The ICCF is recognized by the USCF. It seems the ICC is too. How do these all connect together - or do they. Does you ICCF rating affect your USCF rating, etc.
What about the ICC? Someone like me who is a million miles from the nearest club, disabled, and cannot travel long distances would of course like to know. I have been playing with the ICC and the ICCF. Am I wasting my time (other than getting some practice in)?
Jim
What do you mean by “recognized”?
Alex Relyea
Well, actually, I was hoping someone would tell me. I figured the scores from games were “recognized” by the USCF kind of like the scores from the USCF are “recognized” by FIDE.
Jim
ICCF is the International Correspondence Chess Federation, recognized by FIDE, which handles international CC ratings and championships. ICC is a private company. They’re independent organizations. None of the ratings affect any of the others. The USCF has some dealings with ICCF over, e.g., invitations to international team matches. There was some talk last year about USCF taking over the U.S. branch of ICCF after Max Zavanelli retired, but I’m not sure how that worked out.
OK, when you sign up for the USCF, they say they recognize the scores of the ICCF. You get a free membership to the ICCF. Something is amiss. If nothing else, it is misleading. They (USCF) make the ICCF sound like the correspondence “arm” of the USCF.
Jim
I don’t believe there’s any such thing as a “membership” in the ICCF, except perhaps for national federations. The USCF is apparently now handling placement of U.S. players in international correspondence tournaments, in addition to the USCF’s domestic correspondence chess program (see main.uschess.org/content/blogcategory/82/397/).
But is sure does look like your ICCF rating is recognized by the USCF
Q: If I play Correspondence Chess (CC) will I get an official USCF rating?
A: Yes! As soon as you finish your first game, notify the Correspondence Chess Director (CCD) and your game will be rated. You will receive a Provisional rating (less than 25 games rated) and be published, in our annual rating list which appears on our website, when 4 or more games are on your record. After 25 games are rated your rating is recognized as being Established. After each game your result information will be posted in an online searchable database on our website. Your CC rating will also appear on your Chess Life magazine label.
main.uschess.org/content/view/7522/397/
Jim
Excuse me? I don’t see any reference there to ICCF ratings. Those are calculated and published by, well, the ICCF. That paragraph is talking about USCF correspondence ratings. I suppose they might use an ICCF rating to seed a USCF rating for a new player, but that’s going to be a very rare occurrence.
John,
I do not mean to argue – in fact I would love this cleared up. But this seems to follow standard IF/Then logic.
If the USCF recognizes the ICCF, and the ICCF is CC, THEN, it would seem ICCF play would be counted. How else can you play CC within the USCF?
Jim
OK, I found it. It took some searching. The ICCF is not the “official” arm of the USCF for CC. It is recognized which seems to mean absolutely nothing.
You can play CC according to USCF rules by postal. Ancient. The ICCF has an online version that keeps track of stuff.
But, while CC does count toward your rating, and the USCF recognizes the ICCF, you must play CC by USCF rules. Why recognize the ICCF if you are going to ignore the results. It is very confusing.
Jim
Because the ICCF runs international correspondence tournaments, some of which lead up to the World Correspondence Championship. You might as well ask why there are separate USCF and FIDE ratings. If you play in an ICCF tournament, it’s rated by the ICCF. If you play in a USCF correspondence tournament, it’s rated by the USCF (as postal, not OTB)… For that matter, I think there are still some other CC organizations around with their own rating systems.
Looking at this thread, I’m starting to wonder if we’re on the same page. I don’t mean to sound condescending, but let’s try this from scratch.
USCF over-the-board ratings (which are almost always referred to simply as “USCF ratings”) count only games played over the board in USCF-rated tournaments.
USCF postal ratings count only games played in USCF correspondence tournaments (there’s a list at the link I gave above). They have no effect on OTB ratings, and vice-versa. You can see both ratings (assuming the player has one) on the MSA (uschess.org/msa/).
ICCF ratings are earned in international correspondence events organized by the ICCF. You enter those through national affiliates of ICCF. For a long time ICCF-US was run by a guy named Max Zavanelli, but he retired recently, and it seems the USCF has taken over administration of this. ICCF ratings are entirely separate from USCF, though it’s possible some events are dual rated – you’d have to ask Alex Dunne or Mike Nolan about that.
ICC, the Internet Chess Club, is a private business which sells on-line play. Such games are not (and cannot) be USCF-rated, since a) there is no way to verify the playing conditions and b) the Delegates won’t stand for it. ICC runs its own rating system entirely unconnected with the USCF. My understanding is that these ratings are inflated and not terribly reliable, but your mileage may differ.
There SHOULDN’T be a difference between USCF ratings and FIDE ratings. It does chess in the US a big disservice. I have my Ph.D. in statistics and I have investigated this (I could, and probably should write a book).
Minor differences in the provisional ratings start the difference. But the biggest difference comes from the K-values used by the two organizations for higher rated players. It is stupid. International sports scored differently for the same thing. Crazy.
I guess in basketball the 3 point line is different, and international rules are different. But still, chess is chess. Oh well, the USCF uses a different floor and K-value. Because of that the USCF rating will be SLIGHTLY higher than FIDE most of the time.
Thanks all for helping me clear this up.
Jim
See above. There are a number of reasons why USCF and FIDE ratings might differ, but that’s not the point. They are entirely separate systems run by different organizations. The same is true of USCF postal and ICCF (and any other private CC rating systems still active).
FIDE and the USCF use different formulas to rate two mostly separate pools of players and events.
Statistically speaking, what reason is there for them to be the same?
I do not see why masters get treated differently in the two federations. Second, the pool can greatly overlap at the top.
Statistically, you would at least be comparing Apples to Apples.
Jim
If you look at the formulas, you’ll see there are differences. (And if the USCF goes to a Glicko-like formula, there will be even more differences.)
As to the players at the top, looking at the top 100 active players on the FIDE list (ie, players with FIDE ratable games in the last year), only 35 have EVER played in a USCF rated event, and only 13 of those have played in a USCF rated event since January of 2008. None of the top 15 players on the FIDE list have played in a USCF rated event since early 2001.
Nakamura is the highest FIDE rated player active in any recent USCF rated events at #16, the next highest player active in recent USCF events is Onischuk, at # 33.
Looking at the top 500 players, only 140 have ever played in a USCF rated event, just 60 of those since January of 2008.
Red Herring … not much to do with our discussion. The Glicko system is a crock. It had to floor probabilities because, well, they were not probabilities. Dr. Mark Glickman says in in bio as Boston U that he wants to be the “John Travolta of Statistics”. I will send him a disco ball. The more erratic your play, and the longer time between matches, the more of a chance of gaining rating points under Glicko-2. Dr. Glickman tried to get it patented, but was unsuccessful. Dr. Elo, on the other hand, wanted to keep chess in his best interest and calculated by hand all the Elo points for FIDE until 1970 just using an HP calculator or earlier, by hand.
The Glicko system uses a distribution that goes from .3 to 1.3 – huh? Even Dr. Glickman says it needs much more research. Dr. Elo on the other hand was a brilliant man who was also a great chess player (state champion several times). His system is based on a probability curve derived by calculus of empirical data.
OK, I did the same thing, went off on a diatribe. Your point was well made. Here is my main argument. Masters are less than 1% of any playing pool. More players, more masters. Dr. Elo’s system and distribution assured that (especially with the K-values).
I know the pools are different. But why would the USCF adopt a K-value system that causes rating inflation (Glicko is worse)? My only guess is to keep people interested. FIDE is spending money to bring Chess to the developing (formerly called 3rd World) countries. The USCF needs some seed money for clubs. I have no club withing 3 hours drive of where I live. With a little seed money, I would start a club.
But now we are way off topic. My main point is in red.
Jim
Why should their be different “standards” for
First of all, besides Dr. Glickman, who did I insult? Not really my nature. If you were offended, I sincerely apologize. Again not my nature. But it is my experience that discussions can get heated. Again, I apologize.
Dr. Glickman has published his system and comments. That opens him up to criticism. He will be the first to tell you Glicko-1 was published in a referred journal. The idea behind such journals is for other scholars to confirm or debunk the system. If it stands up to scrutiny, well then it is a great paper.
Now, you did attack me. I have a Ph.D. in Applied Statistics. Dr. Glickman has his Ph.D. in Health Science Quality Control. Believe me, I know what I am talking about. We may disagree on the APPLICATION, but I know what I am talking about.
Why attack me? Tell me this, what are the K-values for masters for both the USCF and FIDE? That is where a majority of the difference is – and which one is more inflationary? Discussion, not an attack.
You what I find insulting? 12 years of college, a bunch of research into this, and I am dismissed as an attacking and insulting person. I am 51. I still owe $29K in student loans! I will die before they are paid. I spent 8 years in grad school learning my stuff. I do know what I am talking about.
A better, but much more complicated system would be Finite Markov Chains if the goal is to predict with greater accuracy who will win. But it is quite complicated. Dr. Elo came up with a beautiful system, based on data. After he did the calculus on the data curve and came up with a distribution, floored the delta (to 400 Elo points difference) it really was and is a work of art. I have great respect for the man and the system. Why change the K-values?
It seems everyone is an expert except maybe, an expert.
Jim
I notice you didn’t actually answer any of my follow-up questions. Convince me that you understand how the current system works by answering them, and we can continue the conversation:
Do you not understand the reason that inflationary features MUST be built into our rating system to counter the deflationary ones? (at least if you want the overall system to be stable)
Do you understand the things that cause deflation in the rating system?
Do you have any evidence that the USCF ratings are ACTUALLY inflating?
I’ll add two more:
4. What ARE the values for K used for USCF and FIDE ratings of games played by masters (assume a rating of 2300, just for discussion)?
Right on Jim! I understand where you are coming from, and I can’t tell you how many times that when you give construvtive advice to make a situation better, people who disagree attack the messenger. Don’t give up, and keep posting!