Stephen B. Gray wrote:
The former. It comprises a much bigger advantage than adding 6 stones to the weaker player would be. The latter idea is never used so far as I know.
Steve Gray If by latter idea you mean adding stones to score of the weaker player, then the latter idea is komi, and it is used. Komi is typically 4.5 stones for the first move. It could be used to balance finer differences in strength than you can achieve by handicap stones (Komi, also serves to elimate ties, because Komi usually has a half point.)
To put the 6 & 7 handicap stones into perspective, for those of you not familiar with handicaps in Go, a 9 stone handicap can let a merely competent amateur player compete roughly evenly against the strongest professional. Using the normal handicapping system, I don't think the difference between the strongest and weakest professional players is more than 2 stones. A long time ago, when I had time to play, I learned that no special skill was required to reach the rank of amateur 1 dan (supposedly, if you played N games while paying attention you'd attain the rank of 1dan [N varied in the telling, from 100 to 1000]). The kyu rankings were weaker (a 7 kyu rank was 7 stones lower than a 1 dan rank, and a 3 dan rank was 2 stones higher than a 1 dan ranking). In any case, an amateur 1 dan could take 9 stones and play against the strongest professional. So 6 stones is quite a lot. (As far as I know, a 1 step delta in professional rank is about 1/3 the gap between two consecutive amateur ranks, and typically the handicap between two amateur players with no track record between them (the handicap that is supposed to make the game roughly even) is (rank1 - rank2) stones [where a "1 stone handicap" is translated into moving first]: so a 3 dan would give 2 stones to an amateur 1 dan. I think 9P (9 dan professional) is equivalent to a 9 dan amateur rank, but a 1P (1 dan professional) is equivalent to a weak 7 dan amateur rank. So a 6 stone handicap would allow an amateur 3 dan to play roughly evenly against the strongest professional players. ) Although this is a very coarse measure of playing strength, and 6 stones is a lot of stones, it is still pretty amazing. I know it's not really fair to try to guesstimate a ranking based on winning games (without knowing about the 1000s of other lost games), but assuming these were representative games against the professionals, and assuming the margin of victory was small: August 2008: 8P dan professional w/ 9 stones = somewhere around 2kyu or 1 dan amateur Feb 2009: 9P dan professional w/7 stones = somewhere around 1 or 2 dan amateur : 1P dan professional w/6 strones = aomewhere around 1 dan So it seems as if this program may play as well as an amateur 1 dan. Not stupendous, but competent. Which is pretty amazing considering how weak computer go programs have been even fairly recently.