CFC Provisional Ratings

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • CFC Provisional Ratings

    Hi, I'm wondering how the provisional rating equation works. I'm having difficulty understanding how the (W-L) section ties in to the final result.
    I recently played a tournament where I lost nearly 200 rating points (from 2100 down to 1918), and I'd like to confirm the rating calculation was correct.
    My new rating says I have played 12 rated games in total, however since my previous total was 7 games, and I just played 5 more, it appears maybe some games
    which should not have counted towards my new rating have been added incorrectly. Specifically, 2 games where I played opponents who's ratings
    were more than 400 points below my rating. (1300ish, and 1655.) Should my total games not equal 10 now? I appreciate any help on this question!


    413. Matches.Individual matches are rated with the following restrictions:

    a) the maximum change allowed as a result of a match is 50 points;
    b) match results must be dated and signed by both players;
    c) the rating difference between the two players must not exceed 400 points.

    414. Equations.The equations used to calculate ratings are as follows:
    a) For previously unrated and provisionally rated players, the performance rating is:

    Rp = Rc + 400 (W - L) / N

    where Rp is the performance rating,Rc is the average rating of the player’s opponents,Wis the number of wins,L is the number of losses,N is the total number of
    games played.


    Thanks,
    Ken Ferguson

  • #2
    Note that the rules for matches (one versus one player; multiple games) are different (so I think 413 does not apply in your case since you stated it was a "tournament")

    In any case, I am editing this reply to include what I should have added: I'll wait for Bob Gillanders to weigh in on this for you... best of luck.
    Last edited by Kerry Liles; Saturday, 29th October, 2022, 06:27 PM.
    ...Mike Pence: the Lord of the fly.

    Comment


    • #3
      Thanks Kerry, I should have added that the tournament arbiter told me the 2 games where there was a +400 point difference would not be rated, or (later he said) at least the points would be
      very minimal. I won vs. the 1350, but lost vs. the 1655. In my other 3 games: -1960, =1707, =2043. The CFC site says my performance rating was 1663, and to me that seems low? Perhaps my game vs. 1655 was given full value?
      Last edited by Ken Ferguson; Saturday, 29th October, 2022, 06:31 PM.

      Comment


      • #4
        While your rating is provisional there is no "protection" for beating low rated opponents (this goes the other way as well). Once you have your 25 games in, there is rating protection for opponents > 400 points away from you.

        The arbiter was incorrect, in this case - but the calculating of ratings is not really their job.

        Comment


        • #5
          Hello Ken

          The calculation of your performance rating from the last event is correct:

          Average rating of opponents = 1743
          Number of wins = 1
          Number of losses = 2
          Number of games played = 5
          (W - L) / N = (1 - 2) / 5 = - 0.2 x 400 = - 80
          Performance rating = 1743 - 80 = 1663

          The 400 point difference has nothing to do with this formula. Additionally, since you are considered a provisional player (i.e., played fewer than 25 games), your new rating is calculated by applying this same formula cumulatively to all twelve games you have played so far. The formula is heavily dependent on the average rating of your opponents. For the three events you've played so far that has been 2280, 1817, and 1743, hence the large changes in your rating.

          Comment


          • #6
            Thankyou very much for your replies, Stephen and Fred! This clarifies the situation. Now I understand that
            playing in events where my opponents have a lower average rating can effect my rating, and my rating will not be protected until after 25 games. I am slightly disappointed to hear this news, but I had a good time at the tournament, and I’m glad to be playing again after 20+ years away from the scene! Thanks again for taking the time to help answer my questions.
            Regards,
            Ken Ferguson

            Comment


            • #7
              As the new CFC Rating Auditor, I should make a comment. First thank you Stephen and Fred for clarifying the situation. This anomaly in the calculation of Provisional ratings was first brought to my attention by David MacLeod 49 years ago on the bus home from the Toronto Chess Club. He told me that for my provisional rating it is best to play against as high rated players as possible, even if you lose. Back in those days, you used to be able to pay a fee and have your rating calculated from scratch. I think Dave was one of the few people who did this. As Fred points out, just play lots of rated games and everything will come out in the wash.

              Comment


              • #8
                Perhaps the CFC Provisional rating system should be reviewed from scratch. I believe that it has been in its current form for at least 50 years (looks pretty much the same as when I started).

                25 games perhaps was appropriate in the good old days, but the new follks seem to getter better faster as they have probably played hundreds on on-line games during that 25 game span, etc.Perhaps get it more in tune with FIDE which effectively has no provisional system after a rating is established, other than the K-40 for 30 games.

                Locally, we have a player (173468), who played in his first OTB event in July - with a low resultant 7 game provisional. Fast forward to November - he is performing at a 1950+ level (games 8-11), but carries a sub 1600 prov. rating after 11 games. If he performs in the current 5 rounder (In the first round, he defeated a bewildered 1900 who wondered if we were using accelerated pairings; we know the underrated player from online and local non-CFC play) - at, say, 2100 level, his 16 game provisional would only be 1748. For illustration purposes, let's say he performs at a 2300 level over games 17-25 - final provisional is 1927. Basically he would be running at about 400 below the 'true' rating for most of the provisional period, meanwhile his opponents may be suffering a bit of deflation. I suspect this scenario is playing out in other areas of the country with fast progressing juniors.

                Comment


                • #9
                  Originally posted by Tony Boron View Post
                  Perhaps the CFC Provisional rating system should be reviewed from scratch. I believe that it has been in its current form for at least 50 years (looks pretty much the same as when I started).

                  25 games perhaps was appropriate in the good old days, but the new follks seem to getter better faster as they have probably played hundreds on on-line games during that 25 game span, etc.Perhaps get it more in tune with FIDE which effectively has no provisional system after a rating is established, other than the K-40 for 30 games.

                  Locally, we have a player (173468), who played in his first OTB event in July - with a low resultant 7 game provisional. Fast forward to November - he is performing at a 1950+ level (games 8-11), but carries a sub 1600 prov. rating after 11 games. If he performs in the current 5 rounder (In the first round, he defeated a bewildered 1900 who wondered if we were using accelerated pairings; we know the underrated player from online and local non-CFC play) - at, say, 2100 level, his 16 game provisional would only be 1748. For illustration purposes, let's say he performs at a 2300 level over games 17-25 - final provisional is 1927. Basically he would be running at about 400 below the 'true' rating for most of the provisional period, meanwhile his opponents may be suffering a bit of deflation. I suspect this scenario is playing out in other areas of the country with fast progressing juniors.
                  25 games is mathematically sound but your point is quite valid - not just juniors but adults as well. A newbie starts off in the U1000 and all those 3 digit rating games from their first tournament are a millstone around their neck even once they are experienced and better players for probably a year or more. No bonus points either to reflect any rapid change in strength.

                  Comment


                  • #10
                    I think deflation is mostly a problem at very low ratings. It would make sense to me that at least for low ratings (say U1400) if your performance rating exceeds your present rating, the performance rating should be your new rating (up to 1399), assuming the newest performance is based on at least four games.

                    So, for example, someone has a rating of 1212/15 and has a performance of say 1522/4 the new rating is 1399/19. It's possible they might be a bit over-rated but it's almost certain their strength will catch up to their rating.

                    In examples where the new rating would be higher under the old system, the old system would prevail.

                    Unless this has already been addressed since back in the old days when I input crosstables?
                    Last edited by Tom O'Donnell; Wednesday, 9th November, 2022, 06:30 PM.
                    "Tom is a well known racist, and like most of them he won't admit it, possibly even to himself." - Ed Seedhouse, October 4, 2020.

                    Comment


                    • #11
                      Originally posted by Tom O'Donnell View Post
                      I think deflation is mostly a problem at very low ratings. It would make sense to me that at least for low ratings (say U1400) if your performance rating exceeds your present rating, the performance rating should be your new rating (up to 1399), assuming the newest performance is based on at least four games.

                      So, for example, someone has a rating of 1212/15 and has a performance of say 1522/4 the new rating is 1399/19. It's possible they might be a bit over-rated but it's almost certain their strength will catch up to their rating.

                      In examples where the new rating would be higher under the old system, the old system would prevail.

                      Unless this has already been addressed since back in the old days when I input crosstables?
                      Sounds like a good idea. Of course COVID really made this problem much worse.

                      Comment

                      Working...
                      X