Dieter: move of discussion on stone counting method (2006-08-12 09:02) [#1306]

(copy of the original page of Stone counting as teaching method)

Benjamin Teuber:

As far as I know, most teachers in asia just explain how to capture at the beginning. But they don't use Atari-Go. Instead, they just continue teaching capturing methods until kids are able to understand simple life & death and then territory. I don't like this so much, as the kids can't play themselves in the beginning, so I thought a while about alternatives to Atari-Go.

Then, a few months ago, I was very astonished when I recognized by myself, that Go actually is not about territory on a basic level, but actually about providing life for as many stones as possible (recently I heared that Ing stated this before).

That made me think of a different game for beginners to play directly after learning to capture:

  Real Go Using Stone Counting

That means that both players try to get more stones on the board than the opponent.
Rules of it are very simple:
Instead of playing a stone, players are allowed to pass. After both players passed in a row, just stones are counted and the one with more wins (if you dislike the idea of not counting the two eyes each player should leave, you could also point out that stones plus they liberties are counted, which is equivalent to normal chinese rules).
After hearing this, of course people start exactly as in Atari-Go, as it's obviously good to reduce the number of enemy stones on the board. But as they don't have capturing as official goal, I could imagine that the idea of "must capture/must escape" is not burned into the beginners mind as much.
In the first game, the player with less territory (just gained by capturing) will probably put himself into self-atari first and therefore lose big. When the teacher points out that it's better to pass then to put yourself into self-atari, they can figure out eyes and life&death alone.
After this, you can introduce territory as a new strategy without changing even one rule. Just tell them that it might be good to build yourself a kind of "house", where each intruder will be caught and therefore many own stones can be put later.

In my opinion, it should work out very nice, as it is as easy, encouraging and straightforward as Atari-Go without being a game different from Go and hiding something. Also, as players continue until pass, life&death will become self-explaining by this.

What do you think about this?

Mef: This sounds pretty interesting because it doesn't focus on capturing, but doesn't avoid it either. Also it avoids trying to explain territory to people (that always seems to be one of the concepts that eludes beginners). I'd be interested in hearing what kind of results this has.

Anonymous: deg wrote a beginner's pamphlet using a stone scoring variant, and the Strasbourg rules use straight stone scoring. So yeah: other people think this is a good idea too.

(Sebastian:) Yes, I think it's a great idea. This seems to be the same way as it has been done in ancient times. (Or similar - don't ask me about the differences people discuss on that page.) Often it is easier to start with the way people did it historically.

C.S. Graves: William Cobb's The Book of Go describes the basics of the game in a very similar fashion. It starts out with atari-go, then extends it to the degree that both players start filling in their own territory, and a player wins the ongoing capturing race if he has more space surrounded by his stones, therefore postponing self-capture.

Dieter : Agree completely. I have been a fan of the atari go teaching method, but lately, though not through the practice of teaching, I have become more prone to its limitations. Not as much the "obsession with capture" but rather the dullness of the game and in particular its being different from the real game.

Your proposal is actually to have the game played in its simplest form, postponing the Japanese beauty of omission for the greater sake of simplicity.

Another big argument for this teaching method is that you don't have to worry about when to pass. In conventional go, the beginner's question "When does the game end - when both pass - when do you pass - when you think there is no good move left ..." is very frustrating to both beginner and teacher. Here, it is plain to see when no good moves are left: the only available points lead to self-atari.

Consequently, the idea of two eyes is introduced very naturally. It will take only one big capture due to auto-atari for the pupil to understand the idea once and for all. In this way, life will become a unified concept, regardless seki, eyes or seemingly false but real eyes.

Yes, that's the way to do it.

Robert Pauli: Fully agree, Benjamin. Already expressed the very same opinion on Teaching Go to Newcomers / Discussion myself. BTW, maybe that's the better place to discuss it.

Bill: At his ''Internet Go Cafe'' site a fellow named Iehiro presents his method for teaching go. Instead of seeing Atari Go and Stone Counting as alternatives, he uses both.

Briefly, he starts with having the winner be the first player to capture a stone. He progresses to requiring the capture of 5 stones to win. At this point, when the game ends because one player has to fill the second eye of one of his own groups, allowing it to be captured, he switches to what he calls Zaru Go. In Zaru Go you count captives, not stones on the board, but it comes to almost the same thing.

GoJaC: Counting many stones is quite troublesome. (Both on 9x9 and 19x19.) If you have one more black stone than white stone, or an equal number is fine too) the winner is the one that runs out of stones first (prisoners given back to the person they belong to). I don't know too much about area scoring, and should read up on it more. Still, in my limited experience I believe Chinese players sometimes looks at the stones left in their bowl to have an idea of who wins. (Chinese sets usually have 181 black and 180 white stones, don't they? I know this isn't always the case with e.g. Baduk sets - the ones used most often at our club generally has around 160 white and black stones.)

Andrew Grant: As you say, counting many stones is troublesome. So any system that requires an equal number of stones beforehand (and thus requires counting before the game starts) is flawed.

Oh, and don't bother to reply telling me about Ing bowls which hold exactly 180 stones. I've tried them, and if they're the answer, it must have been a pretty stupid question. Not only do stones frequently get jammed sideways in the holes in Ing Bowls, but also there's no guarantee that you won't drop and lose a stone or two during the game. That's not a problem if you have an effectively unlimited supply of stones, but if you're depending on having exactly equal numbers of stones it wrecks the game.

GoJaC: I didn't know about Ing bowls. Interesting. Anyway, I also definately like territory scoring more (though I do feel area is definately easier for beginners). Once I played a couple of games against a Chinese person in China (we were both between 20k and 25k at the time). Near the end of our last game he said I won, but I knew he won. Due to lack of a common language the source of the confusion was hard to determine. My conclusion was that he might be looking at the stones left in the bowls, so I illustrated by putting all the stones on the board - I recall about 40 black stones were missing (only about three white ones were missing). He seemed to indicate he understood and now agreed, though for all I know that might just have been him getting tired of the situation and just wanted to "end" it. (I took white, so my stones had started "running out" - I had requested he keep his prisoners during our second game, urging him to now do prisoner exchage in the end-game after he felt he was losing was losing far. Using area scoring meant he usually didn't score, so probably didn't care much for yose. ;)

Andrew Grant: I agree that area scoring is easier for beginners; my only problem with it (apart from the difficulties of counting the score) is that it may reinforce bad habits. The beginners at my club are quite fond of making unnecessary defensive moves inside their territory. With territory scoring you can tell them that they're losing points, and they can understand this. With area scoring all they lose is sente - a much harder concept to get beginners to appreciate.

DnF: But this is not the case. Consider the situation in the diagram to the left. Clearly, if black plays inside his own territory he will lose by one point (as white gets 2 dame points and black one). If black plays a dame he wins by one. So not playing a dame point and defending instead can lose a point.

Andrew Grant: You've missed my point. Of course I know that giving up sente can lose a point (or more). What I was saying was that this is not an obvious concept to beginners.

DamienSullivan: I don't think you have to invoke sente at all. If they play within their area they don't increase their score, if they play dame they get another point. Ask which they'd choose, a move worth one point or a move worth zero points?

Dieter: You do not need an equal number of stones to use stone counting. You need it only in order to prove that it is equivalent to territory scoring. When I explain the game as "more stones on the board wins" I have no restriction whatsoever. Players are allowed to pass but it is a clear disadvantage (until the end). In Haeng-ma tutorial for beginners the author goes a long way explaining that Go is not about surrounded empty points but about living stones. Territory is a derived concept, not a prime objective. If you teach this from the start, people will less likely grow up with mutilated ideas on the nature of the game, as I did and I think many of my compatriots.

On reinforcing bad habits: the bad habit in atari-go is too much focus on capturing the other's stones (not enough on safety of own stones). In territory-go it is too much focus on surrounding empty points (not enough on safety of stones). In living stones-go (which is IMHO the real nature of the game) the alleged bad habit of adding stones in an area which is already safe is not really a matter of sente but of efficiency. In territory counting, it is supposed to be dealt with by indicating that it is a loss of point, but that can lead to the bad habit of not reinforcing. To reinforce (stabilize) or not reinforce (but develop instead): that's the fundamental question in Go. Emphasis on living stones puts emphasis on stability. Emphasis on territory puts emphasis on development. The first is "urgent". The latter is "big".

AshleyF: Here is a pure stone counting engine (modified version of GnuGo). Try pointing a total beginner (someone who even doesn't know the rules) at this and watch what happens. I've done this with a four and a five year old and it's amazing to see them figure out life and death on thier own! http://www.gogoigo.com

(end of copy original page Stone Counting as teaching method)