Site Ideas
13 Jan 09
14 Jan 09
2 edits
Originally posted by shortcircuitThat's why they typically wait to set the new floor until a rating class is maintained for X games [say, 30].
If they can implement the rating floors as you say, then I can see a definite value to them. I was under the impession that the floors would move as the players' rating move (i.e. lways 200 points below their playing level). My only question would be how to set it so the player who hits their alltime high and then goes legimately tumbling off the cart, ...[text shortened]... ed indefinitely to playing over their head because they had a run of good fortune and outcomes.
For example, after years as a 1600 player, player A has a hot streak and rises to 1710 after a string of wins. His rating floor does not move up until he has maintained a rating over 1700 for 30 games. If he goes tumbling back to reality before that time, then the floor never goes up.
14 Jan 09
Originally posted by SwissGambitI would be in favor of this then, under those parameters.
That's why they typically wait to set the new floor until a rating class is maintained for X games [say, 30].
For example, after years as a 1600 player, player A has a hot streak and rises to 1710 after a string of wins. His rating floor does not move up until he has maintained a rating over 1700 for 30 games. If he goes tumbling back to reality before that time, then the floor never goes up.
- Joined
- 07 Jun 05
- Moves
- 5301
14 Jan 09
Hiya Swiss
We have argued about this before, so I'll not go back over all the old arguments straight off.
3 issues I see with this, for your comment:
- Rating point Inflation
- administrative burden
- advantages implementing an actual floor compared to a Tournament Entry floor
I'll leave it there for now, as you know my position on all of these.
Cheers,
Gezza
14 Jan 09
2 edits
Originally posted by gezzaFor the benefit of the viewing audience, my response to the three points:
Hiya Swiss
We have argued about this before, so I'll not go back over all the old arguments straight off.
3 issues I see with this, for your comment:
- Rating point Inflation
- administrative burden
- advantages implementing an actual floor compared to a Tournament Entry floor
I'll leave it there for now, as you know my position on all of these.
Cheers,
Gezza
- Inflation from floors is negligible; and besides, the main point of ratings is to compare relative playing strength.
- Admin burden is negligible; there is no need to even consider floor-removal requests unless the player has been stuck on their floor for some time [and not due to absence].
- TER as implemented now has serious defects; after 100 days, the floor starts rising and people can sandbag all over again. Even if that flaw was fixed, TER does nothing to help with open invites, clan challenges, etc. Even if it were extended to clans, open invites etc. it still unfairly punishes a player who has a 'hot streak' and shoots up to an uncharacteristically high rating; also, it does not address the real problem at hand: the fact that certain players are grossly underrated due to mass-timeout/resignation.
Edit: Also, TER punishes the 2000+ player who goes absent by keeping them out of banded tourneys until they can get their rating back up. Floors allow them to immediately join tournaments against players close to their own strength.
14 Jan 09
Originally posted by IchibanovYour highest rating of 2004 would set a floor of 1804 I would guess, seems ok, and I'd even argue a rating floor of -150.
I'd be in favor if this idea. Not sure what would be a good number of games used to establish the rating floor, but in principal it sounds good.
That would put my floor near 1380 to 1430 based on a highest rating of 1580ish. I could live with either floor, and if I stalled there perhaps they can be evaluated every 6 months.
If they were evaluated by game, someone could make a ton of games to manipulate the floor lower, I feel it should be based on days, not games.
P-
14 Jan 09
Originally posted by SwissGambitI've been in these threads before also, and agree with all the points you've made.
For the benefit of the viewing audience, my response to the three points:
- Inflation from floors is negligible; and besides, the main point of ratings is to compare relative playing strength.
- Admin burden is negligible; there is no need to even consider floor-removal requests unless the player has been stuck on their floor for some time [an ...[text shortened]... Floors allow them to immediately join tournaments against players close to their own strength.
P-
14 Jan 09
1 edit
Originally posted by PhlabibitJust to make things clear, the USCF rounds down to the nearest hundred.
Your highest rating of 2004 would set a floor of 1804 I would guess, seems ok, and I'd even argue a rating floor of -150.
That would put my floor near 1380 to 1430 based on a highest rating of 1580ish. I could live with either floor, and if I stalled there perhaps they can be evaluated every 6 months.
If they were evaluated by game, someone could ma ...[text shortened]... ton of games to manipulate the floor lower, I feel it should be based on days, not games.
P-
Also, the floor is not set until the rating level is maintained over several games. In other words, Ichibanov would have to stay over 2000 for several games in a row. Once he did, his new floor would be 1800.
In your case, if you have maintained a rating over 1500 over several games in a row, your floor would be set at 1300.
The actual number of games that would constitute 'several in a row' is debatable. I suggest using 30 games, just to make sure that atypical 'flash in the pan' performances do not raise the floor. We should wait to establish or raise a floor until the player shows that they can consistently play at a higher level.
Also, players cannot manipulate their floor lower than it already is. Floors can only go up, not down [this is one reason that it is important not to move the floor up too soon]. The only exception is if a player is stuck on their floor for a long time while active on the site. In that case, they can petition an admin [or game mod] and the floor can be lowered. This is an extremely rare case. In my USCF experience, the only ones I see stuck on their rating floors are seniors who have lost much of their former ability due to aging.
Here's the logic I would use if programming rating floors:
After each rated game is finished, and rated:
1) Grab the player's last 30 ratings (the most recent 30 values from their rating graph), excluding provisionally rated games. If they have not yet played 30 non-provisional games, do nothing [skip remaining steps].
2) Find the lowest rating in the group; round it down to the nearest hundred.
3) Subtract 200 from that value.
4) If result is lower than the current rating floor, do nothing [skip remaining step].
5) If result is higher than the current rating floor, set it as the new rating floor.
The "30" is open to adjustment. Experience should show if needs to go higher, or lower.
14 Jan 09
Originally posted by SwissGambitVery Rusty cleared some of this up for me, we've been talking about the subject along with clan score ideas.
Just to make things clear, the USCF rounds down to the nearest hundred.
Also, the floor is not set until the rating level is maintained over several games. In other words, Ichibanov would have to stay over 2000 for several games in a row. Once he did, his new floor would be 1800.
In your case, if you have maintained a rating over 1500 over several ...[text shortened]... on their rating floors are seniors who have lost much of their former ability due to aging.
I had a USCF friend who used to play at 1900. He aged, and was stuck at 1700 for a floor for a year or more. He said something about it being evaluated, but I believe he passed on before they had time to check into it.
I wasn't suggesting ratings floors be lowered 'willy nilly'.
I wouldn't mind a floor near 1200 or even 1300. The only way I'll drop that low is either resigning all my games, or a devastating blow to the skull.
P-
14 Jan 09
Originally posted by SwissGambitI am very much in favour of rating floors.
- Inflation from floors is negligible; and besides, the main point of ratings is to compare relative playing strength.
That said, isn't this open to abuse?
Imagine player X has a floor of 1800 and starts 300 games with player Y. By letting them all timeout after the minimum required moves, player Y is effectively gaining a much higher rating that what he would gain without a rating floor. This would make "Golden Kings" (User 285560)much more effective.
14 Jan 09
1 edit
Originally posted by PalynkaEven if it did make this sort of rating abuse go faster [debatable - Player X still needs 20 provisional games plus 30 games over 2000 to get his 1800 floor] - this form of abuse is very obvious, and the abuser(s) would just get banned like GoldenKing did.
I am very much in favour of rating floors.
That said, isn't this open to abuse?
Imagine player X has a floor of 1800 and starts 300 games with player Y. By letting them all timeout after the minimum required moves, player Y is effectively gaining a much higher rating that what he would gain without a rating floor. This would make "Golden Kings" (User 285560)much more effective.
20 Jan 09
Originally posted by SwissGambitGood idea, and rec'd on my part.
It's high time RHP adopted rating floors similar to what the US Chess Federation uses.
The idea is this: After X games at a certain rating class, a player's rating cannot fall 200 points below that class.
Example: Player A's rating remains over 1900 for 30 games. This establishes a rating floor of 1700. His rating can no longer fall below 17 ...[text shortened]... that player has been stuck on their rating floor for a long time [and not due to absence!].
The only problem I see is a guy could throw his 30th game, and retain a lower floor.
You know the deal, rule breakers tend to use and abuse the rules and laws to their advantage etc etc.