this is exactly same with problem (8) )
Problem 10) : Upper right corner - W's turn.
There is simple and clean(no ko fight) move for white to kill corner black. Easy problem, No hint.
2 - Problems of advanced level
Problem A) : Upper right corner - W's turn. This problem is of white's best yose move in black's corner, using aji of white's marked stones. W can not save its marked 2 stones, but can make something(?) cool in b's corner.
Problem B) : Bottom side - W's turn.
Against , is one of expected black's resisntance, but it can not work. W can get huge profit, playing brillant "meak". The goal is to kill black's 6 stones in left side, or destroy black right side territory completely, killing black 3 stones.
Problem C) : Right upper corner - B's turn.
Now, Black has just one eye. To make one more eye. Black should kill one of w's mark stones and save it. We alreay saw how to make miai by using belly attachment move in problems (1-3).
Problem D) : Bottom side - W's turn. White cut, black played atari agaisnt it. Now, Can white kill black 7 stones on right side? If so, how to?
(Hint : Actually, this problem is similar to problem (4) in intermeidate level section )
Problem E) : Upper right - W's turn to kill black.
The first thing to do is to read out why W's move 'a' can not work to kill black. Next, after finding W's right attack move, then think about black's best resistance agasint "White's right attack" and what is reasonable result for both.
(Hint : As always,,, belly attachment move is very effective to make miai)
3 - Jeong suk and the use of belly attachment move in it: Old version and comtemporary one ( Rui naiwei jeong suk)
Upper right corner sequence is showing its start. Upper left corner sequence is the old version and, lower left one is comtemporary jeong suk (so-called Rui naiwei joeong suk).
4 - Case study. From Sakata 9p's game
is one of the most famous moves of Sakata. We will discuss this position and the reasons why this 's attachment move was very effective for defending white's dragon on bottom side by examining relevant variations...
Notice - I wrote this section just for tim. So, this topic will not discussed in my lessons for other padook pupils.
1 W's turn to connect"
1) Think about what is the most efficient move to connect from 'move - a' to 'd' for white at this position.
2) Also, think about w's best way to connect to defend its weakness in lower left corner.
2 Some simple cases
This is a simple quiz for fun. How many ways(moves) is it possible for white to defend its weakness at 'a'?
More specifically, think about how many moves are there for w to defend w's weakness 'a' so that white can save its marked 2 stones after black's cutting at 'a'.
3-1 Problems (Basics)
In all of following problems, B's turn to play.
3-2 Problems (intermediate)
Problem G(Upper side) - Black can save its marked 2 stones, or kill white in corner, making miai.
Problem G(Bottom side) - Save one marked black stone, or kill white in corner.
4 More things to think about.
It's not always good for us to connect and save our stones. Sometimes, Letting opponents to cut and capture our stones is much better, sacrificing them. These 2 positions are elementary cases in which sacrificing is much more profitable for us.
we will discuss how to play agasint 's threat to cut, and examine some relavant variations.
What others say about Minue622
A strong amateur 6 dan Korean player from Seoul on KGS.
He is known by his strong endgame.
He loves go so much that he comments and shares his knowledge.
His favourite word "btw.."
The picture at his kgs account's info belongs to his neighbour. (and she is very beautiful)
A good example of sensei's style: http://files.gokgs.com/games/2005/5/7/minue622-Bigriver.sgf
(Pupils of Min Sensei, Sebosai and Besh wrote this)