PDA

View Full Version : Dynamically DQ'ed --- Why


JLyon
06-02-2009, 04:51 PM
The following scores were by player A:
L 6-1, 6-1
W 6-4, 7-5
W 7-6, 6-3
W 6-4, 6-3
W 6-3, 6-4
W 6-0, 6-3
W 6-3, 6-1
W 0-6, 7-6 retire

Note this player was rated 4.5 end of 2005

rabidturtle
06-02-2009, 04:58 PM
There isn't enough information to tell you why player A got dynamically DQ'ed.

If we had the ratings of players A's Opponents and any other additional information (past history, grievances, etc...)

cak
06-02-2009, 05:23 PM
What level was he playing in 2009?

What were the scores of his matches since 2005?

Did he play against baseline (B rated) players? If so, what were those scores?

Did any of the guys he played against also get DQ'd?

Are you sure it was dynamic, and not a grievance?

Nanshiki
06-02-2009, 05:58 PM
Probably had to do with the 7-1 record. Federer doesn't have a 7-1 record. (it's 4-1 BTW). This would be different if it was 3.0, 3.5 tennis where you have a lot of people who are terrible at matchplay or play too high. But 4.5 is a different thing all together.

JLyon
06-02-2009, 06:15 PM
it was a Dynamic DQ, because he now has a "D" for the rating. He beat a mix of B,C, and recent bump-ups.
The 1,1 loss was against 2 Benchmarks who go to Senior Sectionals every year.
He did not play any of the Super Ringers present.
He also did not play league in 2006-08. He self-rated 4.5 for this year, although he could have by rule self-rated at 4.0 since you can self-rate down .5 rating.

Nellie
06-02-2009, 06:53 PM
I have said it before and poeple disagree, but my belief is that the USTA believes that new players should lose. If you self rate, even if you played before, and win most of your matches, you can get DQ'ed by the computer. By the way, I bet he was already DQ'ed before the last match, because the score and the retirements would not suggest to me that that result would be a thrid strike, whereas a 3 and 1 win could easily get you a strike.

Jim A
06-02-2009, 07:39 PM
hmm..so I thumped someone who has turned out to be a B player last week, 2 and 0

of course it was the first singles match he'd ever played

I can only assume as a self-rated that this is a strike against me, regardless of the fact he never plays singles

kylebarendrick
06-02-2009, 09:09 PM
B doesn't mean they are at the top of the level. It means they played in a playoff match last year. It may have even been a playoff match at a lower level before being bumped.

JavierLW
06-02-2009, 09:22 PM
hmm..so I thumped someone who has turned out to be a B player last week, 2 and 0

of course it was the first singles match he'd ever played

I can only assume as a self-rated that this is a strike against me, regardless of the fact he never plays singles

Having a B is meaningless, doesnt tell you what his real rating is, just tells you he went to the playoffs.

With the OP's example, it's totally meaningless because we dont know who he played and even if we did, we dont know their ratings anyway.

The Threshold is lower for 4.5 then 3.0 and 3.5, this much I know from the descriptions of how this system worked when it came out.

(the theory was they understood that at 3.0 and 3.5 players would improve as the season went on so they didnt want a mass DQ'ing of everyone, but as you go higher in level, the odds of rapid improvement go down so if someone is throttling everyone it's because they dont belong)

I wonder if this guy played any other levels as well?

HookEmJeff
06-02-2009, 10:01 PM
Well, a self-rated teammate of mine got Dynamically DQ'd in three matches this year.

W: 0-6, 6-4, 1-0 (10-point breaker) over probably one of the top-rated players in our area at his level
W: 6-3, 6-1 victory over a pretty solid singles guy who also doesn't have a lot of data on him
W: 6-4, 6-2 victory over a weak 4.0 and a low 4.5 in doubles

Basically, he probably tipped the scale on the first match. The guy my teammate beat in his first match...that guy made Nationals and has won a few age group titles around here (not that those count towards NTRP this year from what I am hearing).

So, three strikes seems kind of deceiving. It's not necessarily the wins over those players per se. It can be one BIG strike that does all the damage, if that strike is a top dog in a level. You basically shoot past his Dynamic rating...and for our guy, that win was his VERY FIRST MATCH of the season. I'm guessing the other two matches after that first one his calculation probably flatined or barely blipped up and might even have gone down a hair. But it was still above the threshold from that one big win.

Makes me wonder if the calculation for self-rated players is somehow different (or tweaked) this year ...different from the Benchmark or Computer-rated players. I think it also goes to show ORDER of matches in those three strikes might play a part more than anything. Based on my teammate's limited data on him, I wonder if he would've been promoted that fast if he had played those matches in reverse order? Just a question I don't know the answer to but would love to find out.

Jeff

amarone
06-03-2009, 03:04 AM
Probably had to do with the 7-1 record. No it doesn't. Win-loss record is irrelevant. The only things that count are the strength of your opponent (and partner, in doubles) and the margin of victory/defeat.

amarone
06-03-2009, 03:10 AM
The Threshold is lower for 4.5 then 3.0 and 3.5, this much I know from the descriptions of how this system worked when it came out. Correct. At 3.0 and 3.5 the DQ level is 0.3 into the next level, e.g. 3.0 DQ level is 3.3. At 4.0, it falls to being 0.2 into the next level.

amarone
06-03-2009, 03:13 AM
hmm..so I thumped someone who has turned out to be a B player last week, 2 and 0

of course it was the first singles match he'd ever played

I can only assume as a self-rated that this is a strike against me, regardless of the fact he never plays singles You cannot tell. As already stated by others, the fact that he is Benchmark is completely irrelevant.

However, one of the dangers of playing singles is that you can play a good doubles player whose rating does not reflect their singles ability. You beat them handily and bang, you have a strike, because your rating is affected by their inaccurate (for singles) strong rating.

amarone
06-03-2009, 03:17 AM
I have said it before and poeple disagree, but my belief is that the USTA believes that new players should lose. The USTA does believe that you should self-rate not for the level you are right at the time you self-rate, but where you expect to be in a few weeks/months' time. Example - an athletic person who used to play 4.0 is just starting to play again but at the moment is incredibly rusty. This person should not rate as a 3.0 just because at the moment he can hardly get the ball over the net - he should rate at least 3.5 because in a few weeks' time he will be back to that level.

raiden031
06-03-2009, 03:23 AM
B doesn't mean they are at the top of the level. It means they played in a playoff match last year. It may have even been a playoff match at a lower level before being bumped.

Exactly. I am a 4.0B player because I played at championships, which happened to be at the 3.0 level. I am one of the weakest 4.0 players in my league right now, at least according to my lame singles record this year.

amarone
06-03-2009, 03:34 AM
As the benchmark misunderstanding appears in thread after thread, let's explain what being a benchmark player is all about.

Benchmark players are used to normalize the ratings across the country. The issue is how to try ensure that a 3.5 in NorCal is similar to a 3.5 in Southern. And withn Southern, a 3.5 in Arkansas is similar to a 3.5 in Georgia. The answer is to calculate ratings using players from these different sections/districts who have played each other - and that means players who have reached Nationals, Sectionals etc.

Hence a player who has reached playoffs and therefore gets to play against players from other districts etc. is used as a "benchmark". When the national ratings are run each November, the players that reached nationals are used as a first level of calibration, then players that reached Sectionals, and so on.

This ripples down and affects the ratings of everyone, albeit possibly in a very minor way. This calibration can affect a whole group of players. A couple of years ago many 3.5 players in Georgia got bumped up. I suspect it was because a Georgia 3.5 team came second in Nationals that year, and hence the calibration process determined that Georgia was strong and everyone got pulled up a little in the national ratings.

JLyon
06-03-2009, 04:48 AM
I guess it is confusing, just feel bad for the player and Captain since they were playing within the rules, technically. While another self-rate has basically lied on self-rate form, score managed and will maintain a 4.5 rating going into Sectionals, outside of an obvious Grievance, that should be filed.
A Former Top 150 ATP player and D1 All-American (< 50 y.o.)has no justification playing 4.5 Tennis as a self-rate

KFwinds
06-03-2009, 07:58 AM
No it doesn't. Win-loss record is irrelevant. The only things that count are the strength of your opponent (and partner, in doubles) and the margin of victory/defeat.

I'm glad you answered that; I still can't believe the number of folks out there who don't understand that the USTA ratings don't factor wins and losses. A player could go 187-0 and not get bumped if all the people they played had a much lower dynamic rating and/or the scores were competitive (our district coordinator defines "competitive" as being 6-3, 6-3 or closer).

Jack the Hack
06-03-2009, 10:36 AM
I guess it is confusing, just feel bad for the player and Captain since they were playing within the rules, technically. While another self-rate has basically lied on self-rate form, score managed and will maintain a 4.5 rating going into Sectionals, outside of an obvious Grievance, that should be filed.
A Former Top 150 ATP player and D1 All-American (< 50 y.o.)has no justification playing 4.5 Tennis as a self-rate

If you have proof that this player lied on his self-rating form, you can file a grievance even if you haven't played that team yet. I did this last year with a guy in our league that self-rated at 4.5 and was a former top 15 ranked college player that was in his mid-20s. He was cruising through the season going undefeated, and I expected the system to dynamically disqualify him. When that didn't happen, I e-mailed the captains of several of the teams that lost to him, but they didn't want to file a grievance in fear of retribution. Anyway, a week before my team was set to play them, I went ahead and filed the grievance myself. Turns out that he never disclosed his college playing experience on the online self-rate form, and the USTA automatically disqualified him. Funny thing is that he never denied being a 5.0 player, but argued instead that the self rate form was flawed and the punishment for missing the college question was too harsh. The USTA grievance committee in my section didn't have too much sympathy for his "ignorance" defense. All of his match wins were reversed, and since a majority of his team's wins were by 3-2 scores (which were reversed by his score reversals), they went from having a 6-2 record and playoff spot to 2-6 and near last place. I also understand that a captain can be penalized for knowingly playing a self-rated player that has lied on the questionaire, but that didn't happen in this particular case. The fact that it knocked their team out of the playoffs was punishment enough.

I am not relating this story to brag. I ticked off a lot of people on that team, and several of the other captains in my area thought I was being too anal at the time. However, the fact is that if you know a player is abusing the system and do nothing about it, you have no right to complain when they get away with it.

JLyon
06-03-2009, 10:45 AM
If you have proof that this player lied on his self-rating form, you can file a grievance even if you haven't played that team yet. I did this last year with a guy in our league that self-rated at 4.5 and was a former top 15 ranked college player that was in his mid-20s. He was cruising through the season going undefeated, and I expected the system to dynamically disqualify him. When that didn't happen, I e-mailed the captains of several of the teams that lost to him, but they didn't want to file a grievance in fear of retribution. Anyway, a week before my team was set to play them, I went ahead and filed the grievance myself. Turns out that he never disclosed his college playing experience on the online self-rate form, and the USTA automatically disqualified him. Funny thing is that he never denied being a 5.0 player, but argued instead that the self rate form was flawed and the punishment for missing the college question was too harsh. The USTA grievance committee in my section didn't have too much sympathy for his "ignorance" defense. All of his match wins were reversed, and since a majority of his team's wins were by 3-2 scores (which were reversed by his score reversals), they went from having a 6-2 record and playoff spot to 2-6 and near last place. I also understand that a captain can be penalized for knowingly playing a self-rated player that has lied on the questionaire, but that didn't happen in this particular case. The fact that it knocked their team out of the playoffs was punishment enough.

I am not relating this story to brag. I ticked off a lot of people on that team, and several of the other captains in my area thought I was being too anal at the time. However, the fact is that if you know a player is abusing the system and do nothing about it, you have no right to complain when they get away with it.

I have filed my share of grievances and won several including a former Top 10 D2 player trying 4.5 and another D2 who was solid 5.0 trying to self-rate. I am no longer captaining, but I am more upset that the state association knows this player and has done nothing and in my opinion in this case. a captain should not have to police this type self-rate play.

Jack the Hack
06-03-2009, 10:47 AM
The following scores were by player A:
L 6-1, 6-1
W 6-4, 7-5
W 7-6, 6-3
W 6-4, 6-3
W 6-3, 6-4
W 6-0, 6-3
W 6-3, 6-1
W 0-6, 7-6 retire

Note this player was rated 4.5 end of 2005

As to your original question... so this was a computer rated 4.5 at the end of '05 that just came back to the league this year and self-rated at 4.5 again?

If that's the case, I can totally understand your confusion regarding why he got bumped to 5.0. The only thing that I can see might be that his first 5 matches must have been against top level 4.5ers, and that 6-0, 6-3 win in his 6th match must have been his first strike. Once he rose above the dynamic threshold in that match, it doesn't look like his subsequent match wins did enough to move him back down out of the range.

This story is a bit scary to me because the top player on my 4.5 team is a "mixed" (M) rated 4.5er. He actually was offended by getting rated 4.5 and appealed to be moved back up to 5.0, but was denied. Since mixed ratings are subject to dynamic disqualification, he could be on the bubble. We haven't been hiding him, and he is undefeated this year by similar scores as you've listed above. :shock:

Nanshiki
06-03-2009, 08:35 PM
I'm glad you answered that; I still can't believe the number of folks out there who don't understand that the USTA ratings don't factor wins and losses. A player could go 187-0 and not get bumped if all the people they played had a much lower dynamic rating and/or the scores were competitive (our district coordinator defines "competitive" as being 6-3, 6-3 or closer).

Well, if that's true it doesn't make much sense. If you win 1000-0 against people who average somewhere near 3.5 dynamically (even if it's by 6-4 or higher every time), how could you possibly be a 3.5?

raiden031
06-04-2009, 02:48 AM
Well, if that's true it doesn't make much sense. If you win 1000-0 against people who average somewhere near 3.5 dynamically (even if it's by 6-4 or higher every time), how could you possibly be a 3.5?

As long as those scores don't push your dynamic NTRP too high....

There is a weird phenomen where some players are at your level, yet you can always find a way to win. For example, there is this guy I play indoors frequently during the winter months where we would have close scores and often go into 3rd sets, but for some reason I would always win. So if I went like 12-0 in the winter against him, does that mean I'm an entire NTRP better than him? No, I am marginally better but for some reason I always rise just enough to beat him. I'm playing 4.0 right now and he's playing both 3.5 and 4.0, and he has a better record at 4.0 than I do!

amarone
06-04-2009, 03:44 AM
Well, if that's true it doesn't make much sense. If you win 1000-0 against people who average somewhere near 3.5 dynamically (even if it's by 6-4 or higher every time), how could you possibly be a 3.5?
Let us say you are a high 3.5 and you play 1,000 average 3.5 players. The computer algorithm expects you to win approximately 6-4, 6-4. If you win 4 & 4 every time, your rating will not change because you have done exactly what the algorithm predicts you will do if you are at the correct level. And that is how it should be.

Nanshiki
06-04-2009, 08:29 AM
The law of averages says that there's no 3.5 in the world who can beat average 3.5s 1000 times in a row without losing, much less with the same score every time...

slick
06-04-2009, 08:57 AM
Let us say you are a high 3.5 and you play 1,000 average 3.5 players. The computer algorithm expects you to win approximately 6-4, 6-4. If you win 4 & 4 every time, your rating will not change because you have done exactly what the algorithm predicts you will do if you are at the correct level. And that is how it should be.

Then they need to change the computer algorithm because the real world doesn't work that way.

raiden031
06-04-2009, 08:58 AM
The law of averages says that there's no 3.5 in the world who can beat average 3.5s 1000 times in a row without losing, much less with the same score every time...

Hence that is why there are no players with a 1000-0 record. So there is no issue with the system ignoring wins and losses.

raiden031
06-04-2009, 09:00 AM
Then they need to change the computer algorithm because the real world doesn't work that way.

That is why they take averages. If you lose 6-1, 6-1 to every player, then you get moved down. If you win some 6-1, lose some 6-1, and play a couple close matches it evens out during averaging.

amarone
06-04-2009, 03:09 PM
Then they need to change the computer algorithm because the real world doesn't work that way. While the idea that 1,000 consecutive matches would go the same way is obviously unrealistic, if this hypothetical scenario did occur, then it is totally correct that the player stays at exactly the same rating.

Let us try a golf analogy, for those that follow golf. The NTRP rating is equivalent to a golfer's handicap. If a golfer has a handicap of 14 and (hypothetically) every round he played he played to exactly 14, it would clearly be nonsense to say that his handicap should be anything other than 14.

If our 14-handicap player played 1,000 times against a 16-handicap player, who played to his handicap every time, our 14-handicap player would beat him by two strokes every time. Should his handicap of 14 be reduced? Clearly the answer is no.

And this is what we see in tennis. A player with a handicap/rating of 3.45 can play a player with a handicap/rating of 3.3 and beat him by the same narrow margin every time and of course his handicap/rating should not change.

Sticking purely to tennis, our high 3.5 player beats the mid-3.5 players every time. This same player may also lose narrowly to 4.0 players all the time. Should the high 3.5 player be moved to low 4.0 (where he always loses) just because he can consistently beat players who are weaker than him? Again, the answer is clearly no.

Jim A
06-05-2009, 07:13 AM
some random thoughts on the run:

1. Potentially the NTRP ratings shouldn't be a year long process, or basically 2 years. If you play a year at 3.5, and come up a bit short, you have an entire year to dominate the level before moving to 4.0. Let's face it there should be more 4.0's and fewer 3.5's in most areas, this will help give us more than 3.0/3.5/4.0 as strong levels in regions. Some sort of point system through league/tournament/mixed/etc to bump people at least every 6 months

2. League play is the thorn in the side for all of this. Calendars don't match for all the districts, so some people are nearly done and waiting for Nationals (in October!!) while others are early in league play. If you had a system that allowed B/C players to be bumped after league play and a playoff system that was closer, there could be more movement of the proper players. However you could bump a 3.5 to 4.0 and then have him come back 4 months later to 3.5 Nationals, and that would be a mess.

3. If you have an NTRP rating, you play it and nothing up (unless OPEN). I like my district rule where you can't play up

Nanshiki
06-05-2009, 10:12 AM
While the idea that 1,000 consecutive matches would go the same way is obviously unrealistic, if this hypothetical scenario did occur, then it is totally correct that the player stays at exactly the same rating.

Let us try a golf analogy, for those that follow golf. The NTRP rating is equivalent to a golfer's handicap. If a golfer has a handicap of 14 and (hypothetically) every round he played he played to exactly 14, it would clearly be nonsense to say that his handicap should be anything other than 14.

If our 14-handicap player played 1,000 times against a 16-handicap player, who played to his handicap every time, our 14-handicap player would beat him by two strokes every time. Should his handicap of 14 be reduced? Clearly the answer is no.

And this is what we see in tennis. A player with a handicap/rating of 3.45 can play a player with a handicap/rating of 3.3 and beat him by the same narrow margin every time and of course his handicap/rating should not change.

Sticking purely to tennis, our high 3.5 player beats the mid-3.5 players every time. This same player may also lose narrowly to 4.0 players all the time. Should the high 3.5 player be moved to low 4.0 (where he always loses) just because he can consistently beat players who are weaker than him? Again, the answer is clearly no.

Tennis ≠ Golf

There are no handicaps in tennis. There are no scores or standards based purely on an individual's playing; tennis is a game played between two competitors, not a comparison of individual scores. A 4.0 will not always beat a 3.5. A 3.5 will not always beat a 3.0. And a true 3.5 should not have a 14-1 record against true 3.5s. However, 7-1 is more reasonable.

FWIW, Rod Laver had a win ratio of less than 4. And Jimbo Connors, who has the most career wins of anyone, is only 4.5 or so.