From: bogin YAHOO CO JP> Date: 30 may 2003 Subject: Re: Dual ratings Hello everyone, I am not a mathematician or a statistician or any kind of tician at all. Also, I couldn't even try to pretend to pretend to find some relationship between Shogi Club 24 ratings and the various other rating systems out there. I know the NSR doesn't really help things by granting equivalent ranks for actual play as well as for solving tsumi problems and best next move problems. It is possible to gain a 5 dan rank from answering 19 out 20 questions from various newspaper web sites. Perhaps things would be less confusing if they offer tsumi problem solver 5 dans and best next move 5 dans along with actual OTB 5 dans. I also know that each club may have their slightly different way of rating players and then equivalent ranks in different parts of the world may be different simply because the overall level of play may be different in different areas. A Tokyo 5 dan may be a few ranks stronger than an Okinawa 5 dan or vice versa. With regards to Shogi Club 24 ratings, of course they are not completely accurate. The time control can be varied, some people may have bad Internet connections, some people may play better sitting across from an actual person than in front of a computer monitor. In postal chess, it is not all that rare for an average OTB player to be an strong postal player and for an strong OTB to be only average at postal chess. Playing online may require a different way of thinking than playing an actual person. Also, on Shogi 24 you can choose you opponents but that's how it is on most Internet sites, excluding online tournaments of course. But, even though some people may try and cheat I think the number is very small. Also, I think most people tend try and play in their own rating range or play up since that`s the best way to gain points. A 2100 rated player can play 30 games in a row against a 1500 win 29 and then lose the 30th game and end up losing more points from that 1 loss then they gained from the previous 29 wins. It's just not practical. If you constantly play down then you will get burned eventually. Some sites, including Shogi 24 realize then so they prevent players from challenging others 400 or more points below them. In some of the small clubs in the US and Europe the same people play each other over and over and over. Of course, theirs no one else to play in a way that's sort of like choosing your opponent. How does beating the same player over and over again become a good way to base a rating system on? How does that guarantee a more accurate rating system than one used on the Internet by a club that has over 130,000 members. Of course, you can play the same person again on the Internet as well, but not like I've seen in the cross tables of some of the FESA and USSF tournaments. Once on Yahoo English Shogi, I played a nice man 3 or 4 games in a row. We were both provisional so our ratings and relative playing strength were an unknown. I won all of the games and after our last game my opponent complimented me and said that I must be a very high dan. I told him I was only a 2 dan and he said that I was mistaken since he was a 2 dan and I was obviously much better. He wasn't a bad player but the games were relatively easy and he made some pretty simple errors so I was have estimated his strength around 5 or 6 kyu. Anyway, he assumed that since he was a 2 dan and that since I beat him that I must be underrated. The opposite possibility of me actually being a 2 dan and him even being slightly overrated never seemed to occur to him at all. In reality, it was probably a combination of both or perhaps I was just lucky. After all, we could've played 3 more games and I could've lost them all. It seems that since there are a lot of people out there with high USSF ratings and FESA ratings who have much lower ratings on Shogi 24, the assumption being made is that Shogi Club 24 is vastly under rating everyone and that our system is much more accurate in determining a players actual strength. Otherwise why would all these "strong" players be losing to people 200 or 300 hundred points below them? It make a person feel good to think that "Hey, they're beating me because in reality they are actually higher in rank than I am. Shogi 24 is under rating them. That must be why I'm losing to these guys. What other explanation could there be?" Isn't it remotely possible that it's the other way around for a lot of the people out there. Isn't it possible that except for the top players out there that the USSF and FESA rating systems are slightly over rating people. Could it be that people are starting out way to high with their initial rating when they first join Shogi Club 24 and are only slowly moving toward their true playing strength. If I think I'm a 5 dan and sign up as a 5 dan but it reality I'm way over my head, won't I mostly lose and drop points until I end up where I really should have been in the first place? I can understand 100, 200 or even 300 points difference between the 2 sets of ratings and just figure it's a fluke or bad luck. Like I said not everyone may be that good at online shogi. But, how can you explain 500, 600, 700, 800 and in some cases higher differences between both ratings? It seems strange to me to just say it's because Shogi Club 24 under rates everyone. I don't have a current USSF rating but I did play in small 3-game tournament in Chicago a few years back. I went 3-0 and I think I ended up with a provisional rating of around 1970 or so. But, I think that was more of a reflection of my opponents' numerical ratings than the actual quality of play. I really am only a NSR 2 dan and I also happen to be a Shogi Club 24 2 dan as well. I started at the bottom and worked my way up. I lose games to lower rated players but I also beat higher rated players as well. I win games on time and lose games in which I should win because of time. But, it seems to all even out over the long haul and I feel I'm right where I should be in playing strength at this time. I consider my Shogi club 24 rating to be more accurate than any other rating I have. Isn't there anyone else out there like that as well? Doesn't anyone out there have the same USSF/FESA rating as what they have on Shogi Club 24? Is everyone out there under rated? Do any of the Japanese members of the Amaren (Japanese Amateur Shogi Association) have Amaren ratings that our much different from their Shogi 24 rating? Bill Gaudry --- Larry Kaufman COMCAST NET> $B$+$i$N%a%C%;!<%8!' (B > The following is a listing of all ratings of players rated on > both the Pan-Atlantic and ShogiClub24 rating systems fulfilling the > criteria given below. Of course, only those players whose identities > on ShogiClub24 are known to me could be included. The first column > is Pan-Atlantic, second is ShogiClub24. > > 2422 2284 > 2285 1993 > 2255 2013 > 2195 1753 > 2174 1925 > 2118 1455 > 2101 1465 > 2094 1555 > 2074 1064 > 2044 1181 > 2012 1453 > 1980 1513 > 1926 1378 > 1924 1124 > 1908 1161 > 1852 1377 > 1809 1298 > 1784 993 > 1769 1149 > > All players listed have at least 80 games on ShogiDojo, which > reduces the effect of starting rating to under 5%. All players > listed have at least 30 games on Pan-Atlantic, where starting rating > plays only a tiny role and is in any case based on known rank. > Player must be listed on 2002 or 2003 Pan-Atlantic rating list. No > names are given in case some players prefer anonymity. > There are 19 players on this list. The average spread is 557.5 > points, more than I expected. We need to do a linear regression on > the above numbers. If anyone can perform this for us easily, we > would appreciate it, so I don't have to rack my brain trying to > remember how to do it on my calculator or how to work downloadable > software. > > Larry Kaufman > __________________________________________________ Do You Yahoo!? Yahoo! BB is Broadband by Yahoo! http://bb.yahoo.co.jp/