Question on dynamic ratings

Discussion in 'Adult League & Tournament Talk' started by raiden031, Feb 2, 2007.

  1. raiden031

    raiden031 Legend

    Aug 31, 2006
    If a 5.0 computer rated player lost in an open tournament to a 3.0 computer rated player, would their rating immediately drop a level or more? What do you think would happen? Lets say they are both in the middle of their level (4.75 and 2.75 respectively).
  2. Jack the Hack

    Jack the Hack Hall of Fame

    Jan 18, 2005
    The NTRP algorithm was explained in this thread from a while back:

    The article that was cited in the thread appears to be gone now, but it went into great detail about the formula, which appears to look like this:

    (PRD - CRD) divided by 2 = RDD (which is reapplied to the winner and loser's rating number)

    The PRD is the "player rating differential", which is calculated by subtracting the difference between the higher rated player's NTRP and the lower rated player's.

    The CRD is the "computer ratings differential", which is a chart the USTA has devised which shows the number that should be used in the formula depending on the score from the match. This chart is reported to assign the following numbers for the match score:

    Table 1
    Match Results = CRD

    7-6, 7-6 = 0.06
    7-6, 6-4 = 0.09
    6-4, 6-4 = 0.12
    6-3, 6-4 = 0.15
    6-3, 6-3 = 0.18
    7-5, 6-2 = 0.18
    6-2, 6-3 = 0.21
    6-2, 6-2 = 0.24
    6-3, 6-1 = 0.24
    6-1, 6-2 = 0.265
    6-3, 6-0 = 0.27
    6-2, 6-0 = 0.295
    6-1, 6-1 = 0.295
    6-1, 6-0 = 0.325

    The RDD ("ratings differential discrepancy") is the difference between the PRD and CRD, divided by two. This number is then applied to the winner and loser's rating.

    I'm not sure if this formula is 100% correct, but it seemed to work in the example in that previous thread.

    In your hypothetical scenerio, the exact calculation would depend on the score. However, let's say the 3.0 beat the 5.0 by a score of 6-4, 6-4. Here is how that would effect their ratings:

    Lower rated player's rating (2.75) minus the higher rated player's rating (4.75) = 2.00, which is the PRD.

    With a 6-4, 6-4 score, the CRD would be 0.12.

    2.00 - 0.12 = 1.88 divided by 2 = 0.94 for the RDD, which is added to the winner's rating and subtracted from and loser's.

    Winner = 2.75 + .94 = 3.69
    Loser = 4.75 -.94 = 3.81

    I really wish that I had the original source article that explained the formula because I could then know if I've done this correctly. However, it does look like that one loss could potentially drop a 5.0 down to a 4.0 dynamically and raise the 3.0 to a 4.0. Such a scenerio is very, very unrealistic, but I suppose it's possible. On the other hand, I've also heard that there is a piece of programming within the NTRP computer that throws out abnormal scores like this to eliminate blatant manipulation of rating via thrown matches or a one-time injury. Since the system is supposed to give you a year end rating that represents your average level of play, I could see how they might throw out your two highest and two lowest scores within TennisLink to balance the rating appropriately (but I don't know if they do this for sure).
  3. cak

    cak Professional

    Feb 23, 2004
    The other little piece of information included in the formula was if the players were too far apart, I think it was 1.0 apart, the scores would be ignored. So a 4.75 playing a 2.75 wouldn't have the score factored in.
  4. ras

    ras Guest

  5. Mark Jensen

    Mark Jensen New User

    Feb 3, 2007
    Wow, brilliant link. This might generate a bit of discussion ;)

    I'm saving it for a read later tonight.
    Right now I'm actually off to hit a fuzzy yellow imagine that!
  6. Mark Jensen

    Mark Jensen New User

    Feb 3, 2007
    Assuming this is a factual account of dNTRP

    What strikes me is, that as the league year progresses, a dNTRP rating is an average number...weighted primarly from your last 3 matches.

    As I understand it:
    1. At the beginning of the league year you start off with your year-end-rating (YER)

    2. After your 1st match (you won), your adjusted-winners-rating (AWR#1) is calculated and the 2 ratings are averaged to give you your 1st dNTRP rating of the year: (YER + AWR#1)/2 = dNTRP#1

    3. After your 2nd match (hey, you won again), your AWR#2 is calculated and now the 3 ratings are averaged to give you your 2nd dNTRP of the year: (YER + dNTRP#1 + AWR#2)/3 = dNTRP#2

    4. After your 3rd match(oops, you lost, here comes a new acronym...the ALR or adjusted-losers-rating), your ALR#3 is calculated and now the 4 ratings are averaged to give you your 3rd dNTRP of the year: (YER + dNTRP#1 + dNTRP#2 + ALR#3)/4 = dNTRP#3

    Now this could go on for the whole year and by the end of it your latest dNTRP would be an average of the total number matches you had played in Adult and Mixed (for me that would have been 28 matches in 2006 plus Tourneys).

    But the USTA has another nuance in the algorithm. Once you hit 4 NTRP ratings, you start to drop the oldest number from the mix. In other words, after your 4th match the YER is dropped from the average in favor of the 4th match result....after the 5th match, the dNTRP#1 is dropped in favor of the most current result...and so on.

    That's cool (in my book).
    It's not "you're only as good as your last match" but more like "you're only as good as your last 3 matches and and your oldest dNTRP".

    Well, that's not very poetic, but accuracy counts too. :)

    What are your thoughts? (On this point or any other's in the document).

    See you on the courts.
  7. Cindysphinx

    Cindysphinx G.O.A.T.

    Aug 31, 2006
    For those of us who don't do math, this means that players shouldn't worry that they'll be DQ'd because they improved at a normal rate. It would take a sudden spike performance to cause a problem, right?
  8. Raiden.Kaminari

    Raiden.Kaminari Semi-Pro

    Nov 21, 2006
    No, it would be a consistent spike in performance to really cause a problem.

    By the way, the table that was in the document now really only applies to Senior (50+ year olds). For players lower than 50, they have a new algorithm.

    Tournaments are usually applied to year end rating, so unless the 3.0 player also played league, he would not feel the effects of beating a 5.0 until the year end ratings.

    The document doesn't also mention the weight / history of a particular rating.

    Let's say, someone has been a career 3.0. Over 100 matches logged, average rating generated.

    The 5.0 was self-rated the year before, but only had 10 matches logged.

    Guess whose rating weighs more?

    Vice versa, someone is a career 5.0. Over 100 matches logged, average rating generated.

    The 3.0 only has 50 matches logged, with the average rating consistent over a long period of time.

    Guess who's going to move up fast, especially if a captain files a self-rating grievance?
  9. kevhen

    kevhen Hall of Fame

    Feb 20, 2004
    Auckland, New Zealand
    That rating system is so tragically flawed. If two guys rated 3.25 play each other and one double bagels the other indefinitely, they eventually will end up with a 3.43 rating and a 3.07 rating and will have to keep playing each other since they are still at the same level.

    A score a 6-1, 6-0 should not be just a .325 level difference but at least 0.50. A 6-0, 6-0 score should be like .75.

    If the USTA uses a formula like that then there will be a lot of overlap within each level with weak players that should be lower ranked and strong players who should be ranked higher. I see alot of this myself in the USTA so it does appear they use a formula similar to this. I (4.0 rated) beat some 4.5 guys in doubles two weeks ago with a 4.0 partner and then in the same night lost badly to some other 4.0 guys (one who was stronger than the 4.5s). People complain about sandbagging but it is this tragically flawed system that allows it to happen.
    Last edited: Mar 20, 2007
  10. kevhen

    kevhen Hall of Fame

    Feb 20, 2004
    Auckland, New Zealand
    I think a simpler system would be to bump players up who have 10 more wins than losses at their current level playing against players also ranked at that level. If they lose 10 or more matches than they have won they would drop a level. Then everyone would know how close they were to moving up or moving down. There is no reason to tank as you would have to tank half you matches to stay at a lower ranking.

    I know a guy who was 20-4 last year and he didn't get bumped up. He is stronger than a number of 4.5s that I know and yet he is allowed to play 4.0 doubles.
    Last edited: Mar 20, 2007

Share This Page