2018 YE Ratings Hype Thread

leech

Rookie
So any trilevel played at this point will not count?
Yup, that is my understanding, that any matches after 10/28 won't factor into 2018 year end ratings. I have a buddy that didn't do as well as he'd hoped at Nationals, so his TR estimated rating dropped down to the 3.35 range. He crushed it in his last two TriLevel men's matches, so his TR rating might be in the 3.62 range at the next update, but to no avail. He was crushed to learn that the last two matches won't count, ha. (He is one of the rare men who WANT to get bumped up)
 

time_fly

Hall of Fame
Yup, that is my understanding, that any matches after 10/28 won't factor into 2018 year end ratings. I have a buddy that didn't do as well as he'd hoped at Nationals, so his TR estimated rating dropped down to the 3.35 range. He crushed it in his last two TriLevel men's matches, so his TR rating might be in the 3.62 range at the next update, but to no avail. He was crushed to learn that the last two matches won't count, ha. (He is one of the rare men who WANT to get bumped up)
If someone's level is jumping around by 0.3, which more than half an NTRP level, based on the last couple matches then the algorithm is way too unstable. Hypothetically if they updated the leagues every month with an algorithm like that, half your roster could change each time which makes no sense.
 

ShaunS

Semi-Pro
If someone's level is jumping around by 0.3, which more than half an NTRP level, based on the last couple matches then the algorithm is way too unstable.
Tennis Record is not the same as USTA NTRP. I would typically agree with you, but potentially he's only played a handful of matches? On the other hand I'm skeptical that losing at nationals would have a dramatic impact on your rating. Everything suggests you should be playing a pretty tight band of rated players.

Remind him that he can always play up even if he doesn't get bumped.
 

leech

Rookie
If someone's level is jumping around by 0.3, which more than half an NTRP level, based on the last couple matches then the algorithm is way too unstable. Hypothetically if they updated the leagues every month with an algorithm like that, half your roster could change each time which makes no sense.
Agree, the TR algorithm is likely way off. But he went from 3.38 to 3.50 after a 1&0 win, and then had a second 1&0 win that hasn't been calculated on TR yet, but I wouldn't be surprised if he's at 3.62 next update.
 

leech

Rookie
Tennis Record is not the same as USTA NTRP. I would typically agree with you, but potentially he's only played a handful of matches? On the other hand I'm skeptical that losing at nationals would have a dramatic impact on your rating. Everything suggests you should be playing a pretty tight band of rated players.

Remind him that he can always play up even if he doesn't get bumped.
Whether it's accurate or not is up for debate, but there were some surprisingly (suspiciously?) low rated players we faced at Nationals. So losing to them (even though they were better than us) really lowered our ratings on TR. Like he lost to a 2.96 guy...and the match wasn't that close.

He does play up already at 4.0 with decent results, but I'm guessing it's more of a status thing for him to get to 4.0 -- and he wants to play on the 4.5 team he captains.
 

time_fly

Hall of Fame
Whether it's accurate or not is up for debate, but there were some surprisingly (suspiciously?) low rated players we faced at Nationals. So losing to them (even though they were better than us) really lowered our ratings on TR. Like he lost to a 2.96 guy...and the match wasn't that close.

He does play up already at 4.0 with decent results, but I'm guessing it's more of a status thing for him to get to 4.0 -- and he wants to play on the 4.5 team he captains.
I filed my first official grievance last year against a self-rated 3.5 that was about 29 years old and had a winning record in Division II college tennis according to an old team webpage I found. He played only mixed doubles, and I think TR had him around 3.20 even though he would be a 4.5 on anyone's "eyeball test." Since he played 7.0 matches and the other teams still won some games, the TR algorithm concluded he was within range.
 
Last edited:

brettatk

Semi-Pro
I played a regular season 40+ match against a team from the eventual 4.0 National Champions. Didn't get blown out but lost the match. TR showed my rating went down since both of them had lower ratings than mine. Not surprising since they've been tanking for years at 18+. Another reason why tanking sucks because it doesn't just effect your own rating.
 

J_R_B

Hall of Fame
What is the appealable range?
I don't think they publish the actual range (and it varies by the number of matches you've played and possibly by rating level), but it definitely is not 0.08 in any circumstance.
 

J011yroger

Talk Tennis Guru
I played a regular season 40+ match against a team from the eventual 4.0 National Champions. Didn't get blown out but lost the match. TR showed my rating went down since both of them had lower ratings than mine. Not surprising since they've been tanking for years at 18+. Another reason why tanking sucks because it doesn't just effect your own rating.
When teams don't play their highest rated players in the playoffs and win anyway suspicious Jolly gets suspicious.

J
 
10/28 was indeed the cutoff for any matches counting toward 2018 year end ratings.

Tennisleaguestats is as reliable and credible as a Florida vote count. They are the fake news of NTRP projections.

New ratings will be released on or after December 1.
 
10/28 was indeed the cutoff for any matches counting toward 2018 year end ratings.

Tennisleaguestats is as reliable and credible as a Florida vote count. They are the fake news of NTRP projections.

New ratings will be released on or after December 1.
Does TLS not keep ongoing stats for the current season? I can't find myself on there at all. Seems like it only has last season and prior stuff.
 
Best I can tell, they only look at a single year and then only post a player with a rating if he or she has played three "Adult" league matches. They don't count in Tri-Level (probably because it depends on the local area/section if the results even count.)

I'm not saying they're never right. Usually if they predict someone going up by .20 or more into the next level, they nail it, but that is common sense. Otherwise, they should carry a disclaimer similar to horoscopes. "Ratings are for entertainment purposes only."
 

OnTheLine

Hall of Fame
Can I be on your team?

Well, it depends.

J
You, probably willing to take you ... can you wear a skirt, change your name, move across the country and re-rate yourself as a 3.5?

But I am changing my requirements this year and intend to completley re-shuffle my 40+ team:
a. will you attend at least 50% of our practices ... if not, are you hitting at least 2 other times per week outside of a match? If answer is no to both ... not interested
b. Are you capable of responding to texts/emails/phone calls ... promptly. If not go find someone else to torture.
c. Do you arrive to all matches at least 15 minutes early and ready to play? if not, again, go find someone else to torture
d. Do you play with an attempt to be competitive out there or are you just there for the giggles. I don't care if you lose as long as you try, but if only there for the giggles .... there are other teams for you.
 
You, probably willing to take you ... can you wear a skirt, change your name, move across the country and re-rate yourself as a 3.5?

But I am changing my requirements this year and intend to completley re-shuffle my 40+ team:
a. will you attend at least 50% of our practices ... if not, are you hitting at least 2 other times per week outside of a match? If answer is no to both ... not interested
b. Are you capable of responding to texts/emails/phone calls ... promptly. If not go find someone else to torture.
c. Do you arrive to all matches at least 15 minutes early and ready to play? if not, again, go find someone else to torture
d. Do you play with an attempt to be competitive out there or are you just there for the giggles. I don't care if you lose as long as you try, but if only there for the giggles .... there are other teams for you.
Id like to submit my name. Can you please provide extra credit opportunities if such player does not meet but 1 - 2 of your requirements. Or possibly do you round up? Is meeting 3 out of 4 requirements "passing?"
2 is close to 3. Just asking for a friend.
 

Nostradamus

Bionic Poster
Id like to submit my name. Can you please provide extra credit opportunities if such player does not meet but 1 - 2 of your requirements. Or possibly do you round up? Is meeting 3 out of 4 requirements "passing?"
2 is close to 3. Just asking for a friend.
How does USTA calculate exactly who gets bumped up or down ? They claim their system is totally different than those other system that is published, UTR or whatever. but they won't say what system they really use and it is their own only
 

OnTheLine

Hall of Fame
Id like to submit my name. Can you please provide extra credit opportunities if such player does not meet but 1 - 2 of your requirements. Or possibly do you round up? Is meeting 3 out of 4 requirements "passing?"
2 is close to 3. Just asking for a friend.
Extra credit for consistently hitting TS groundie behind the service line.

This is ladies' 3.5 ... more extra credit if your 2nd serve is anything other than a dink.

If you can't get extra points for either of these, I will offer additional extra credit if you can rally with me for over 30 shots.

The only one that is an entire deal breaker is #3. If you are not on time for matches you're out!
 
You, probably willing to take you ... can you wear a skirt, change your name, move across the country and re-rate yourself as a 3.5?

But I am changing my requirements this year and intend to completley re-shuffle my 40+ team:
a. will you attend at least 50% of our practices ... if not, are you hitting at least 2 other times per week outside of a match? If answer is no to both ... not interested
b. Are you capable of responding to texts/emails/phone calls ... promptly. If not go find someone else to torture.
c. Do you arrive to all matches at least 15 minutes early and ready to play? if not, again, go find someone else to torture
d. Do you play with an attempt to be competitive out there or are you just there for the giggles. I don't care if you lose as long as you try, but if only there for the giggles .... there are other teams for you.
Solid list. 2/3 of the local league team I captain fail at A-C. C. really irritates the **** out of me. A. may be a little much to expect out of my league, but it's annoying when they lose, come off the court like "I could've beat that guy but I'm just rusty." Oh you're rusty? No ****. You play twice a month.
 

OnTheLine

Hall of Fame
Solid list. 2/3 of the local league team I captain fail at A-C. C. really irritates the **** out of me. A. may be a little much to expect out of my league, but it's annoying when they lose, come off the court like "I could've beat that guy but I'm just rusty." Oh you're rusty? No ****. You play twice a month.
Yes, C. kills me. B just really irritates me.

A and kinda D .... I really don't care if you are the worst player in the league ... as long as you put forth the effort to improve / play well ... If you put forth that effort I will keep you around no problem
 

OnTheLine

Hall of Fame
They used to pull data on Tuesdays and then push the new numbers on Fridays ... now they are a little less predictable ... last week pulled on Wednesday night and included match results that I had entered into tennislink that same morning
 
And then a few more days for it to actually calculate that match into your rating... that's even stranger to me. I'd think it be like a spreadsheet where when you enter in a number, it automatically/instantly recalculates the totals.
I'm just picturing some guy sitting in a basement somewhere with a piece of paper and a calculator doing each person's info by hand, haha.

Don't get me wrong, it's a great website. I wish Tennislink were as navigable. I'm just severely impatient.
 

kevrol

Hall of Fame
You, probably willing to take you ... can you wear a skirt, change your name, move across the country and re-rate yourself as a 3.5?

But I am changing my requirements this year and intend to completley re-shuffle my 40+ team:
a. will you attend at least 50% of our practices ... if not, are you hitting at least 2 other times per week outside of a match? If answer is no to both ... not interested
b. Are you capable of responding to texts/emails/phone calls ... promptly. If not go find someone else to torture.
c. Do you arrive to all matches at least 15 minutes early and ready to play? if not, again, go find someone else to torture
d. Do you play with an attempt to be competitive out there or are you just there for the giggles. I don't care if you lose as long as you try, but if only there for the giggles .... there are other teams for you.
Unless of course the person clearly slipped through the cracks and should have been bumped to 4.0. Then I'd be willing to overlook a couple of these.
 

ShaunS

Semi-Pro
And then a few more days for it to actually calculate that match into your rating... that's even stranger to me. I'd think it be like a spreadsheet where when you enter in a number, it automatically/instantly recalculates the totals.
I'm just picturing some guy sitting in a basement somewhere with a piece of paper and a calculator doing each person's info by hand, haha.
Admittedly, I don't have intimate details on how TR handles their processing, but I'll assume it's like any database of real size.

Excel can appear to do things instantly because the calculations are very simple and humans enter data at a glacial pace. When you're dealing with lots of data it's much more efficient to pull everything in first then run any analysis against the data. I wouldn't be surprised if they do some validation in between too. From a conceptual standpoint the idea of pulling a single match, running the current and recent matches to evaluate that, and then proceeding to the next single match is... not optimal.
 

schmke

Hall of Fame
Admittedly, I don't have intimate details on how TR handles their processing, but I'll assume it's like any database of real size.

Excel can appear to do things instantly because the calculations are very simple and humans enter data at a glacial pace. When you're dealing with lots of data it's much more efficient to pull everything in first then run any analysis against the data. I wouldn't be surprised if they do some validation in between too. From a conceptual standpoint the idea of pulling a single match, running the current and recent matches to evaluate that, and then proceeding to the next single match is... not optimal.
I don't know what TR does, but I regularly add matches and recalculate all my ratings to do reports for people in an hour or two. It isn't (or shouldn't be) that complicated.
 

ShaunS

Semi-Pro
It isn't (or shouldn't be) that complicated.
True, I should clarify that I wasn't implying the days in between were necessarily required, only that it'd be abnormal to pull data and perform calculations at the same time.

If I was taking a total guess I would say they're likely using the same hardware to host the site as perform the processing, so they spread it out to minimize any impact.
 

Zman

New User
I'm sure they fix their "misses" after ratings are published so that is all that happened here. But they were off by a full 0.14 (if not more) so that is a pretty big miss.

Note I do the same corrections when I do miss with my estimates, if I have someone at 3.50 or lower and they are a 4.0, I move their rating up into 4.0 range, but very rarely miss by that much.
0.14 counts as a big miss? Check out Tinsley Dozier, a self-rated 4.0 whose tennisrecord dynamic is 4.6, even though he played eight matches and didn't get dq'ed from 4.0. And actually tennisrecord recently adjusted his rating downward, a couple weeks ago it said he played 6.0-caliber tennis in his second match, which now has been lowered to a more believable(?) 5.5. If tennisrecord were even remotely close to correct, he would have been dq'ed long ago.
 

J_R_B

Hall of Fame
0.14 counts as a big miss? Check out Tinsley Dozier, a self-rated 4.0 whose tennisrecord dynamic is 4.6, even though he played eight matches and didn't get dq'ed from 4.0. And actually tennisrecord recently adjusted his rating downward, a couple weeks ago it said he played 6.0-caliber tennis in his second match, which now has been lowered to a more believable(?) 5.5. If tennisrecord were even remotely close to correct, he would have been dq'ed long ago.
0.14 is actually a pretty big miss, LOL. There are even more extreme cases, though. Paul Caruso in 2016 is another. He played 11 4.0 men's matches as a 4.0S and all 11 are calculated as strikes on TR (and not even close - 4.38 minimum). He was not DQ'd.
 

schmke

Hall of Fame
0.14 is actually a pretty big miss, LOL. There are even more extreme cases, though. Paul Caruso in 2016 is another. He played 11 4.0 men's matches as a 4.0S and all 11 are calculated as strikes on TR (and not even close - 4.38 minimum). He was not DQ'd.
Oaf. I have him with perhaps 2 strikes by the end of the year, the last two were right at the threshold. But agree, no DQ, but a bump up to 4.5 I got this one right.

And @Zman I have the player you mentioned with no strikes, not sure what TR is doing on that one ...
 

Zman

New User
0.14 is actually a pretty big miss, LOL. There are even more extreme cases, though. Paul Caruso in 2016 is another. He played 11 4.0 men's matches as a 4.0S and all 11 are calculated as strikes on TR (and not even close - 4.38 minimum). He was not DQ'd.
I'm sticking with Tinsley Dozier as my current favorite illustration of tennisrecord craziness. Paul Caruso's highest tennisrecord dynamic in 2016 doesn't even crack Tinsley's top 7 tennisrecord dynamics of 2018. And Tinsley's top marks are awesome -- his highest tennisrecord dynamics are 5.23, 5.19, 5.00, 4.92, 4.75. So tennisrecord must have been off by more than 0.80(!!!!) in at least one of Tinsley's dynamics, since Tinsley didn't get dq'ed from 4.0, despite being a 4.0S.
 

kevrol

Hall of Fame
Here are the release dates for the last 5 days according to schmke.
  • 2017 - Thursday 11/30
  • 2016 - Wednesday 11/30
  • 2015 - Monday November 30
  • 2014 - Monday December 1
  • 2013 - Monday December 2
My guess is they are gonna stick with the 11/30 date which would be a week from Friday.
 

J011yroger

Talk Tennis Guru
Here are the release dates for the last 5 days according to schmke.
  • 2017 - Thursday 11/30
  • 2016 - Wednesday 11/30
  • 2015 - Monday November 30
  • 2014 - Monday December 1
  • 2013 - Monday December 2
My guess is they are gonna stick with the 11/30 date which would be a week from Friday.
But I wanna know now!

J
 

schmke

Hall of Fame
Here are the release dates for the last 5 days according to schmke.
  • 2017 - Thursday 11/30
  • 2016 - Wednesday 11/30
  • 2015 - Monday November 30
  • 2014 - Monday December 1
  • 2013 - Monday December 2
My guess is they are gonna stick with the 11/30 date which would be a week from Friday.
Note that those Monday dates were all in the morning I think, the others were late afternoon or 9pm'ish PST, so midnight the next day EST. My guess is that since the Monday would be 12/3, they don't wait for that, and they aren't going to do it on a weekend, and just to make sure they don't do it on a Friday with no buffer or people around to deal with issues as they leave for the weekend, they do it on Thursday the 29th.
 

Nostradamus

Bionic Poster
Note that those Monday dates were all in the morning I think, the others were late afternoon or 9pm'ish PST, so midnight the next day EST. My guess is that since the Monday would be 12/3, they don't wait for that, and they aren't going to do it on a weekend, and just to make sure they don't do it on a Friday with no buffer or people around to deal with issues as they leave for the weekend, they do it on Thursday the 29th.
Have you seen some players that move to another city and self rate themselves to whatever they want and play as fresh new player in USTA ? I suspect this has been happening in some instances. I had some high suspicions of these players in the past but I am such a nice guy that I didn't make noise about it.
 

schmke

Hall of Fame
Have you seen some players that move to another city and self rate themselves to whatever they want and play as fresh new player in USTA ? I suspect this has been happening in some instances. I had some high suspicions of these players in the past but I am such a nice guy that I didn't make noise about it.
Sure, there are a few players that re-self-rate, deliberately or by mistake or misunderstanding. And I've seen cases where when it is discovered the memberships are merged. But sometimes it isn't discovered or reported.
 

Nostradamus

Bionic Poster
Sure, there are a few players that re-self-rate, deliberately or by mistake or misunderstanding. And I've seen cases where when it is discovered the memberships are merged. But sometimes it isn't discovered or reported.
Memberships can be merged ?????? Hummm had no idea...... anyway, these guys were ridiculous. These guys were living in another city and played USTA for years and moved and they made a brand new membership under another account and self rated and start playing at what ever division they felt like.....LOL. There should be better security checks, I think.. What about the Iris scan ? I heard that can't be fooled
 

OnTheLine

Hall of Fame
Have you seen some players that move to another city and self rate themselves to whatever they want and play as fresh new player in USTA ? I suspect this has been happening in some instances. I had some high suspicions of these players in the past but I am such a nice guy that I didn't make noise about it.
Unfortunately, yes. And dude ended up MY mixed partner for half of the awful fall season. He had been a 3.5 in another city under Juan Martinez ... moved here and self-rated as a 4.0 under John Martin (yes altering actual name to protect the infamous) ... dude was *maybe* a low 3.5 ... his 4.0 record for 2018 indicates he will be back to 3.5 (or lower) in no time. Wonder what name he will use to get his vanity 4.0 next year ....
 
Top