Tennis Record site, out of business?

#51
I was hovering around 4.4 for the last 5 or 6 matches, but now the match ratings for most of them suddenly no longer show up in the current results match details.
Well, that probably explains it. Assuming the women were exactly equal, your 15 point differential over the other man means your team should have won by a certain margin and you slightly exceeded it so you gained some points. Even if you had double bageled them, I doubt you would have picked up that many points.
 

J011yroger

Talk Tennis Guru
#52
My TR.com rating stayed the same, but something weird happened - looks like someone tried to manually superimpose my men’s rating onto my last two mixed matches, as the math doesn’t add up. If my partner is 3.0, opposing female is 3.0, opposing male is 4.25, and match score is 6-2, 6-1, shouldn’t the match rating be better than 4.3?
I played a match, my partner was 5.02, opposing girl was 5.01, opposing guy was 4.81.

We won 12 games, they won 9.

My match rating was 4.51 my partner's was 5.02.

Shouldn't our match rating be higher than our opponent's rating?

No respect.

J
 
#53
I played a match, my partner was 5.02, opposing girl was 5.01, opposing guy was 4.81.

We won 12 games, they won 9.

My match rating was 4.51 my partner's was 5.02.

Shouldn't our match rating be higher than our opponent's rating?

No respect.

J
This is more evidence of what I think TR is doing. For a short while, their algorithms actually started to make sense. But now they messed with it again and the algorithms for mixed and men’s doubles are no longer independent or logical. They are trying to force-fit outside data from your men’s matches to come up with your match rating for a given mixed match. They have made it waaaaay too insensitive to individual match performance (in fact the mixed rating sensitivity is now dialed down to zero!), so I hope they hire Nate Silver or something and get their act together soon.

Edit: I looked at your record and they still have their old algorithm in effect for yours. The case you mention was actually logical because you won the match while winning fewer games than your opponents.

Edit 2: I rechecked my tr record and see that only the last match in the system is fishy. They have match ratings going back for the last 7 matches for me. My average tr rating for the prior 6 matches before the illogical match rating was 4.45 (all mixed). Based on the score of my last match,my rating for the last match should have been about 4.6 and my overall rating should have been about 4.45. But I speculate tr just slapped in a 4.3 on the last match because that’s what they have my men’s rating at based on a couple of matches I played over a year ago.
 
Last edited:

J011yroger

Talk Tennis Guru
#54
This is more evidence of what I think TR is doing. For a short while, their algorithms actually started to make sense. But now they messed with it again and the algorithms for mixed and men’s doubles are no longer independent or logical. They are trying to force-fit outside data from your men’s matches to come up with your match rating for a given mixed match. They have made it waaaaay too insensitive to individual match performance (in fact the mixed rating sensitivity is now dialed down to zero!), so I hope they hire Nate Silver or something and get their act together soon.

Edit: I looked at your record and they still have their old algorithm in effect for yours. The case you mention was actually logical because you won the match while winning fewer games than your opponents.

Edit 2: I rechecked my tr record and see that only the last match in the system is fishy. They have match ratings going back for the last 7 matches for me. My average tr rating for the prior 6 matches before the illogical match rating was 4.45 (all mixed). Based on the score of my last match,my rating for the last match should have been about 4.6 and my overall rating should have been about 4.45. But I speculate tr just slapped in a 4.3 on the last match because that’s what they have my men’s rating at based on a couple of matches I played over a year ago.
I won more games.

J
 

time_fly

Professional
#55
One thing I'm wondering about: every season I probably play at least 2 or 3 matches against someone who doesn't have an official rating yet. In those cases, TR and probably USTA use my rating to figure out theirs (sorry guys, that means you automatically stink no matter how you played) and my rating stays the same. But after I play another match, does that previous match count in the "moving average" used to calculate my dNTRP? It shouldn't, since my rating wasn't really affected by it.
 

J_R_B

Hall of Fame
#59
Yea, I understand that, it's like if you lose you suck and if you win it's because your opponent sucked.

J
Here's the answer to your mystery. Your male opponent is S-rated in his second match. Therefore, he didn't actually have a DNTRP yet (you don't get one until after your second match). Therefore, all this match did was set a rating for him relative to the DNTRP for the other three players and the score. They are "showing" you getting a match rating of 4.51, but in reality, you didn't get one at all because the match was used to set the rating for the S-rated player who didn't have enough matches for a rating coming in.
 

J011yroger

Talk Tennis Guru
#60
Here's the answer to your mystery. Your male opponent is S-rated in his second match. Therefore, he didn't actually have a DNTRP yet (you don't get one until after your second match). Therefore, all this match did was set a rating for him relative to the DNTRP for the other three players and the score. They are "showing" you getting a match rating of 4.51, but in reality, you didn't get one at all because the match was used to set the rating for the S-rated player who didn't have enough matches for a rating coming in.
Aha! Sabotaged!

J
 
#62
TR is updating matches but still haven't corrected all the errors made with Appeal players. I still have a match from last fall (but after the deadline for counting towards 2018) that shows my 4.0 to 3.5A partner getting a 3.04 for our match.
 
#65
Played my first men’s match in more than a year last week. Didn’t win, but Tennis Record decided that I improved a lot this week. I still have suspicion that participation points are part of the algorithm.
 

schmke

Hall of Fame
#66
Played my first men’s match in more than a year last week. Didn’t win, but Tennis Record decided that I improved a lot this week. I still have suspicion that participation points are part of the algorithm.
I doubt the actual rating TR has is accurate, but the direction of movement might be correct as in a plus league, you can play against one (or two) 5.0s and a close loss can be better than expected. My ratings did show you going up a bit.
 

time_fly

Professional
#67
One thing I notice looking at our team rosters is that TR seems to weigh scores more heavily than the quality of the opponent. On many of our club rosters the guys routinely playing D2 and D3 are rated far ahead of the D1 and S1 players, even though anyone with two eyes can see the D1 and S1 players are better. But there's some weak cream playing the lower courts in most leagues so you see more lopsided wins there.
 
#69
While I do appreciate the updates, I wish they would correct the issue with some appeal players starting at the wrong rating (ex...3.5A from 4.0 to 3.5 starting at 3.01 instead of 3.51). Until that is fixed, my rating is way off.
 
#70
While I do appreciate the updates, I wish they would correct the issue with some appeal players starting at the wrong rating (ex...3.5A from 4.0 to 3.5 starting at 3.01 instead of 3.51). Until that is fixed, my rating is way off.
Not sure if I already posted, but i am a 4.5A from 5.0 and it started me at 4.51 which seems reasonable.
 
#72
I'm an appeal player as well, and mine has fluctuated from the 4.00 it started me at with each match. As of now it has me at 4.07. I have a split set win against weaker players and a loss against comparable players that hasn't been scraped yet. Curious to see if those results put me back down below 4.
 
Top