daal wrote:For simplicity's sake, let's just look at Black to Kill type problems.
I believe "White to live" is considerably more complex, and that suggests there is something artificial about this business.
But seriously, if it helps for black to kill but not for white to live, wouldn't that still be a good thing?
Might be something here of the old joke about looking for the lost keys under the lamp post, because that is where the light is.
In practical play and shinogi, finding an extra half eye or weakness in the surrounding groups are both complex reading issues. There are problem sets about bridging under, I recall, which are quite tough; and any sort of shortage of liberties issues occur. These are big areas, and "combinatorial" (probably not reducible to a short list of things to try).
Charles Matthews wrote:In practical play and shinogi, finding an extra half eye or weakness in the surrounding groups are both complex reading issues. There are problem sets about bridging under, I recall, which are quite tough; and any sort of shortage of liberties issues occur. These are big areas, and "combinatorial" (probably not reducible to a short list of things to try).
Pure curiosity: Are they tough because the stones end up in unusual patterns which make normal candidate moves for similar shapes not work? Outside of factors like reading depth etc. Interested from a "how much of high dan problems are taking the solver away from the usual shapes/patterns they are familiar with rather than just being a case of needing deeper reading" perspective. Might be a trite question.
Charles Matthews wrote:In practical play and shinogi, finding an extra half eye or weakness in the surrounding groups are both complex reading issues. There are problem sets about bridging under, I recall, which are quite tough; and any sort of shortage of liberties issues occur. These are big areas, and "combinatorial" (probably not reducible to a short list of things to try).
Pure curiosity: Are they tough because the stones end up in unusual patterns which make normal candidate moves for similar shapes not work? Outside of factors like reading depth etc. Interested from a "how much of high dan problems are taking the solver away from the usual shapes/patterns they are familiar with rather than just being a case of needing deeper reading" perspective. Might be a trite question.
I suppose the point I may be trying to make is that computers may be better than humans in some such positions, anyway. For example, the technique to live is miai, but the key play for a double threat only shows up after you have explored two complex variations.
