USCF Rating Distribution

I used the USCF membership rating database to compute the rating distribution for “active” players with established (that is, not provisional) ratings. My definition of “active” is at least one USCF tournament result in the last year. The graphs are below.

I’m thinking about writing a script to automatically harvest the data and regenerate these graphs regularly. In my “copious spare time,” of course.

This entry was posted in Chess. Bookmark the permalink. Both comments and trackbacks are currently closed.

10 Comments

  1. Posted June 6, 2008 at 9:11 am | Permalink

    Very interesting! It seems that the US has a lot more players down the lower rated end than the UK. This I guess shows that chess is more popular in the US at the grass roots level.

    http://www.bluehorizonweb.com/blog/2008/03/general/ecf-grading-statistical-distribution/

  2. Steve
    Posted June 6, 2008 at 9:24 am | Permalink

    Hi Dave,

    Thanks for commenting! I was intrigued to see that someone else had the same urge that I did to look at the rating distribution.

    I plan to post some additional comments about the USCF distribution when I get a chance.

  3. Steve
    Posted June 6, 2008 at 9:30 am | Permalink

    I think the high number of low ratings in the USCF distribution is because of the relatively large number of scholastic players.

  4. Jordan
    Posted June 25, 2009 at 3:19 pm | Permalink

    Quoting Steve: “I think the high number of lower rated [USCF] players is because of the relatively large number of scholastic players.” This is probably correct, though the phrasing is akward…In particular what we are observing with the American distribution is more or less a bi-modal distribution super imposed onto a normal bell curve. The British distrubtion on the other hand, is a classic bell curve. Anyway, interesting graphs! I am glad somebody did this, so that we lazier chess players have a quick reference to see where our ratings put us in the scheme of things. Perhaps there is a way to seperate out the scholastic players? I know the USCF website does not list birthdates (unlike the FIDE website)…so this may be difficult. Anyway thanks again!
    -Jordan

  5. Steve
    Posted June 25, 2009 at 5:44 pm | Permalink

    Jordan—The USCF ratings database distributed to tournament directors does not have birthdates in it, or even membership classes.

  6. Steveross2852
    Posted May 19, 2010 at 5:46 am | Permalink

    These graphs, while illuminating, show only part of the story. The USCF is playing a double game of deliberately keeping ratings as low as possible while denying that they are doing so. In chess as everywhere in politics, the politicians lie with impunity since their political supporters don’t mind their lies as long as the people they support are in control.

    The above graphs show an average rating well below 1200 which is a beginner level rating. But in printed annual rating lists from recent years (since provisional ratings are included) an overwhelming majority of ratings, maybe two out of three seem to be below 1000! Arpad Elo envisioned a rating system where the “average” player would be rated about 1500 and during part of the 1980s, the average USCF rating actually approached 1500. In fact you almost never saw ratings below 1000 at tournaments during most of the 1980s. But although all modern rating systems owe Elo a debt of gratitude, they long ago discarded his idea of the average player being rated about 1500.

    Three things have depressed ratings: (1) a rating system top heavy with children playing initially in scholastic tournaments and given arbitrary low initial ratings so that they have “nowhere to go but up”, creates a “rating black hole” sucking points from more experienced players; (2) deliberate changes in the USCF rating system deflated adult ratings to FIDE rating levels whereas FIDE ratings had been traditionally much lower than USCF ratings; and (3) computers made chess and chess skill much more accessible but at the price of the loss of the scarcity value that chess skills fomerly had. Obviously the easier something is to obtain, the lower the value is that will be assigned to it.

    USCF policy decisions undoubtedly caused tens of thousands of able bodied adult members, disgusted with severe rating losses, to simply abandon playing in USCF events and not renew their membership. Without question the USCF loses hundreds of thousands of dollars a year if not more, due to this folly. Sadly, most masters if they will level with you, approve because of the “respect” they imagine the lower USCF ratings has gained them internationally, since they imagine that US ratings are now considered “more genuine” abroad. Many masters resent rank and file players anyway, considering them to be bums. These masters fail to realize that without interest and support from the rank and file players, the masters would be nothing. Like Ruben Fine during an even more bleak era for chess, they would have to give up chess for more lucrative but less satifying work.

    Most grownups who have stayed in USCF organized chess since the ratings were deflated think their ratings dropped steeply simply because they got much weaker, but a comparision of their games from the mid 1980s with their recent games shows a different story. If a player rated in the 1900-1950 range during 1985 played the same chess today, such a level of chess skill would only be good enough to maintain a rating in the 1700s at best. Such a player might even drop into the 1600s.

    But one of the biggest lies is the cliche that people go into politics with “good intentions”. Usually they go into politics, including chess politics, with bad intentions. The “volunteers” who run the USCF and who, with a few notable exceptions, run the chess tournaments and chess clubs, want to be big fish in a small pond. It is all about power and control and not at all about promoting chess. These mostly Class A and Expert level players who run most clubs and club tournaments, resent the more casual players and don’t want them around. They tell the weaker players to “play for fun” but take their own chess very seriously. They have “rabbit ears” for any real or imagined noise problem as if anyone but they themselves cared about their wholely insignificant chess results. Some chess clubs consider members who are casual players seeking a friendly game and who don’t play in club tournaments, to be freeloaders. The officers of such clubs sometimes make clear in thinly veiled ways that the casual non-tournament players are not really welcome in “their” club.

    With disfunctional chess federations that only value chess players who are masters or children while despising the adult players who used to be the backbone of organized chess, it is no wonder that chess is a dying. Why would anyone admit to non-chess-players that they play in chess tournaments in such an unforgiving rating environment, where the ratings of most players used to signify a beginner level and in the minds of many people such ratings still do! After all if you tell people you play in tournaments, the next question is likely to be “so what is your rating?”

    Even if we assume that it is good business to keep scholastic ratings low, there is no need, given the flexibility of mathematics in the hands of a competent mathematician, why adult ratings must remain at such deflated levels. That is bad business, especially for a federation like the USCF which has usually been strapped for funds over the years. The 1980s in the United States saw a signficant rating inflation but also saw the biggest sustained boom in adult interest in chess. Contrary to popular belief, the short lived “Fischer boom” of the early 1970s was not the real hey day of organized chess. The 1980s was.

    The end of the hey day of American chess coincided with end of the rating inflation that contributed to it at least in part. Should the main focus of organized chess be only to support the pet projects of power hungry elitists or should it be to popularize chess among all people?

  7. steveross2851
    Posted May 19, 2010 at 6:01 am | Permalink

    Anyone who was around in the 1980s knows that the USCF has deliberately kept ratings as low as possible while denying that they are doing so. The 1980s saw a significant and unplanned rating inflation but contrary to popular belief this was also the hey day of organized chess in the United States, not the short lived “Fischer boom.” Was this a coincidence?

    Now the USCF only values chess players who are masters or children even though there is no mathematical reason why adult ratings must be deflated just because scholastic ratings are low. The beauty of mathematics is its flexibility. But most masters in the United States want American ratings to be low because they imagine that their ratings will thus be regarded abroad as more “legitimate”. Yet tens of thousands of able bodied adults and their membership dues and tournament entry fees have been lost because their ratings have been severely devalued. For a federation like the USCF that is usually in poor financial condition that is bad business.

    It is futile to argue that people should not pay attention to ratings. They do care about them. Should organized chess primarily focus on the pet projects of power hungry elitists who only care about scholastic chess and FIDE events or should it focus on popularizing chess among all people?

  8. Tad
    Posted August 2, 2010 at 12:46 pm | Permalink

    Why is there even such a thing as a 400 rating? Is there some compelling reason our ratings system has to account for people who aren’t even sure how the horsey moves?

  9. Michael B player
    Posted August 2, 2010 at 8:39 pm | Permalink

    In 1986, my 1600 rating was at the 56th percentile. There were 30,000 players, 95% adults. Then I went into teaching 5th graders and quit playing tournaments. In 1995, after many children had started playing Chess all over America and getting ratings below 1000 at tourneys, another Percentile graph (26,000 adults and 18,000 children) showed my untouched rating at the 81st percentile. I hadn’t played a single game !
    Obviously, the kids were segueing the numbers. I know damn well I won’t win 4 out of 5 games at an open tournament. I asked U.S.C.F. to section off the children (no reply). I wonder how many home runs Babe Ruth would have hit if the Little League were combined with the Major League.
    U.S.C.F. Rating Percentile Graphs are useless.

  10. Nicholas Webster
    Posted October 26, 2010 at 4:01 am | Permalink

    I think the low ratings are because of the lack of integration between newly registered USCF players.
    For instance, my college recently began a new chess club where essentially no one had a rating at all,
    excluding myself who had a 4 game provisional rating from when I was 10 yrs old. (of 826)
    We all started out at the provisional 1000 rating and moved from there.
    After our first tournament I tied for 2nd but only moved up to 923.
    I play online regularly and have an active rating of 1600-1700.
    Our entire organization’s ratings are extremely low because we are not integrated with other players, so there is no controlled variable for computing our ratings besides my provisional rating from 12 years ago that says I am 826.