From: Larry Kaufman WIZARD NET> Date: 2 mar 1997 Subject: odds and ends This to reply to several postings. Charles Smith asked what is the average shogi rank. Well, the average rank of a player in a Japanese club is probably around 1 kyu or even 1 dan; the average rank of a Japanese who plays shogi occasionally but not at clubs is perhaps 5 kyu or so; the average Westerner who plays at a club is perhaps 3 or 4 kyu; the average Westerner who plays with friends but not at a club is perhaps 7 or 8 kyu. These are my guesses, don't take them as precise numbers. Schazworth said much of Morphy's success at handicap play in chess was due to taking advantage of the absent pawn to bring out his pieces faster. This overlooks two facts: most of Morphy's published handicap games were at knight odds (or even rook), since very few players in the world could beat him at knight odds. Also, when he did give pawn odds, it was always the king's bishop pawn, which does nothing for development. Thanks to Eric Cheymol for his detailed statistics, which confirm what I said about the European ratings, namely that they imply a 64% win ratio for a 1 Dan rank difference (Eric's figures average 63.8%). After fully digesting his remarks as well as some of the other people who have posted on the ranking topic, I have reached this conclusion: Each European rank is the equivalent of a half a rank in Japan. If we accept that 4 Dan has the same meaning in Europe and Japan, then European 3 dan = Japanese 3 1/2 dan, European 2 dan = Japanese 3 dan, European 1 Dan = Japanese 2 1/2 dan, European 1 kyu = Japanese 2 dan, etc. This is consistent with the claim that the European players are typically ranked a grade or so below the rank they would have in Japan, and is also consistent with the fact that Europe assumes a 3-1 win ratio means a two rank difference, while all correspondents but one (Michael Facer) agree that in Japan a 3-1 ratio means at most one rank difference. To fully appreciate the extent to which Japanese ranks are inflated at the lower dan and kyu range, let me give you some examples of Japanese players in our D.C. shogi club who return from time to time to Japan and play in the clubs there (usually in the biggest one, Shinjuku Shogi Center). One is a middle aged gentleman who is ranked 1 kyu in our club, yet who plays as 2 Dan in Shinjuku and holds his own at that rank there. I have played him many times in our club, always at 5 piece handicap, and although the games are close, he has never won. I doubt that he could win one game in a hundred playing even with our 4 dan players (assuming they were playing seriously), and yet the European system would imply he should score about 1 out of 4! My second example is an elderly Japanese man who learned shogi in his golden years; he is ranked 1 Dan in Japan (2 kyu in D.C.); he plays me at six piece, and wins perhaps 1 out of 4 games; he has a losing record in even games with my son, who is barely out of the novice range (7 kyu in our club when these games were player; now 5 kyu). I suspect that these two Japanese players would be ranked something like 5 and 8 kyu respectively in Europe! Anyway, it is clear to me that we cannot possibly match both the European and Japanese ranks; we have to choose one or the other, or else aim at somewhere in the middle. However, there is no good reason why our Elo ratings in America should not be comparable with Europe's, since we both use compatible formulas for rating even games. If anyone (Eric?) has statistics on the average European Elo of Japanese players ranked in Japan as 1 dan, 2 dan, 3 dan, and 4 dan that would help to determine if our scale needs to be adjusted to make it compatible with Europe's. Larry Kaufman, 301-309-0904, 9213 Wooden Bridge Road, Potomac, Md. 20854