Hi,
As the Shatranj top lists only contain players that have played the last two weeks, and I have been told that Shatranj was implemented on Chess.com already back in 2018, I thought it would bee interesting to collect information about the highest ranked players.
In the following Excel link:
https://docs.google.com/spreadsheets/d/15-7Zdo-AHm36zCq-LAeQ8WVSed9UGJse/edit?gid=1292100573#gid=1292100573
I am happy to present 3 Shatranj rating lists. The first Best list shows all Shatranj players I have found that have made 1900+ rating at some point, either in blitz or rapid (dark-green). The second list shows the 100 current highest rated players by 2. January 2025. The closed accounts are excluded, and only blitz ratings are included here. The last list is a country list where the mean of five highest ranked players is calculated. As this list is the last I have worked with it has huge potential of improvement. In order to get started many of the lowest ranked countries are supported by artificial agents with low ratings.
Digging into old games and players on Chess.com is almost as hard as looking into the beginning of Universe. The archive supports only your last 40 games, and only your last 40 games against each player in two-player seek. The oldest games I have found are from late 2020. So if you know of players missing in my lists – let me know! I hope of publishing this list in future every year in January.
I have also done some correlation analysis (Pearson) between Shatranj ratings and (western) chess ratings, and between Shatranj ratings and amount of Shatranj games played. This can of course be done in several ways (live ratings, best ratings, blitz games only, total amount of Shatranj games) but for both measures the correlation is around 0,45-0,50. One important thing to remember here however is the fact that in Shatranj everyone starts on 1500 rating, and as there are few players compared to western chess, getting your correct shatranj rating might in did in some cases take hundreds of games.
Regards
Rungne