With few exceptions, most players who earn a money prize floor are improving players who fairly quickly breeze past that floor.
It’s the exceptions that justify having the money prize floors at all, though.
Here is an exception that might justify abolishing money floors: uschess.org/msa/MbrDtlMain.php?12234690
Amazing a player can stay on a 2000 floor forever and the USCF does nothing about it. I have similar players in my club where 1400 players have beaten or drawn a 2000 floor so many times it’s embarrassing. Everyone talks about skewed ratings. Well this is a perfect example of not letting a player keep a floor like this for 20 years and do nothing about it. I had another player who’s floor was 2100 and he stepped away from chess for awhile. No skewed ratings possible here. When he came back and joined my club again I found out the USCF arbitrarily dropped his floor to 1800. I’ve asked these players with high floors to contact the USCF and request a lower floor so they can possibly become competitive again. No go there as they love the expert rating and can just sail thru the tournaments not worrying about winning or losing. I have no problem with the prize based floors, but allowing these high floors to continue indefinitely without any improvement just does not fit the bill.
I would suggest to Mr. Hunt that his players enjoy being experts might stop playing if their floors were removed. A few players with a rating too high for their strength doesn’t skew the system as badly as other things do, and falls out pretty quickly. I doubt his story of US Chess arbitrarily dropping a 2100 floor to 1800 is true. Some players would rather have a high rating than be competitive for prizes. Some prefer the opposite. It is wrong to say that an OLM, for example, shouldn’t be able to choose which he prefers.
Alex Relyea
Money floors were created for the benefit of affiliates, especially those running events that pay out large cash prizes.
As a group, the OLMs probably represent the largest group of floored players whose current strength is well below their floor, but if you look at a graph of rating frequencies you’ll see significant spikes at all of the 100 point floor levels up to 2200. The spikes are less noticeable below 1400 but that’s probably because floors at 1300 and 1200 didn’t exist until relatively recently.
But good luck getting the OLMs to abandon their hard-earned 2200 floors.
I am not aware of any players who have had their floors lowered except by request or possibly a few cases of obvious data error. (Note that by EB action from several years ago, playing in a match while at one’s floor is considered to be a request to have one’s floor lowered.)
It is possible that we lost track of some floors for players who were not very active but had floors earned prior to the start of our detailed data (events rated on or after December 8, 1991), but whenever someone has brought such a case to the attention of the ratings manager in the office, that floor has generally been quickly reinstated. I’m also aware of a number of OLMs who never received their 2200 floor when it was earned but had it retroactively added to their records.
I’m actually more interested in your reaction to this post and especially this one on page 1 of this thread.
Bill Smythe
I’m one of those horribly honest people who taught all 8 of my kids that even white lies are terrible. My friend did in fact receive the drop in his floor and it did happen before 1991 without his knowledge. I only “suggested” to the players with high floors who complained when they continually lost to lower rated players that they might want to consider dropping their floors to rejuvenate the old spirit of competitiveness.
Regarding the 5 minute base time for regular, dual, and quick chess with 25 second(or more) delay (earlier in this thread). Chess is chess and if players want to play in that format let them. It’s their money and time. I see no valid reason to force TD’s to restrict tournaments to certain time controls. We do promote chess after all.
Prior to 2005, manual changes to ratings and floors were not recorded at all. Printed ratings reports from events did not show anyone’s floor, either. At times there were some codes attached to players’ ratings that sometimes indicated what someone’s peak rating (and thus floor) was, but that data was known to contain errors and inconsistencies. (And that doesn’t get into the issue of how floor policies changed over the years.)
Without knowing more facts (and the most relevant of those facts may be lost in the sands of time), it seems quite possible to me that this floor change could have been the result of an operator error or inadvertent data loss.
There have been cases where players have requested floors be added (or reinstated) where those floors could be verified by the office from published ratings data. For example, if the peak published established rating for some during the 1980’s was 1651, then there’s sufficient information to prove that the player should have a 1400 floor. However, if the player’s peak post-event rating was 1701 but that rating was never published because it dropped below 1700 before the next ratings supplement, then there is no evidence to prove the player deserves a 1500 floor.
Hanlon’s razor also applies: Never attribute to malice that which is adequately explained by stupidity.
In early 1997, the floors were dropped, even for OLMs. My rating dropped under 2200 after a so-so event. After a few more events, even getting a chance to play in the US Amateur, the rating magically went to 2200 again. The same thing seemed to happen to Mr. Cox’s 2000 floor. His rating dropped into the 1800’s and then magically became 2000 again. Now, how did that happen? Was it the USCF that made that rating change, or did one affliliate complain in order to protect its interests? Only one affiliate that I know of has its own rating floors.
The polite way to phrase it is that there was a lot of bumbling around with ratings floors in the mid 1990’s. When you make lots of ill-considered changes, mistakes happen.
The EBs of that time period were not exemplars of inspired decision-making.
Something goofy happened when 1xx floors were lowered to 2xx in 1997. Other floors—such as the 2000 money floor for the player I listed—were lowered by mistake.
Life Master floors at 2200 were also lowered, though I don’t know if that was by mistake or deliberately. Tom Dorsch, then on the PB/EB, got roasted for winning an Expert class prize at a major open around then, after his 2200 floor was lowered. But it was not just guys named Tom. It was all Life Masters. (Not sure if the “Original” was part of the title then.) At some point 2200 floors were reinstated for LM/OLM.
The player I mentioned had his 2000 money floor restored in 1998, thanks to a conversation at a CCA tournament, but CCA had nothing to do with it.
Point is, the guy is in his 60s and has been playing at 1700-1800 level for many years. That will only go in one direction at this point. He got a 2000 money floor from a fluke result at a tournament in Las Vegas in the '90s. He told me the secret was he did not gamble that time—but if he had ever been over 2000 before that, it was just barely and not for long.
Now it’s 20-some years later, he is a senior citizen and still active in rated chess. I doubt he will play big-prize tournaments at this point and even if he did his chances to win a large cash prize in Class A or U-2000 are not good. But he should be allowed to enter a section that reflects his actual playing strength. And if he were super-active in rated club play, he could have a non-trivial inflationary effect on the ratings of those he played regularly.
But we stopped him taking that second bite at the apple.
Keep in mind that prior to 2005 US Chess had NO WRITTEN RECORDS of floors other than the floor in the computer system, and the floor in a member record did not indicate why the floor was there or when it had last been changed, so there was no way to easily distinguish between a peak rating based floor and a money prize floor, an OLM floor or some other kind of floor that might have existed. And I believe there was a time when peak rating based floors went all the way up to 2200 if not higher. (Money prize floors may have, too.)
Thus when they were fiddling with the floors (200 points ending in 00, 100 points ending in 00, floating, then back to 200 points, I think, though others may know more than I do), each policy change could have resulted in updating the floor information in someone’s member record and overwriting the previous floor information.
Update: While there were some records kept back in the pre-2005 era, when I started on the redesign of the ratings system, I was told those records were no longer available.
I would have written the rule as: A minimum base time of 30 mins is required to be regular rated. Personally, I do not consider 25 5 and 30 0 equivalent; 25 5 is effectively g/28 for 40 moves; it should be quick rated only.
By extension 25 5 and 5 25 are not equivalent even under the rules. 5 25 for 40 moves is 5 mins plus a 1000 seconds about 22 mins clearly quick.
Most people who compare time controls with delay use 60 moves as the break-even point - not 40.
Well, that horse has already sailed. G/25 d/5 was explicitly made legal as regular, because many organizers were already doing it, and because of support from the scholastic community.
A more reasonable requirement is the one we’ve already been talking about. Minimum mm should be 25 minutes, and minimum mm+ss should be 30.
Bill Smythe
True but a much higher percentage of games go around 40 moves so I think comparing time controls based on a 40 move game makes sense. I am changing the time control at the quad tournament I run once a month from G/45;inc15 to G/50;inc15 so players will get around 60 minutes total for most games.
Let us know how that goes - which time control your players prefer and how your total time for the event works out.
I doubt the extra 5 minutes messes things up and if your players prefer it then that’s great.