Counting without Tears
Over the board it is usually practical to make use of estimates of the count and the value of plays. Away from the board, positions may be analyzed at leisure. Doing such analysis may help you in making estimates over the board, and learning the sizes of plays and local counts can provide a ready reference.
Judging from their writings, it appears that over the board pros typically either rely upon memory or make an educated guess whether a play is sente or gote, and take the difference between the counts of the resulting positions, working them out in similar fashion, if need be. This rough and ready approach is good enough, as a rule. With enough time they will also project the play up to the end of the game. In the old days, with no time constraints, endgame mistakes by strong players were rare.
John F. You may in general be right, Bill, but Liu Yajie was shocked when I suggested some thing similar. She said that Chinese pros count each position carefully except in fast games. No doubt, though, this is made easier by memorising lots of standard cases.
Bill: Thanks for the comment, John. I did not mean to imply that the pros were not careful. And I know that a good deal of counting occurs early, before yose and thus, before byo yomi, so that they have time to be careful. So maybe rough and ready is not appropriate. I do think that memory plays a big role.
(Ah! I see where I suggested too much guesswork. I have altered that.)
However, I do not believe that most pros do better over the board than the yose books and articles, which show the earmarks of the approach I suggest. One major reason I have not come out with a translation of a yose book is that there are many errors in calculation and some errors, as a result, in play. Much revision is required, with new diagrams.
Anyway, my point for talking about that approach is to reassure fellow amateurs that you do not have worry too much about exactness. Approximate methods and educated guesswork can be effective. :-)
Here I plan to show how to calculate simple gote, sente, and reversals, as well as simple kos, and indicate how to apply the technique to more complex positions. I will leave out any discussion of hyperactive positions, which few people understand, anyway. ;-)
Fhayashi: By the title, I thought it was some way to always win games, because when I count my losing games, I cry... =P
Gabaux: I do think, the idea is very good. An analyses after the game helps to learn on our errors. I can hardly wait for examples! :-))