Would you like being a CFC member more if a USCF-style rating floor was used?

[Vlad Drkulec]

Re: Would you like being a CFC member more if a USCF-style rating floor was used?

After 18 votes it appears there's not much interest overall in having a USCF-style rating floor for the CFC, at least among chesstalk respondents.

If this trend remains, this goes against an idea put forward in the CFC long-term planning committee's final report in 2012, i.e. that a rating floor might help increase the number of CFC members.

You would first have to know that the respondents on Chesstalk were representative of CFC members before arriving at such a conclusion. In any case the rate of response would argue against any conclusions one way or the other.

[Kevin Pacey]

Re: Would you like being a CFC member more if a USCF-style rating floor was used?

You would first have to know that the respondents on Chesstalk were representative of CFC members before arriving at such a conclusion. In any case the rate of response would argue against any conclusions one way or the other.

I was noting that the trend so far is not that supportive of a rating floor, rather than reaching a definite conclusion yet. Whether the poll is representative of both non-CFC members and CFC members is anyone's guess, as would be whether the low rate of response so far indicates that most chesstalk member viewers of this thread are uninterested in having a rating floor and also don't bother answering the poll, assuming they find any of the poll answer choices palatable. Instead of that, some viewers may be undecided at this point, still, but don't wish to vote for that as an answer to the poll question.

In any event, it's possible that many times an imperfect poll (in terms of knowing who it represents) which is done for free is better than no poll at all.

[John Upper]

Re: Would you like being a CFC member more if a USCF-style rating floor was used?

I voted against. But the idea might be tweaked to be useful:

One option would be to use a rating floor only as an anti-sandbagging measure, but not to calculate that player's opponents' ratings. e.g. if your rating is now 1780 but you once were over 2000, you would not be eligible to win an U1800 prize, but your opponent's ratings would be calculated on the basis of playing someone who has earned that 1780 rating (not someone who once-upon-a-time was over 2000).

A slightly more sophisticated tweak is to take account of how long ago the player's highest-ever rating was achieved: if it was years earlier, then it's reasonable to believe that player just got worse, and isn't sandbagging. That's what some big US tournaments do and what we were looking at doing at the 2013 Canadian Open, though AFAIK we didn't end up classifying anybody as ineligible on the basis of a suspiciously large recent ratings drop.

Pro: it's easy: TDs can do this now without any extra work or formula changes from Paul L: just look at the "highest ever" rating field in the CFC database.
Con: it is a financial disincentive to play an event when you are not eligible for a prize.


It might be worth noting that a player's highest ever rating is almost certainly not an accurate reflection of their best ever chess strength. Since chess players skills vary (like every other natural phenomenon), it is almost certainly true that each player's personal best rating is actually higher than their greatest skill level, and was achieved by playing at or near their best, but benefited from opponent's who (through whatever random the factors that cause variation) played worse than normal during the tournament that pushed that player to a new personal high rating.

Thinking that your highest ever rating is a fair indication of your real chess strength (current or past) is like a 10 handicap golfer who think that his best-ever round (the one where every hook and slice into the trees bounced back into the fairway) is the "real" measure of how well he plays. It's a pleasant self-deception, whether you're a duffer or Bobby Fischer c.1971-72, but not something I think we should encourage by creating a semi-fictitious rating.


BTW, every club-level chess player I know already employs a rating floor:
When I ask them what their rating is they all say two numbers: their current rating, and their highest ever rating. ;)

[Kevin Pacey]

Re: Would you like being a CFC member more if a USCF-style rating floor was used?

...
One option would be to use a rating floor only as an anti-sandbagging measure, but not to calculate that player's opponents' ratings. e.g. if your rating is now 1780 but you once were over 2000, you would not be eligible to win an U1800 prize, but your opponent's ratings would be calculated on the basis of playing someone who has earned that 1780 rating (not someone who once-upon-a-time was over 2000).

A slight practical hitch might be having to occasionally explain to a player (or perhaps a parent) after an event finishes that this is a rule for prizes, if it was not extremely common knowledge or well publicisized/announced (and even then... )

A slightly more sophisticated tweak is to take account of how long ago the player's highest-ever rating was achieved: if it was years earlier, then it's reasonable to believe that player just got worse, and isn't sandbagging. That's what some big US tournaments do and what we were looking at doing at the 2013 Canadian Open, though AFAIK we didn't end up classifying anybody as ineligible on the basis of a suspiciously large recent ratings drop.

I don't understand the part in boldface, unless you mean that's what happened in the US before a rating floor became part of the USCF rating system. [edit: or do some big US organizers not think the USCF rating floor is a good enough anti-sandbagging measure? On at least two occasions I've dropped something like 100 points CFC in only about a year, I suspect mainly because of either substantially altering my repertoire or not preparing my openings for my opponents nearly as much. In these cases, I did something unwise for my chess, rather than getting worse (in ability) IMO.]

Pro: it's easy: TDs can do this now without any extra work or formula changes from Paul L: just look at the "highest ever" rating field in the CFC database.
Con: it is a financial disincentive to play an event when you are not eligible for a prize...

In my opinion the 'Con' would far outweigh the 'Pro'.

[Roger Patterson]

Re: Would you like being a CFC member more if a USCF-style rating floor was used?

I voted against. But the idea might be tweaked to be useful:

One option would be to use a rating floor only as an anti-sandbagging measure, but not to calculate that player's opponents' ratings. e.g. if your rating is now 1780 but you once were over 2000, you would not be eligible to win an U1800 prize, but your opponent's ratings would be calculated on the basis of playing someone who has earned that 1780 rating (not someone who once-upon-a-time was over 2000)......

ahhh newbies. Once what was old is new again.

My memory may be wrong but I believe the reason the CFC maintains a "highest rating" field is because back in the day, somebody believed this was a real problem and proposed a "highest rating" field just so that organizers who were concerned about the alleged problem of sandbagging could specify just such a rule for prize money. But, nobody does.

I would guess the reason nobody uses the highest rating as a rule for prize giving is that sandbagging is not really a big problem whereas telling people they are not eligible for a prize consistent with their rating would be a huge problem. Not to mention, some level of unnecessary complications.

To be concrete, take the example of the recent BC Open. http://victoriachessclub.pbworks.com...n-Event-Report. The winner of the U2100 prize :-) would have a rating floor of 2100 under the proposal made by Kevin. Are you really going to tell me that either I am a sandbagger or that I should not be allowed to win that prize because hypothetically, someone else might try to sandbag? Tell 7% of CFC members that they can no longer win a class prize because of the greater good?

In short, if the intent of a rating floor is to use it for anti sandbagging, the tool already exists to accomplish that (the "highest rating" listing). If the intent is to allow people to remember their "glory days" then they already have that with class titles and the "highest rating" field. Kevin's discussion, in so far as he is proposing a change, only makes sense if he wants to gut the rating system so that people's rating don't sink to their natural level.