Life in 19x19. Go, Weiqi, Baduk... Thats the life.
https://lifein19x19.com/
Fair enough, but I don't have a clue what that means. Would you care to explain.Bill Spight wrote:DrStraw wrote:Every time I read this thread about PON I find myself wonder what the ponic value (or is it pony) of a ponnuki at tengen would be.
PON estimate = int(0.33(9 + 0.5(12) + 0.35(16) + 0.25(20)) - 1.96) = int(6.488) = 6
What it means is that if the opponent gets 7 attacks moves in a row, he kills; i. e., if you answer the 6th attack move, you can live.DrStraw wrote:Fair enough, but I don't have a clue what that means. Would you care to explain.Bill Spight wrote:DrStraw wrote:Every time I read this thread about PON I find myself wonder what the ponic value (or is it pony) of a ponnuki at tengen would be.
PON estimate = int(0.33(9 + 0.5(12) + 0.35(16) + 0.25(20)) - 1.96) = int(6.488) = 6
No, understand the 6 part. It is the highlighted part I need an explanation of.Bill Spight wrote: PON estimate = int(0.33(9 + 0.5(12) + 0.35(16) + 0.25(20)) - 1.96) = int(6.488) = 6
What it means is that if the opponent gets 7 attacks moves in a row, he kills; i. e., if you answer the 6th attack move, you can live.
Bill gave a worked example of the PON algorithm in his first post on the subject: http://www.lifein19x19.com/forum/viewto ... 05#p211105.DrStraw wrote:No, understand the 6 part. It is the highlighted part I need an explanation of.Bill Spight wrote: PON estimate = int(0.33(9 + 0.5(12) + 0.35(16) + 0.25(20)) - 1.96) = int(6.488) = 6
What it means is that if the opponent gets 7 attacks moves in a row, he kills; i. e., if you answer the 6th attack move, you can live.
Let us call an empty point that has a Manhattan distance of N from a group an Nth order dame of the group. The diamond shape has 9 first order dame (8 outside the eye, 1 inside), 12 second order dame, 16 third order dame, and 20 fourth order dame. The weights for the different orders of dame and the linear equation were empirically derived, but I am highly critical of the methodology and lay no store in the estimate. OTOH, I do think that the 4th order dame are in general relevant to the PON, so there is something there.DrStraw wrote:No, understand the 6 part. It is the highlighted part I need an explanation of.Bill Spight wrote: PON estimate = int(0.33(9 + 0.5(12) + 0.35(16) + 0.25(20)) - 1.96) = int(6.488) = 6
What it means is that if the opponent gets 7 attacks moves in a row, he kills; i. e., if you answer the 6th attack move, you can live.
Tajima and Sanechika were at the Electrotechnical Laboratory in Tsukuba, Japan, in 1998, which was when their paper was presented, at the same conference where I presented my paper on evaluating superkos and multiple kos. I did not meet them, however.Uberdude wrote:Btw I presume Tajima and Sanechika and names of 2 Japanese people. Who are they? I found this reference to the PON paper https://www.semanticscholar.org/paper/E ... 84c26cc397
Okay, that makes sense. Thanks. I did something similar 15 years ago when I did my counting lessons. But I used the ideas to do an estimate of potential territory that may be gains from the influence a group generates.Bill Spight wrote:Let us call an empty point that has a Manhattan distance of N from a group an Nth order dame of the group. The diamond shape has 9 first order dame (8 outside the eye, 1 inside), 12 second order dame, 16 third order dame, and 20 fourth order dame. The weights for the different orders of dame and the linear equation were empirically derived, but I am highly critical of the methodology and lay no store in the estimate. OTOH, I do think that the 4th order dame are in general relevant to the PON, so there is something there.
How do you estimate the territory for White's right side group? I estimate around 10 pts., which suggests that the group is not in any immediate danger and can ignore at least one attacking play; i. e., it has a PON of at least 2.DrStraw wrote:Okay, that makes sense. Thanks. I did something similar 15 years ago when I did my counting lessons. But I used the ideas to do an estimate of potential territory that may be gains from the influence a group generates.Bill Spight wrote:Let us call an empty point that has a Manhattan distance of N from a group an Nth order dame of the group. The diamond shape has 9 first order dame (8 outside the eye, 1 inside), 12 second order dame, 16 third order dame, and 20 fourth order dame. The weights for the different orders of dame and the linear equation were empirically derived, but I am highly critical of the methodology and lay no store in the estimate. OTOH, I do think that the 4th order dame are in general relevant to the PON, so there is something there.
The methods I described in those lessons would give it about 10-12 points, but it wasn' that clear cut. I used both probable and possible points, counting the possible for half weight.Bill Spight wrote:How do you estimate the territory for White's right side group? I estimate around 10 pts., which suggests that the group is not in any immediate danger and can ignore at least one attacking play; i. e., it has a PON of at least 2.DrStraw wrote: Okay, that makes sense. Thanks. I did something similar 15 years ago when I did my counting lessons. But I used the ideas to do an estimate of potential territory that may be gains from the influence a group generates.
, above, is good for black. White's shape is cramped an inefficient. Later in the game, maybe it can be attacked.
No, it's not clear cut. I didn't mean to give the impression that it was. My methods also yield about 10-12 points, but I discount their results in the center.DrStraw wrote:The methods I described in those lessons would give it about 10-12 points, but it wasn' that clear cut. I used both probable and possible points, counting the possible for half weight.Bill Spight wrote:How do you estimate the territory for White's right side group? I estimate around 10 pts., which suggests that the group is not in any immediate danger and can ignore at least one attacking play; i. e., it has a PON of at least 2.DrStraw wrote: Okay, that makes sense. Thanks. I did something similar 15 years ago when I did my counting lessons. But I used the ideas to do an estimate of potential territory that may be gains from the influence a group generates.
This variation was proposed in one of the go magazines that I read back when the Chinese opening was starting to become popular.Kirby wrote:From this I gather that a weak group or a group that can be attacked is ok in an area where the opponent already has an advantage. If you can minimize what your opponent gets in an area where your opponent is strong, that's ideal.
In the Chinese opening, black's marked stone gives a local advantage:
So if white gets a weak-ish group in the 'a' area, maybe it's OK - if white sufficiently reduces Black's advantages area.
These days, it's not uncommon to see pros try variations as white in the Chinese opening, such as the following:
Since these more recent variations deviate from "standard", I suspect that pros these days are not 100% satisfied with the result of the classic way for white.
That being said, the old way still seems played sometimes, so perhaps there is variation in pro assessment.