play chess online with friends

Online Chess

Tag and you will find!
Funny name, real hobbies
[ Sign up | Log in | Guest ] (beta)

Have you ever played online chess? Online chess can be a lot of fun. Come play online chess with us!

» Play Chess Online

gibo 79 ( +1 | -1 )
Tournament performance ratings I recently played in a swiss otb junior tournament. This tournament used unique junior ratings only used in NSW Australia. I was seeded 5th and came equal first with 9.5/11 which i was quiet happy with.
The rating system used for this tournament are nswjcl ratings. These ratings vary from 250-1350, the highest rated (1350) is also a 2400+IM. My rating is currently 914. It is most difficult to compare these ratings with fide ratings. Thus when i used the tpr system provided on the fide website for the nswjcl ratings, the result was ridiculous. In the tournament i played in the ratings ranged from 250-1011, there were 205 players. What tpr system do you recommende i can use to find my performance ratings is this 11 round tournament?
fmgaijin 26 ( +1 | -1 )
Try BCF Conversion The old British Chess Federation system had similar ratings and your regional federation probably set up a system along those lines (given the connection between Australia and the UK); in general, add about 1000 points to a BCF-style rating to equal FIDE.
gibo 29 ( +1 | -1 )
fmgaijin I must say I dont aggree with you on that. I have a good knowledge of the strength of players with nswjcl ratings and adding 1000 to equal fide would not be correct. It would for some but for most, there fide rating would be grossly overated.
fmgaijin 84 ( +1 | -1 )
All Ratings Are RELATIVE Remember, even when one uses the same Elo rating equation, every rating pool will be different. All the system does is to establish the RELATIVE strengths of the players. Hence, the adjustment I gave you might not be accurate if the rating pool is relatively weak, but it gives a conversion to an Elo-type numbering system that is probably accurate WITHIN the group. If the top players are 1300's and the bottom 200's, then that would shift them to a 1200-2300 range such as that here on GK. However, when we at GK play a match with IECC (as currently), the ratings don't match up because there are two different rating pools. I have OTB ratings with 3 national federations and FIDE and my OTB ratings range from the low 2300's to the high 2400's, but they converted my GK rating of 2243 to 2586 for that match!
werwolf 6 ( +1 | -1 )
thats right I` agree with fmgaijin - ratings are realy relative....
gibo 5 ( +1 | -1 )
But isn't there a general ratings system i can use to work out my tpr?
fmgaijin 35 ( +1 | -1 )
WITHIN most Elo-modeled Systems . . . . . . you get a performance rating by 400 (W - L) + Total Ratings of All Opponents and divide that result by the # of games. (Was that what gave you the "ridiculous" result?) So, for example, if you had 9.5/11, W - L = 8. 8 X 400 = 3200. Say your opponents had an average rating of 800 (in the system you were discussing), so 800 X 11 = 8800. 8800 + 3200 = 12000. 12000 / 11 = 1091 (rounding), which would be your performance rating.
gibo 31 ( +1 | -1 )
fmgaijin this system looks like it could work for me, but in the tournament i didn't lose any games. i scored 8 wins and 3 draws, so how could i work it now. M average rating of opponents was 797. But the problem with this system is it does not recongize the types of wins and draws, as it take average rating.
fmgaijin 74 ( +1 | -1 )
Gibo 8 wins - 0 losses still = 8 as in my example. Your opponent's average rating is almost exactly as in my example. So it's easy to figure out the performance rating; it's just slightly lower than the example (1088). As for your other comment, the rating system is a system of mathematical probabilities and so it DOES NOT CARE which players you beat, drew, or lost to. It takes your performance AS A WHOLE. Why would it make any difference which players you had which results against when considering your overall performance? If I beat a 2300 in one game and lose to a 1900 in the next instead of drawing two 2100 players to get the same 1 point score with the same performance rating of 2100, my overall performance is not any different.
gibo 9 ( +1 | -1 )
but shouldn't it recognize my draws? in your system it only recognizes wins and loses?
fmgaijin 43 ( +1 | -1 )
Not MY System, Arpad Elo's! Draws are half a win and half a loss, so for those games you're performing exactly at your opponent's rating. Think of it this way: (1) Every win, add 400 points to your opponent's rating. (2) Every loss, subtract 400 points from your opponent's rating. (3) Every draw, take your opponent's rating. Now add up all of those and divide by the number of games. You'll find that it's exactly the same as what I gave above.