Re: revised "Quality Ranking" (long)

[abe@kaleida.com (Masumi Abe)]

My formula is not for predicting the next ranking but to calculate
performance of makuuchi rikishi in previous basho. The performance ratio
can be used to predict the outcome of a match between two rikishi during
next basho, if all conditions are the same as the last basho. For example,
even if lower Daizen beats upper Mainoumi, it should not be considered
"upset".

We also could use the performance ratio to see if next official ranking is
based on the actual relative strength of makuuchi rikishi or not, if we
believe the performance ratio is reasonable way to calculate the rikishi's
strength.

I would like to fix the discrepancy in my previous posting.
Please swap Minatofuji and Kasugafuji in Mb rank. The correct ranking for
Mb rikishi should be as follows.

Mb
M4  Hamanoshima (M8,  8-7)        83
M4  Minatofuji  (M15, 9-6)        80 
M5  Daizen      (M4,  7-8)       105 (rule#2)
M5  Misugisato  (M4,  6-9)        85 (rule#2)
M6  Kotonowaka  (M3, 5-10)        84 (rule#2)
M6  Kasugafuji  (M9,  8-7)        78 (rule#3)
M7  Kyokudozan  (M9,  8-7)        77 (rule#3)
M7  Kirishima   (M11, 8-7)        76 (rule#3)



sxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsxsx
Masumi Abe                                      Tel: (415) 966-0824
Kaleida Labs, Inc.                              Fax: (415) 966-0496
1945 Charleston Road                            Internet:  abe@kaleida.com
Mountain View, CA 94043                         AppleLink: ABE.M