|
< Earlier Kibitzing · PAGE 1 OF 2 ·
Later Kibitzing> |
| Sep-01-15 | | dumbgai: According to this Chessbase article, this kid now has a 2601 FIDE rating (the youngest ever to reach 2600+). Yet he has no games in the database here! http://en.chessbase.com/post/septem... |
|
| Sep-01-15 | | dumbgai: A further examination shows that John Burke achieved his rating gains thanks to a k-factor of 40, gaining between 60 and 120 rating points in each tournament he's played. He's defeated a couple of GMs in his career, but mostly his wins have come against 2300-level players. One suspects he'll have trouble maintaining his 2600 rating as he faces tougher competition. |
|
Sep-01-15
 | | MissScarlett: Any idea what a k-factor is? |
|
| Sep-01-15 | | supertimchan: Why does he have a K-factor while nobody else does? |
|
Sep-01-15
 | | perfidious: <supertimchan: Why does he have a K-factor while nobody else does?> Immense talent carries its own reward.
A spot of baksheesh never hurts, either. |
|
| Sep-01-15 | | markz: <supertimchan: Why does he have a K-factor while nobody else does?> FIDE changed the k-factor rules recently so that the young kids have more chances to take advantage of k=40. The "old" kids didn't have the same advantage when they were young. It is unfair. FIDE sucks |
|
| Sep-01-15 | | sonia91: <markz: The "old" kids didn't have the same advantage when they were young. It is unfair. FIDE sucks> I agree with you, I hope FIDE will revise the k-factor for under-18. It's worth noting that this kid played only in American tournaments, while Wei Yi gained his 2600+ status playing in international tournaments abroad. |
|
Sep-01-15
 | | alexmagnus: Another fix to this "bug" would be applying the 2300 border for k=40 not based on the official rating but on the live one. That is, once you reach 2300 live your K gets down from 40 to 15 or whatever it is now for adults with the same rating. |
|
| Sep-01-15 | | markz: K-factor: new ELO = old ELO + K * P
Normaly, 2400+ players K=10, all others K=20. But young kids can have K=40 depending on rules |
|
| Sep-01-15 | | Maatalkko: If Burke was actually at a 2600 level, he could have scored 8.5/9 at those tournaments and been suddenly rated 2800+. |
|
| Sep-01-15 | | AzingaBonzer: Notably enough, Burke's actual performance rating during the tournament that got him up to 2600+ was only 2389. The combination of high k-factor and late updating by FIDE is the only reason his rating got as high as it did. |
|
| Sep-01-15 | | notyetagm: <AzingaBonzer: Notably enough, Burke's actual performance rating during the tournament that got him up to 2600+ was only 2389.> How in the world can you get a 2600 rating with a 2389 performance??? A normal person would require like a 2700-2900 performance to get a 2600 rating, assuming you were like 2500 or 2550 or such. |
|
| Sep-01-15 | | markz: <notyetagm: How in the world can you get a 2600 rating with a 2389 performance??? A normal person would require like a 2700-2900 performance to get a 2600 rating, assuming you were like 2500 or 2550 or such.> How in the world you can't understand? In the last two months, when his true performance clearly below 2500, he jumped from 2258 to 2601 easily because his k-factor was 40, which is a "bug" (<alexmagnus>quote) in the new FIDE elo system. It will be much much more difficult to jump from 2500 to 2600 because k-factor is 10. You can check his performance on FIDE website. By the way, it is a very good example to show how stupid the FIDE is to change the rules without foreseeing the bugs. |
|
| Sep-02-15 | | supertimchan: <Notably enough, Burke's actual performance rating during the tournament that got him up to 2600+ was only 2389. The combination of high k-factor and late updating by FIDE is the only reason his rating got as high as it did.> So... If he performed even better like 2400+, he'd be one of the top 10 in the world? |
|
| Sep-02-15 | | supertimchan: <Another fix to this "bug" would be applying the 2300 border for k=40 not based on the official rating but on the live one. That is, once you reach 2300 live your K gets down from 40 to 15 or whatever it is now for adults with the same rating.> Does that mean if he had more tournaments where he performed like 2400 before this rating list is published, he'd be 2800+? |
|
| Sep-02-15 | | supertimchan: By the way, what's the motivation for introducing K-factor? |
|
| Sep-02-15 | | markz: <supertimchan: So... If he performed even better like 2400+, he'd be one of the top 10 in the world?> Yes or no. I guess he need to perform like 2600, he'd be one of the top 10 in the world. Imagine that if k=400, he can be 3000+ even 4000+ overnight. <supertimchan: By the way, what's the motivation for introducing K-factor?> In theory, there is always a k-factor. k=1 will be too slow, k=100 will be too fast. It is all about math and rules. ELO is a mathematical system. |
|
| Sep-02-15 | | dumbgai: The reasoning behind giving a higher k-factor to low rated players and young players is that in case of a mismatch between rating and playing strength, it allows the rating to "catch up" more quickly. Higher k-factor makes ratings more sensitive to recent results, because the player gains more points for a win and loses more points for a loss. At the high end of the rating scale (and for non-junior players) the k-factor is reduced to help stabilize the ratings. |
|
| Sep-02-15 | | Helkost: Hello Everyone. Can someone explain to me of the k-factor works? Have there been any changes recently? I knew that, for a k-factor of 40 to be applied, he should: -have played less or equal to 30 games (he has played much more than that, after 30 games, my understanding is that the k-factor has to be lowered to 20) -have a published rating of less or equal to 2400.
Am I missing something? |
|
| Sep-02-15 | | sonia91: <Helkost> 3. Under-18 players have k=40 until they reach a published rating of 2300+ |
|
| Sep-15-15 | | Helkost: so, just out of curiosity, how did the boy make a jump from 2538 to 2601 in only 9 games in Aug.? That's 60+ points when his published rating was already above 2300. |
|
| Sep-15-15 | | markz: Those 9 games were played at the end of July (not Aug) when his published rating was below 2300. |
|
| Sep-15-15 | | Namawiz: This K=40 is distorting the rating system already. Too many overrated new players being added to the rating lists and a growing number of overrated kids emerging with each new ratings list. Eventually, ratings just like titles such as CM, FM and IM, will become meaningless numbers and that no longer serve as measures of playing strength as they have been all along. FIDE must go back to the previous rating system. It worked just fine. |
|
| Sep-18-15 | | AzingaBonzer: FIDE should reduce everyone's rating by 3.5%. That should stabilize things. (Semi-serious suggestion; do you like my alliteration?) |
|
| Sep-19-15 | | Whitemouse: Chessgames.com only has one game from this 'youngest ever to reach 2600'? |
|
 |
|
< Earlier Kibitzing · PAGE 1 OF 2 ·
Later Kibitzing> |
Kibitzing
