I have seen a ratings site mentioned here, with approval, which I think has Maruyama in its name. I have never looked at the site, and may have misremembered, but FWIW I have just seen a long article (in Japanese, in Go Weekly) by someone called Maruyama Susumu. He describes his system, and gives lots of results. I assume it is the same one.

I personally have no interest in numbers, but I know many people here do, so I am giving a brief description.

Maruyama notes that there are various systems used for internet games, chief among them being Elo. He claims not to understand Elo and so "reluctantly" has not used it. He has invented his own.

His system is for Japanese official games only (Nihon Ki-in and Kansai Ki-in), and excludes international games. His results are based on 45,000 games.

He starts with a base rating for each player of 1500. Players' ratings go up or down after each game. If they are evenly matched (which seems to mean they are within 99 points of each other, the winner scores +8 and the loser -8.

If the there is a difference of 100 to 199, the winner scores +10 if he is the lower rated one, and the loser -10. If the winner is the higher rated player, he scores +6 and the loser (the lower rated) scores -6.

A similar formula is used for a rating difference of 200 or higher, but the scores are +/-12 and +/-4.

There is no reference to oddities such as void games.

On this basis, the ranking list as of end-February 2020 puts Ichiriki Ryo top on 2114, above Shibano then Iyama.

He gives his top 50. I show here the top 20.

1. 2114 Ichiriki Ryo
2. 2080 Shibano Toramaru
3. 2076 Iyama Yuta
4. 2035 Kyo Kagen
5. 1978 Cho U
6. 1964 Kono Rin
7. 1928 Yu Zhengqi (aka Yo Seiki)
8. 1923 Motoki Katsuya
9. 1919 Ida Atsushi
10. 1906 Fujita Akihito
11. 1905 Murakawa Daisuke
12. 1904 Son Makoto
13. 1904 Hane Naoki
14. 1902 Takao Shinji
15. 1896 Mutsuura Yuta
16. 1893 Yamashita Keigo
17. 1890 Suzuki Shinji
18. 1890 Shida Tatsuya
19. 1869 Onishi Ryuhei
20. 1857 Seto Taiki

The rating figure for place 50 is 1750. Top female is Fujisawa Rina on 1784 in 35th place (just above naughty boy Yoda; Ueno Asami is on 1779 in 38th place). Luminaries such as Cho Chikun are absent. That doesn't always mean they've got much weaker, by the way. The Ki-in's are largely self-administered by the players, and the older (wiser!) ones are expected to become directors, etc. and so play less. Cho Chikun is also spending a fair amount of time playing in the Korean Seniors League.

It appears that there is no effort to measure activity, and so we get Sakai Hideyuki at 29th in the list on 1807, despite the fact that he's gone back to medicine.

There is also no attempt to add a weighting for title games, which may make Ichiriki's ranking above Shibano and Iyama look strange, but Maruyama also gives table going back to end-2014, and up to end-2018, and in all those Iyama was top, but from 2016-2018 Ichikiriki was in second place, so he has been consistently high.

As regards Shibano, he is just too young to have featured much, and only crept into the top 20 at end-2016. But by end-2017 he was in 5th place, and yet by end-2018 had dropped to 7th. I'm guessing he may have hit a wall as he moved up the tournament ladders and started meeting many more top players. If so, he seems well over that wall by now. He has also shown the biggest improvement in this January and February combined: +142. (And if you want another name to watch for in the future, apart from Sumire, newish Kansai Ki-in 1-dan Tanaka Koyo has posted +136 in that period.)

Maruyama himself points out an unexpected oddity in this system. Because of the differing structure of the various tournaments, it can be hard or easy to make big numerical gains even if you are improving fast. In the women's tournaments, where players generally have lower ratings, +8 is the norm for each game. But in a fairly randomised event like the Ryusei, games with players of +/-200 are pretty common. Ueno Asami, who made the Ryusei final last year, scored big because she clobbered the likes of Kyo Kagen, Murakawa Daisuke and Takao Shinji, but only lost -4 in rating points when she lost to Ichiriki in the final.

Little Nakamura Sumire is 14th in a separate table for females, with a rating of 1507.

Thanks, this is very kind of you!


It looks like he's invented something fairly similar to Elo!

I guess if you don't play chess and don't read English, then information on Elo is hard to come by?


In the chess world, Elo tends to go hand in hand with Swiss format tournaments (not by design, it's just a fortunate historical accident). The later rounds of a Swiss (or McMahon) tournament provide lots of opportunities for games against equal strength opponents, as well as for emerging pros (or strong amateurs in pro-am events) to play against slightly stronger opponents without being way out of their league. It's a pity the professional go circuit hasn't been keen on these formats. I think the current system reinforces the gap between established pros and the rest, and is one of the barriers to professional go becoming a truly international affair.