GameKnot: GameKnot related, Rating floor

« TAP TO LOG IN

GameKnot related: Rating floor

« Back to forum

Pages:

« Previous

From

Message

tactical_abyss
23-Nov-10, 14:55

Ganstaman,
Its just that the proximity from 2399 to 2400 is so close,almost magnetically,that yes,it should be counted and rounded up to 2400...thus a floor of 2200.Thats why I mentioned a grace of "perhaps" 2 or 3 points...maybe more,but within reason....I just happed to pick that # out of the hat.I basically have not agreed with alot of the USCF rules over the last 30 years,just like Magnus Carlsen not agreeing with some FIDE rules.It is his right and privilege as well as any other FIDE player or USCF player to disagree with anything they so choose.So,I do not expect everyone to understand my thoughts.I simply do not agree that a 2399 player should have a floor rating of 2100,period.Its the full additional minus 99 points that I disagree with.This is my opinion,take it or leave it.

baronderkilt
24-Nov-10, 20:24

It is clear ...

Anything over 47 pts should be rounded up. As those of my generation know, "47" is the Universal answer. The case is clear . . .
}8-D

baronderkilt
24-Nov-10, 20:36

BTW ... assuming it is still in active use ,

USCF has an added "protection" for those over 2100 losing rating, in that their games are counted at 1/2 K instead of full K. Losses hurt less. Of course this "protection" is also what makes it harder to go from 2100 to 2200 since games count half as much, wins also.

It makes me wonder if that is inflationary or deflationary. It used to be also, that USCF gave Bonus Points in Rating during the 70's, which helped a player get up to true rating faster if they had a really good tournament, and went up a certain amount then they would get Bonus Added. And USCF ratings became inflationary compared to FIDE during that time. EG Fischers 2800 USCf vs 2700's FIDE.

I wonder how they compare now? Since Bonus Points were stopped and FIDE seems to be inflated 100 from the 70's ratings. Or did Bonus ever resume? Seems to me they should be much closer now, but I have no idea if so. Can anyone say or show some comparisons?

vermonty
30-Nov-10, 20:18

Copying USCF isn't necessarily good policy!

Adopting the USCF rating floor rule is a political or economic decision, not a statistically valid
decision. The concept of rating floors is valid, but the floor should be 200 points below a
player's highest established rating (also called "peak rating"). The rounding down -- as much as
99 points down -- was a political/economic decision by USCF because its ratings were inflated,
and the rounding down was designed to reduce the number of rating points in the system.
It makes no sense mathematically for a player whose highest established rating was 1998 to
have the same rating floor as another player whose highest established rating was 1901. Both
should have a rating floor exactly the same number of points below their peak rating.

kingdawar
01-Dec-10, 02:31

vermonty

It's clearer this way with easy-to-understand levels.

kingdawar
20-Dec-10, 05:21

Someone hit their rating floor:

gameknot.com

It works

nathanman22
08-Jan-11, 12:54

Partially agree

i partially agree with the rating floor. In the case of timeouts its good. . . But if someone has a bad streak of play then he should face a rating deduction. A rank of 2000 means less if more people now have it due to the rating floor. . . Maybe we should just not penalize time outs and remove the rating floor. . . I personally faced computer failure leading to time outs dropping my rating. I sat out a year disenchanted by chess and my horrible rank. I finally pulled it together and im back and slowly returning to my capacity. . .wish there was someway to differentiate between timeouts or losses for sandbagging and timeouts or bad play due to life situations.

kingdawar
08-Jan-11, 15:49

Of course timeouts are and should be penalized as if they are normal losses

levellerlevvie
21-Jan-11, 04:10

What if timeouts are interrupted by a regular loss

For eg ... one plays 100 games and starts timing out a few dozen until the rating is just above the rating floor. THen that person looses a game in a regular way by being check mated. Will he end on his rating floor or just below it?

If he ends below it and starts timing out again ... will he end on the very rating bottom like before or not?

But I guess the crucial question is whether the rating floor triggers when you have a regular loss ...

ganstaman
21-Jan-11, 09:11

"Will he end on his rating floor or just below it?"

As per the definition of a rating floor, you can never get a rating below it. Ever.

kingdawar
21-Jan-11, 11:31

levell

Check the link provided above, he doesn't fall below 1500

kansasjayhawk
21-Jan-11, 11:52

Rating floors are not necessarily "hard" with the USCF

Insofar as that one can petition the USCF for them to be removed, but I wonder how many players would want to do so. If for some reason I hit my USCF floor of 1600, I wouldn't want to go any lower (even if it would open me up to win prize money at tournaments, which isn't a good reason to play).

Somebody mentioned that they combat deflation, but the way they do so is by artificially inflating the rating pool. For example, near where I live, a formerly strong player had a stroke and could no longer see the entire board at once, so while he was floored at 1600, his actual strength may have been 1200-1300. In effect, he was giving away rating points to players who might not have beaten a 1600-strength player but were getting credit for easy wins over somebody of that rating.

Pages:

« Previous

« Back to forum