Chess Forum
[ mod this thread ] << - < - > - >>
| From | Message |
| spurtus 11/20/2003 18:33:27 [ report this post ] |
 Subject: ELO too high?... too low?Message: Hi, Dunno if this is the best place to post this but I feel my 'ELO' rating is too high... now... wait a second... My official rating (OTB ... over the board) is 1150, this is my first grading, and yet can develop but never the less my apparent ranking in my country. Nut my ranking here is 1600+ ...great I'm very pleased about that..in fact cautiously elated!....but is it not the case that others here feel this might just be too high? My own self estimation is a true grade for myself might really be about 1300, 1150 is likely to adjust apwards from the results I get now ( thanks gameknot! )... but some 300 points short of the grading I get here. Question ... although the site masters are computing the ELO carefully, and I Repsect this is not trivial in any shape or fashion...but can they please confirm how more or less roughly this difficult thing took place?... Im not complaining jus want to know if I should aspire to my grade I see here on OTB games OR is the ELO here just not the same for some fundamental reason? If applicable... Please post your OTB rating and GAMEKNOT rating here in any correspondance. Kind regards, Spurtus |
| chess_champion 11/20/2003 22:47:32 [ report this post ] | ratings Message: the gk elo rating system does have some modifications so the ratings could still be off by a little. my gk rating is 1488 and my otb cfc rating is 1223 which i find rather low. recently i went to an otb tournament and scored 4 out of 6 so i expect my rating to raise to about 1350. (i would guess im about 1400-1450.) |
| baseline 11/21/2003 00:25:08 [ report this post ] | rating are only relevant Message: for the organization where they are used. We are talking a different pool of players and even different rules. Just the time control makes a big difference. |
| bigkev 11/21/2003 05:02:06 [ report this post ] | ratings Message: I've heard that there can be up to 400 points difference between Correspondence chess and OTB due to having access to books, databases, etc. So in some ways your rating here should be higher here than your official rating. |
| peppe_l 11/21/2003 07:50:16 [ report this post ] | The difference Message: Between OTB and correspondence chess ratings can be far greater than 400 points, for example in my country there are several players who have ICCF rating 700+ points higher than OTB rating. This topic has been discussed before (many times actually!) and as already pointed out by others OTB chess and correspondence chess are completely different and therefore comparing GK & OTB ratings is pointless. |
| baseline 11/21/2003 11:25:48 [ report this post ] | Peppe_l is right Message: I believe the present world correspondence chess champion is only a fide master otb. |
| fmgaijin 11/21/2003 14:39:47 [ report this post ] | Howard Johnson is Right about Abel Johnson... Message: And as I mentioned elsewhere, Dan Fleetwood, a current finalist in the ICCF World Championship isn't even a national master OTB but is an ICCM and approaching GM status in correspondence. |
| baseline 11/21/2003 15:06:14 [ report this post ] | fmgaijin Message: A couple of books I enjoyed very much by correspondence players are: "The Chess Analyst" by Jon Edwards and "Journal of a Chess Master" by Stephan Gerzadowicz both publish by the thinker's press. It's very interesting to see how correspondence masters think as they play. |
| fmgaijin 11/21/2003 21:19:15 [ report this post ] | baseline Message: True. I've known a number of very strong correspondence players (my college teammate Dan Fleetwood, Golden Knights/CCLA Grand National Champion Anton Sildmets, former ICCF WC finalist ICCM John Kalish are three who come to mind), and while each of them had a different approach to the game, they also had some common strengths (willingness to investigate lines to a deep level, good organization skills to keep track of lines they'd analyzed, reliable evaluation of static positions, for example). BTW, another great CC book is Purdy's about winning the first World Correspondence Championship. I've read one of the ones you mentioned but not the other. P.S. Though I've been an OTB player for more than 30 years, this is my first foray into "serious" <LOL> CC, and I'm enjoying it so far. |
| baseline 11/21/2003 22:54:53 [ report this post ] | fmgaijin Message: I have the book on Purdy, I have Berliner's book too, and loved them both. I really learned chess playing correspondence chess with a WWI veteran from California. I think its safe to say there are few here playing serious correspondence chess. |
| spurtus 11/23/2003 12:21:07 [ report this post ] | Message: So my grading here should have what modifier applied to meet with a kind of 'world' ranking? ... or is my/your grading here best indicitive of a 'world' ranking'?... which it should actually be really since the internet is worldwide?.... or even could this mean that my fellow Scottish players are better than the world and I am playing in a tougher average group? Regards, Spurt. |
| baseline 11/23/2003 13:16:26 [ report this post ] | spurtus Message: I don't need to point out the differences between otb play and play here at GameKnot. There are no modifiers that I and conceive of that would embrace all of these differences. |
| spurtus 11/23/2003 15:41:31 [ report this post ] | OK,... Message: Sure, but since the GK ELO is not what my OTB rating refelcts, and others so far agree that there is a significant margin of difference, why might this be the case? My point centres around the fact that there is a BIG diffence, the question is firstly why?... and why so much? Spurt. |
| fmgaijin 11/23/2003 22:34:03 [ report this post ] | False Assumptions Message: Spurtus: You're mistakenly assuming that your particular case generalizes to players on GK as a whole. I also know a number of players here on GK whose OTB ratings are SIGNIFICANTLY HIGHER than their GK ratings. Three of many possible reasons: (1) They're stronger OTB players than CC players. (2) They haven't played enough games yet to find out if their CC ratings will settle in the same place as their OTB ratings. (3) Every ELO rating ONLY gives your rating relative to other players in that particular system. ELO ratings CANNOT be compared accurately across systems unless ratings from one system can be carried over to the other or the "initial" rating base is established using reasonably-accurate ratings carried over from the other system. The "original" ELO systems were designed to center around a median rating of 1500, with players' ratings going up or down from there. If you started a new group of players in a rating system who were all masters but set the system up to have the median be 1500, then 1500 players would be pretty darned strong! The converse of course holds--if the initial group has many weak players (e.g., beginners or those who do not play serious chess), 1500 might not be very good compared to another system. P.S. A system where new players' games are rated as if they were 1200 rather than waiting until they have played 4 or more games before assigning a rating and only then rating their early games by that rating will have a bit of inflation if the new players coming in are weaker than 1200 strength by that pool's standards and deflation if the new players coming in are stronger than 1200 in the pool. P.P.S. My OTB ratings range from 2340 to 2500 across 3 national federations + FIDE, but I wouldn't make any inferences from that about the rating system here, either. |
| peppe_l 11/24/2003 07:06:38 [ report this post ] | Spurtus Message: I believe you misunderstood what Baseline meant... There is no way to convert GK rating to FIDE or any OTB rating. For example 1800 OTB player can be 1600 OR 2600 here, depending on how good he is at correspondence chess AND how much time & effort he is spending for his games. Same applies to 1150 OTB player - you are 1600+ GK but someone else is no more than 1050. The rules of correspondence chess and OTB chess are completely different. Same applies to internet & OTB etc. Tongue in cheek, you can compare differences between correspondence & OTB chess to differences between marathon and 100m spurt :-) |
| than77 12/05/2003 08:37:25 [ report this post ] | Differences in play/Number of games Message: It is worth noting that several psychological factors can effect tournament play that do not effect CC such as time pressure, being intimidated by your opponent, not wanting to draw with a child as young as your own ;-) etc etc. Also, I have played over 200 games here on GK in the last 14 months while I have only played about 50 OTB tournament games. That leaves a lot more room for my GK rating to reflect recent improvement. For instance, here in the U.S. some tournament results can take weeks or months to actually reach the chess federation and be calculated and in the case of an ongoing weekly tournament like a club or league championship usually the results aren't submitted until all of the league or club games are finished. This meant for me that I lost (or failed to gain) close to 100 rating points last summer due to a series of league games that I lost last April plus one lousy tournament result that took 2 or 3 months to to show up at the Federation. Ultimately my rating was lower at the end of the summer than at the beginning even though I had improved a lot in strength and had been studying consistently. At the end of the day, any rating below 2000 doesn't mean all that much and after all they are only supposed to serve as a guideline anyway. I know that I get a lot more satisfaction out of knowing that I played well in a given game than by knowing that my rating is or isn't where I wish it was. Over the long haul ratings improve as play does, whether OTB or here on GK. The main difference is (besides the time control, use of books etc) that I just don't have the time in life to play 200 + tournament games a year so my rating will always lag a little bit behind my ability. That's fine too, I have won several upset prizes this year! ;-) Cheers all, NW |
| thumper 12/05/2003 11:06:28 [ report this post ] | Inflated ratings Message: can happen here for several reasons. In my case, I've received over 50 unearned points from very strong players who likely would have beaten me, but instead resigned or timed out our games. I could only stop this from happening once when I happened to be online and noticed a player I had two games with resigning games wholesale. I beat him to the punch by resigning our final game before he could. Hehe, that'll show him! I prefer to actually EARN the points that I get, not be gifted points. This only inflates my rating not my relative playing strength. With this in mind, I think it would be a nice option to have the ability to decline the additional points gained from a timeout or resignation. The player timing out or resigning would still loose the points of course :) |
| dysfl 12/08/2003 09:15:11 [ report this post ] | Overall it's O.K within GK. Why compare? Message: I don�t believe that the ELO could be accurate or comparable to other system. It is basically a relative measurement, nothing more. If the rating used in GK is functioning as it was intended, that it�s all matters. I feel big pressure when I play anyone in here with 1600+ GK rating, and usually I lose. However, I also check their past game histories. If the guy/gal became 1550 by playing 1300-1400 players only, I treat it as about 1500-ish. I do not care about their OTB rating. In my limited experience, the rating in GK works just fine even there are some exceptional cases. |
Post a reply to this message: |
