moved from Principles of Yose -- Zook
Someone: I don't know if this is the best place, but I'd like to ask a question about the quality of professional play in the yose. Basically, I was wondering how many more points the pros who excel especially at yose might hope to gain over a typical pro in the endgame. For example, I have read that Lee Chang Ho is considered the best endgame player ever, and I'm wondering about how many points per game that translates to (over what an average pro might score starting with the same late-mid / early end game position)?
BobMcGuigan: I don't know whether there has been any experimentation with this sort of thing among pros. My feeling is that differences in pro performance are probably very small, on the order of a small number of points at most. There was a series in Go World magazine several years ago in which strong amateur performance was compared to pro performance in the same position and the difference was fairly dramatic as I recall.
Charles I have a working figure: difference at the endgame = 30% of difference in level. This seems quite good over most ranges - except that something odd can happen around amateur shodan. Given that a third of a stone is the sort of difference that can be detected in classing pros, I reckon that the average difference between top pros can only be a point or so. (This may seem small: differences in komi are small, too - but noticeable at this level.)
RafaelCaetano: Hmm, if the difference is so small, how come some pros are regarded stronger than others at the endgame? And this "endgame" is yose or shuuban?
Charles No - how come you regard the difference as small? In fact, as most amateur dan players know, it ie easy to look strong in the endgame if the opponent only fights, and assumes the game will be decided by resignation. Pros can quite easily keep a game close, if they are happy only to lose by a few points.
Charles Not that. I'm saying something like: if the players are 30 points apart over a complete game, then attribute 9 points of that to the endgame.
kokiri Similar, yet different - it's not unusual for me to play 4 or 5 stone handicap game against a stronger player (say 4 dan), and limp into the endgame only to get butchered. I think that this relates only partly to my bad play in the endgame, I think that a large part of looking good in the endgame is playing in such a way in the middlegame so as to leave yourself as few weaknesses and as many big plays as possible. This is less relevant to stronger players' play however, I would imagine.
crux: There's an interesting experiment in Get Strong at the Endgame. It has two game records starting at the same early endgame position; in one of them two professionals play out the endgame (using 78 moves), in the second one a 3 dan amateur replaces one of the pros (this game lasts 54 moves). The amateur loses 8 points compared to the professional.
Bob: This is the kind of thing I mentioned above (from Go World magazine)
dnerra I think the most important endgame skill at the pro level is to be able to predict very early who will win the game when reading out variations that simplify the game leading straight to the endgame. I think this is also a part where Lee Chang Ho is considered to excel. Other than that, from reading Go World commentaries, it seems that it is mostly a matter of consistently avoiding mistakes.
Btw, I would think that I would definitely lose more than 8 pts in an endgame against a pro (if early endgame is included).
BobWhoosta: Interesting followup question: Is Lee Chang Ho so noted for his endgame play because of his endgame play, or because he is setting up the game for yose from the middlegame?? I think typically you will find the calmer players will be "better" at endgame on any level, simply because their moves tend to be more solid, and the attacking players "worse" because in the midst of an attack they may leave weaknesses behind. So shape comes into any discussion of yose, and attacking shapes are shapes that can be counterattacked. These counterattacks generally come in the yose stage.