A woman is a benchmark 3.0 player. She goes 6-3 at 3.5 level singles and 11-0 at 3.0 level singles during this season. Not only that, but last year she completely destroyed most of her 3.0 competition as well. How the hell does this person not get disqualified from the 3.0 level play? Most of her 3.0 matches are like 6-1, 6-2, and her 3.5 matches are 6-2, 6-3. I would say she's good enough to get bumped to 4.0 at year's end, but I think she definitely should have been disqualified from 3.0 play, despite not being self-rated.
Explain the logic behind this one
- Thread starter raiden031
- Start date
CrocodileRock
Rookie
That's the problem raiden, there is no logic. A computer did this, instead of a human being with a brain stem.
randomname
Professional
same problem with a woman in our mixed doubles league, she has a mixed exclusive 3.5 rating, but only plays mixed and 4.0, shes gone 6-2 in 4.0, beating all computer rated 4.0s and one benchmark 3.5, with one of the losses in a tight 3 setter to a benchmark 4.0 and most of her wins being blowouts. and to add insult to injury, one of our 3.5 players self rates, then gets disqualified after winning 2 very close 4.0 matches and one 3.5 match.
Ace
Semi-Pro
Try this:
Take a look at all the leagues she played on.
For each league, did she go to districts? If so, how did she do there?
If she did not go to districts, how did she do against the person who did go to districts?
And, how did that person do at districts?
Also, how did the people she played against do against the rest of the people in their league?
I'm not saying this is the case here, but sometimes it looks like people are really great, because they are winning 6-2, 6-1 every match, but then you look at the people they played against, and those people are losing 2-6, 1-6 every match against everyone.
Also, how you do at districts, or against the people that go to districts is a major part of your rating.
Take a look at all the leagues she played on.
For each league, did she go to districts? If so, how did she do there?
If she did not go to districts, how did she do against the person who did go to districts?
And, how did that person do at districts?
Also, how did the people she played against do against the rest of the people in their league?
I'm not saying this is the case here, but sometimes it looks like people are really great, because they are winning 6-2, 6-1 every match, but then you look at the people they played against, and those people are losing 2-6, 1-6 every match against everyone.
Also, how you do at districts, or against the people that go to districts is a major part of your rating.
Without knowing who she is beating 6-1, 6-2, Im not sure this is really that ground breaking. I agree with Ace on this one.
You could beat everyone in 3.5 6-2, 6-3, and that does NOT definately mean you should be 4.0. If the people that you beat happened to clobber everyone else then it should be, but otherwise that's not really how it works.
Islandtennis
Rookie
In addition to what Javier said, we do not know what dynamic rating she started the season with. Just because she is a benchmark does not mean that her rating may be starting at the bottom of the 3.0 scale. To be DQ'd she has to have three strikes, which would mean that her dynamic rating would have to be in the high 3.5 or even 4.0 range. She probably will get bumped up, but its perfectly logical as to how she has not been DQ'd.
As I noted in my original post, she dominated the 3.0 level last year, winning probably over 90% of her matches pretty easily, which means she had to have been near the top of the level.
Also she played in two leagues, and my area is considered to be pretty strong in the 3.0 level. Although I play men's so it could be the women in that level are pretty weak. I do know that 37% of my league's 3.0 men's players got bumped up to 3.5 at the year end last year.
She played alot of matches, so I doubt that all of her opponents were weak. Most teams put their stronger players in singles anyways.
I don't think she was guaranteed to be moved up to 4.0, but I think its rediculous that someone who is good enough to possibly get moved up to 4.0 isn't getting disqualified from 3.0. And the main reason I even care enough to post this whole thing is because I just don't understand how this situation can occur logically. It seems to be a huge anomaly in the NTRP algorithm.
One example I can mention. One 3.5 woman she beat went undefeated in 2006 singles play at 3.5 and won about 80% of her 3.5 singles matches in 2007. I'd say both would be in the upper echelon of 3.5.
You keep saying things like "won 90% of her matches at X level", and "won 80% of her matches at this level".
That's meaningless, the system doesnt account for wins or losses at all, it goes by the game diffrence in relation to what your opponents rating is.
It doesnt seem logical because you keep throwing out these meaningless facts that by themselves have no bearing on what happens to someone's rating.
Even if this person you mentioned here was in the "upper echelon" of 3.5, you dont mention what the score was when the player in question beat her. You just mention that she beat her as if that has some special significance.
And even if that would produce a higher rating, it gets averaged in with the last 3 other results so it wouldnt necessarily produce a strike.
Raiden.Kaminari
Semi-Pro
Exactly. The other problem is we have no clue as to how strong any of the opponents were.
So, say for example a 3.5 player beat all 3.0 players, but lost to all 3.5 players. What does that make the player? Definitely a low 3.5 or a high 3.0 depending on the gravity of their rating.
As JavierLW mentioned, we need to know the scores as well.
For a player beating someone 6-0, 6-0, the old algorithm had the differential at .360. However, the new algorithm, which tries to take into consideration the difference between a young 3.5 and a senior 3.5, may assign more of a differential if age was also factored in.
Anyway, not enough data to post anything real about.
That's interesting that they factor in age.I didn't know that.
On an unrelated note, does anybody know why I can't edit my own posts so that I could maybe add a space between my sentences?
Ace
Semi-Pro
I keep a spreadsheet of what I "think" the USTA algorithm is for dynamic rating. I usually try to guess my opponents rating based on their past performance, and then use my little algorithm-thingy to calculate a "guess" of a dynamic rating for myself.
Its interesting to see, because you can actually beat someone, but end up having your rating go down, with the algorithm I used (based on info from the USTA website), once you factor in what thier rating might be and what the score was.
My spreadsheet probably isn't 100%, but I'd bet its probably pretty close. Plus, like I said, you have to factor in how you do in districts or against the people that go to districts, because that ends up being 50% of your rating. My spreadsheet doesn't do that.
I didn't know the differential for a 6-0, 6-0 match was .360, I had been using .5. Do you know that for a fact?
Its interesting to see, because you can actually beat someone, but end up having your rating go down, with the algorithm I used (based on info from the USTA website), once you factor in what thier rating might be and what the score was.
My spreadsheet probably isn't 100%, but I'd bet its probably pretty close. Plus, like I said, you have to factor in how you do in districts or against the people that go to districts, because that ends up being 50% of your rating. My spreadsheet doesn't do that.
I didn't know the differential for a 6-0, 6-0 match was .360, I had been using .5. Do you know that for a fact?
I guess you aren't very strong in the field of computer science or neuroscience. A computers operation is entirely dependent upon logic. The human brain is not.
I did the same exact thing at one point with the spreadsheet. I had it down to some degree of error to at least figure out who was going to get bumped at year end, but the DQ thing is too hard to figure out unless it's obvious. (like players clobbering other players who eventually got DQ'ed themselves or players beating other players who were just rated down from the higher rating)
I also assumed it was .5 apart. Does anyone know where we can find info to see if it was really .36? That would make it even harder to get rated out which would explain a lot.
Hey genius why don't you explain to me how a 3.0 player can consistently defeat computer-rated 3.5 players and still be less than 3.5? If they are so weak than why are they 3.5?
Im not sure a 6-3 record is exactly "consistantly".
If you beat someone who is a lower 3.5 but it's very close you are going to get a number that isnt much higher then theirs AND it's going to get averaged in with several of your other results before you get anything that could give you a strike.
Plus there are 3.0 rated players in the 3.5 league in a lot of areas, you dont mention that. You just say that she went 6-3 at 3.5. That's useless.
Plus the DQ's at the lower level have a built in threshold. Getting DQ'ed and getting rated up at the end of the year are two diffrent things.
You want to assume that this person is getting rated up to 4.0 (and you are assuming that even though you like to use terms like "likely"), but I doubt it, a 6-3 record is hardly groundbreaking and it's obvious that you are only looking at wins and losses.
Even in this rebuttal, once again you just focus on wins and loses, only you want to use vague terms like "consistantly" when in fact she is only winning 2/3rds of her matches at 3.5. (again winning and losing is meaningless by itself anyway)
As stated in my first post, her average score at 3.5 is around 6-2, 6-3. The one opponent I mentioned as an example won like 25 matches at 3.5 and lost about 8, and the match was went into a tiebreaker, with the 3.0 player winning. Of her matches, the average score was like 6-3, 6-3, with the occasional tiebreaker. So all signs show that this 3.0 player is absolutely crushing her 3.0 opponents and playing strong as a 3.5 player. And I have never seen a league where someone can play 20+ matches and all of their opponents are weak, yet playing #1 singles.
tennis-n-sc
Professional
I don't believe this is accurate. There may be some factor involved if a player is playing in both the adult and senior divisions. But I do not think the chronological age of players is ever a factor. You are gonna have to provide some evidence on this.
Ace
Semi-Pro
Again....
how did she do at districts? and if she didn't go to districts, how did she do against the people that did go to districts?
That is 50% of your rating.
I am not saying this is the case, but I have seen people dominate their local league, because their local league was weak. Then they go to districts or even states and get beat pretty easily. Its supposed to be a National rating, not just how you do against the people in your area.
Topaz
Legend
I don't think 6-2, 6-3 is enough to set the alarms off. And, as Javier mentioned, it depends on who they play and what that opponent's dynamic rating is...for example, I'm a 3.0 who has played up at 3.5 this past year, and sometimes my opponent at 3.5 was...another 3.0 playing up.
You seem dead set to prove something here...what is the real reason this has you so upset? Did you play against this woman in mixed or something like that?
Yes, there are cases where it seems like people aren't in the right level, but if they have a computer rating and have just improved a lot since the ratings last came out, more likely the computer will bump her up when the next ratings come out instead of DQing her. It is 'easier' to generate a DQ when the player is self-rated.
Ace
Semi-Pro
Raiden-
Do I have the correct person?
Did she play a 6-3, 5-7, 1-0 match in 3.0 Flight Playoff against another 3.0 this year?
When she went to 3.5 Flight Playoffs, did she got blown away by a 3.5 player 6-2, 6-0? (Though that person could potentially get bumped up this year, which would help this 3.0 get bumped up as well)
Also, I am not sure, but somebody should be able to confirm this...the 1/2 portion of your rating that comes from the "national benchmark" comparison, I believe (I could be wrong, but I thought I read this once) is based on how you do at the highest level you played. So 1/2 her rating is based on how she compared to the 3.5 benchmarks, and the other 1/2 is her dynamic rating of how she did against everyone during the season.
I think her 3.5 performance in 2006 probably hurt her, and her 3.5 performance at playoffs this year probably didn't help her much either for next year, though she probably played a strong 3.5, so she will probably start next year as a low-mid 3.5.
Do I have the correct person?
Did she play a 6-3, 5-7, 1-0 match in 3.0 Flight Playoff against another 3.0 this year?
When she went to 3.5 Flight Playoffs, did she got blown away by a 3.5 player 6-2, 6-0? (Though that person could potentially get bumped up this year, which would help this 3.0 get bumped up as well)
Also, I am not sure, but somebody should be able to confirm this...the 1/2 portion of your rating that comes from the "national benchmark" comparison, I believe (I could be wrong, but I thought I read this once) is based on how you do at the highest level you played. So 1/2 her rating is based on how she compared to the 3.5 benchmarks, and the other 1/2 is her dynamic rating of how she did against everyone during the season.
I think her 3.5 performance in 2006 probably hurt her, and her 3.5 performance at playoffs this year probably didn't help her much either for next year, though she probably played a strong 3.5, so she will probably start next year as a low-mid 3.5.
Last edited:
Raiden.Kaminari
Semi-Pro
I don't know if you realized ratings were published a couple of times last year (within a day or so of being released).
Age was recently added because of all the problems with self-rated players. Senior players have a lot more data on them (because senior leagues are also National leagues, with DNTRP calculations), and when a younger player beats a senior benchmark player, they get disqualified pretty quickly (within 2 or 3 matches).
You should be able to find the evidence yourself, especially when players are 3 strike DQ'd. If you trace back and find a senior benchmark player, it should make sense. Then again, it isn't as obvious if you don't have access to some of the reports I see.
tennis-n-sc
Professional
In other words you have no reliable evidence that age is an official component of the dynamic system.
Islandtennis
Rookie
Raiden,
There is a huge difference between a dynamic rating of 3.5 and being DQ'd at 3.0. Her dynamic rating is in all probability now above 3.0. That doesn't mean she is instantly DQ'd. Her dynamic may in fact now even be in the 4.0 range, but that doesn't in and of itself warrant an immediate DQ.
tennis-n-sc
Professional
From the Southern Section office:
Q. With regard to the factors used in the dynamic rating system, is age ever considered or is age one of the factors, implicitly or explicitly?
A. No, not in the calculation process. Only in the appeals process where players over age 60 receive more benefit in the criteria to grant an appeal than a player under age 60.
This is a constant topic in NTRP debates but current regulations/methodology reflect current majority opinion at national level.
Director of Adult Tennis
USTA Southern
5685 Spalding Drive
Norcross, GA 30092
www.southerntennis.com
It simply wouldn't make sense to even attempt this. What is the difference in a 70 y/o 3.5 beating a 25 y/o 3.5 or vise versa. The age will eventually be reflected in the scores of the matches and thus in the dynamic process. In addition, the USTA is just as liable to discrimination actions as any other large entity. I think this is being handled correctly by the USTA, not to say that they don't have many other issues.
Similar threads
