|
« TAP TO LOG IN
|
GameKnot related: How accurate is gameknot chess rating
« Back to forum
| |||||||
| From | Message | ||||||
|---|---|---|---|---|---|---|---|
|
tactical_abyss 23-May-10, 09:40 |
 Deleted by tactical_abyss on 24-May-10, 14:34.
|
||||||
|
tactical_abyss 23-May-10, 10:40 |
Deleted by tactical_abyss on 24-May-10, 14:34.
|
||||||
|
ganstaman 23-May-10, 11:53 |
Gameknot could, for example, divide everyone's rating by 100 without really changing anything. If you then compare your gk.com rating vs your OTB rating, you'd see that your OTB rating was higher. But clearly that's not because you play better games OTB than at gk. Actually, doesn't England do something like that? Kingscrusher on youtube always talks about ECF or whatever of like 130. His rating in England of 130 is much lower than his FIDE rating (if he has one). This is no way reflects him playing worse in England than for FIDE events. |
||||||
|
algol 23-May-10, 12:31 |
BCF < 216: FIDE Elo = (BCF x 5) + 1250 Elo ≤ 2325: BCF = (FIDE Elo - 1250) / 5 BCF ≥ 216: FIDE Elo = (BCF x 8) + 600 Elo > 2325: BCF = (FIDE Elo - 600) / 8 These conversions formulas have to be re-evaluated as rating systems shift slowly compared to one another. The English Chess Federation was recently tinkering with their system for example. grading.bcfservices.org.uk A newer conversion is FIDE rating = 650 + 8 * ECF rating. |
||||||
|
algol 23-May-10, 15:32 |
USCF OTB versus CC There are 223 players with USCF >= 2000. Of those, 44 do not have a regular USCF rating. Of the remaining 179 players only 11 have an OTB rating which is higher than their CC rating (Dunne,Chudnovsky,Radomskyj,Stopa,Chessing,Bachler,Hehir,Musgrove,Godin,Lutes,Ferguson) Looking at those for which the CC rating is higher than their OTB rating only: The average difference is 468 rating points. The distribution of this difference is fairly flat to about 750 rating points and then drops of quickly. There are a few outliers with a difference of over 1000 rating points. Only 4 players of the 179 have their CC and ratings match within 25 rating points. 9 within 50 points and 17 within 100 rating points. |
||||||
|
baronderkilt 24-May-10, 01:14 |
A largely similar The main Difference would be this: Chess Rating is generally (Always, as far as I know of) open ended with no set maximum achievable. But the day of the open-ended IQ test derived from a Mental Age vs Physical age, Stanford type formula, result is a matter of long gone history (unless it has comeback in the last decade or so while I was watching TV instead). A thing that might be sought out by persons seeking to make a claim to "Highest" result. *** Otherwise the most widely used tests now are standardized within a specific set range. But this can vary from test to test. The two most common have been to cut off at 160 or at 200. Which would of course represent the top score recordable on each. And equivalent. (I don't say Equal,only in that a large enough test it may be possible that a perfect result and -1 might both score in the same % for instance. To guess at a number, I would suspect this may cause great consternation for 1 person in a million, and joy for several others. lol . *** The means used to make differently ranged tests comparable is to use the Standards of Deviation for the two. So for example a score at 3rd SOD on the one is comparable to the 3rd on the other, tho the actual number is different. Tho as algol is mentioning the shape of the curve can vary between them. But the statistician may joyfully devise fractional standards on any bad TV night . . . }8-D |
||||||
|
algol 24-May-10, 08:23 |
baronderkilt FIDE is considering tinkering with their rating system again and will have a conference about this www.chessvibes.com |
||||||
|
tactical_abyss 24-May-10, 14:31 |
Deleted by tactical_abyss on 25-May-10, 16:10.
|
||||||
| |||||||
