Understanding the Tennis record algo

[Moon Shooter]

I am assuming the "match rating" in TR is the basically the "performance rating" of the player for that match.

Match1: between a player that had a dynamic rating of 2.89 (at the time of the match) and a player that had a 3.19 at the time of the match. The match was won by the 3.19 player 6-0 6-0. The "match rating" of the losing, 2.89, player was 2.71 and the "match rating" of the 3.19 player was 3.37.

Match 2: A dynamic rated player of 3.17 (at the time of the match) beat a 3.77 player 6-0 6-0. There was no match rating for the 3.17 player because it was only the third match for the 3.77 player so it just said "S" for match rating for the 3.17 player. But the 3.77 player got a match rating of 2.69.

So it seems a 6-0 6-0 loss should equate to a "match rating" of .48 below the other players rating. And a 6-0 6-0 win should equate to a performance rating of .48 above the other players rating.

Just roughly looking it looks like a singles match 6-1 6-1 will mean a .38 gain or loss over your opponents dynamic rating. And a 6-2 6-0 loss will be a plus or minus .39.

Now I am assuming that the actual result is weighted mush less (or not at all) when the dynamic ratings going in are wider. But if a mid 3.0 with a dynamic rating of 2.75 played up against a higher 3.5 with a dynamic of 3.39 the higher rated player could not possibly gain any rating points and may not even be able to avoid losing some rating points. The 3.39 could at best get a performance rating of 3.23. The 2.75 could not possibly lose any rating points and may be assured to gain some rating points even with a 6-0 6-0 loss. His performance rating would be a 2.91 at worst.

 

[BallBag]

I remember reading something about your match rating is averaged with your previous three dynamic ratings. Maybe it was on @schmke's blog. The match rating is based on a table with half a point difference is 6-0 6-0 and a tenth is 6-4 6-4.

 

[TennisOTM]

I suspect that neither TR nor any other sensible algorithm would allow a shutout win to cause your (dynamic) rating to go down. TR may display a match rating for a 6-0, 6-0 win that is less than your pre-match dynamic rating, but then perhaps does not actually use that value in the calculation of your new dynamic rating in the next column?

In general, I've found that it's pretty easy to reproduce TR's calculation of the match rating, but less easy to recreate their match-by-match updates to the dynamic rating. Averaging the new match rating with the prior 3 dynamic ratings sometimes reproduces their result, but it's easy to find exceptions.

 

[ChaelAZ]

I wish TR would give their rating as they do, but also a further estimated rating that included S rated matches. If possible I mean.

 

[Moon Shooter]

I suspect that neither TR nor any other sensible algorithm would allow a shutout win to cause your (dynamic) rating to go down. TR may display a match rating for a 6-0, 6-0 win that is less than your pre-match dynamic rating, but then perhaps does not actually use that value in the calculation of your new dynamic rating in the next column?

In general, I've found that it's pretty easy to reproduce TR's calculation of the match rating, but less easy to recreate their match-by-match updates to the dynamic rating. Averaging the new match rating with the prior 3 dynamic ratings sometimes reproduces their result, but it's easy to find exceptions.

I don't know what TR is doing. I just looked at my own rating. And going into a match I had an overall rating of 3.15. I got a match rating of 3.07 and ended up with a 3.06 overall rating. Who knows.

 

[Moveforwardalways]

This spring season my overall rating was lower than each of my individual match ratings. Not sure how they calculate that.

 

[Moon Shooter]

Who runs tr? I suspect there is some connection with usta - albeit an informal one.

 

[silverwyvern4]

Does USTA website have an API so that UTR and Tennis Record can get results? Or does TR have to manually use a web crawler

 

[travlerajm]

I don't know what TR is doing. I just looked at my own rating. And going into a match I had an overall rating of 3.15. I got a match rating of 3.07 and ended up with a 3.06 overall rating. Who knows.

I think the overall rating is based on an avg of x number of matches. So if your new 3.07 match rating caused a really good match rating from 5-10 matches ago (maybe a 3.40?) to drop out of your average, then your avg rating would drop a lot.

I’m not sure how they calculate the average though.

 

[Moon Shooter]

Maybe something like that is going on but my highest match rating was the match right before that match- singles match rating of 3.52. Rating in my area are a bit fubar because I’m trying to create new usta teams and so there are a large number of self rates. That said it’s hard to explain what happened there - unless it discounts your highest and lowest results or something. But then again mfa seems to provide another example of this algo seeming to be broken.

 

[schmke]

Who runs tr? I suspect there is some connection with usta - albeit an informal one.

Don't think so. I spoke with someone from the USTA several years ago and they didn't know who ran it and asked me if I did (I don't) or if I knew who did. Everything the USTA presents goes out of their way to say TR doesn't know the algorithm and is often way off and shouldn't be trusted.

 

[travlerajm]

Don't think so. I spoke with someone from the USTA several years ago and they didn't know who ran it and asked me if I did (I don't) or if I knew who did. Everything the USTA presents goes out of their way to say TR doesn't know the algorithm and is often way off and shouldn't be trusted.

It’s almost as if the usta feels threatened by other algorithms.

 

[travlerajm]

I’ve tried a couple of times to reverse engineer the TR algorithm. I’ve gotten close, but haven’t quite cracked it.

There are two separate parts to this.
1. The formula for calculating the match rating.
2. The formula for calculating the average rating.

Im pretty certain that the averaging formula throws out low outliers.

 

[time_fly]

I find the current Tennis Record algorithm to be a little too unstable. I think its estimates of the odds of score differentials based on rating may be too extreme. I'm not sure where that information comes from. All I know is that when I win matches this year, I often get a match rating in the area of 3.40 to 3.65, but when I lose I often get a match rating in the 2.80 to 3.05 range. These can be matches on back-to-back days and clearly my level of play can't be changing that much, so to the extent that it's trying to estimate my level based on match results it's not doing a very accurate job.

 

[travlerajm]

I find the current Tennis Record algorithm to be a little too unstable. I think its estimates of the odds of score differentials based on rating may be too extreme. I'm not sure where that information comes from. All I know is that when I win matches this year, I often get a match rating in the area of 3.40 to 3.65, but when I lose I often get a match rating in the 2.80 to 3.05 range. These can be matches on back-to-back days and clearly my level of play can't be changing that much, so to the extent that it's trying to estimate my level based on match results it's not doing a very accurate job.

It would be interesting to collect data on the standard deviation of TR match ratings for each player. I will calculate for myself for a few years and report back.

 

[travlerajm]

It would be interesting to collect data on the standard deviation of TR match ratings for each player. I will calculate for myself for a few years and report back.

My TR match ratings, all mixed, for the last year have ranged from as low as 4.24 to as high as 4.67. My official TR average mixed rating has ranged from as low as 4.40 to as high as 4.50.

 

[Moon Shooter]

Don't think so. I spoke with someone from the USTA several years ago and they didn't know who ran it and asked me if I did (I don't) or if I knew who did. Everything the USTA presents goes out of their way to say TR doesn't know the algorithm and is often way off and shouldn't be trusted.

That is interesting but I would think they have an idea who runs it now. I know officially they always go out of their way to say it’s unreliable. But in heather’s interview she also said something like: we used these third party ratings to see if it would influence players. If they don’t have any idea how many people even access sites like this how would she be able to determine the effect it is having? Also she didn’t say too much concrete against it other than to say they can get two people with the same name mixed up.

 

[schmke]

That is interesting but I would think they have an idea who runs it now. I know officially they always go out of their way to say it’s unreliable. But in heather’s interview she also said something like: we used these third party ratings to see if it would influence players. If they don’t have any idea how many people even access sites like this how would she be able to determine the effect it is having? Also she didn’t say too much concrete against it other than to say they can get two people with the same name mixed up.

They very well may now. What I shared occurred several years ago.

And what I've seen Heather say in other webinars, and shared with me directly when I've talked with her, is that she got complaints from players saying their captain didn't play them because TLS/TR said they were a weak 3.5 (say 3.10) when in fact Heather looked up their actual dynamic rating and it was a high 3.5 (say 3.40). She cited the unreliability of the ratings causing captains to make bad line-up decisions as one of the negative effects of captains using TLS/TR.

 

[travlerajm]

My local league coordinator keeps telling me that I shouldn’t trust TR because it can be inaccurate. Then she said, ‘you know… there’s a new rating out now. You should look at that?’

It may very well be the case that TR doesn’t agree with the usta computer. But at the very least, TR’s algorithm seems simple, logical, and mostly transparent in that it updates with individual match ratings (something none of the other public rating sites do).

 

[Moon Shooter]

TR is not supposed to be it’s own rating. It is supposed to try to mirror usta. In my experience it does a pretty good job of that. I have not seen people being off by over .3 as heather described but I am sure it does happen - either because the name draws the wrong person - including an opponent - or because there are few games so a few calls that are off with partners and opponents can exaggerate the algo’s differences.

I agree that tr is transparent and that is good. I never understood why the companies think their own Elo or Glicko system is some super secret formula that no one else should ever know. These systems have been out for decades and their tweaks are no big deal. The marketing they do around this is just idiotic and makes them lose credibility. The service they do is in data management and entering the scores and giving other analysis tools. Those purposes would be better served if people understood how they get the number they do instead of having a number fall from the sky.

But it is hard to call tr logical when you have an overall rating of 3.15 get a performance rating
/match rating of 3.07 and then immediately end up with an overall rating of 3.06.

 

[travlerajm]

TR is not supposed to be it’s own rating. It is supposed to try to mirror usta. In my experience it does a pretty good job of that. I have not seen people being off by over .3 as heather described but I am sure it does happen - either because the name draws the wrong person - including an opponent - or because there are few games so a few calls that are off with partners and opponents can exaggerate the algo’s differences.

I agree that tr is transparent and that is good. I never understood why the companies think their own Elo or Glicko system is some super secret formula that no one else should ever know. These systems have been out for decades and their tweaks are no big deal. The marketing they do around this is just idiotic and makes them lose credibility. The service they do is in data management and entering the scores and giving other analysis tools. Those purposes would be better served if people understood how they get the number they do instead of having a number fall from the sky.

But it is hard to call tr logical when you have an overall rating of 3.15 get a performance rating
/match rating of 3.07 and then immediately end up with an overall rating of 3.06.

How do we know what TR is “supposed to” do?

It’s clear that TR force fits ratings at the end of the season by adjusting ratings to match the usta computer rating band. I see this as simply an annual recalibration.

But other than that, it seems to be its own independent algorithm. If it was trying to mirror the usta dynamic ratings and become better at predicting bumps, then we would see an effort to reduce the number of out-of-band mismatches, with the number continually decreasing each year. But it doesn’t seem that that is happening. I conclude that whoever runs TR doesn’t care about that.

There were a couple obvious algorithm glitches early on when TR first came out, but those bugs were quickly addressed, and now it has just been chugging along doing its thing.

 

[Moveforwardalways]

TR’s website clearly states that it is not trying to mimic USTA, and that it is it’s own rating. Also, being off by 0.3 is HUGE as that is over half a level, and is difference between top of level and bottom of level. And yes, captains use TR not only to create lineups but to determine whether you should be on a team in the first place. While it’s not perfect, until USTA publishes their ratings or until @schmke posts his for free, TR is all captains have to go on. They cannot simply tell, by an eyeball test or by a match they did not see, which partner was the reason a doubles pairing lost a match or whether a loss was a “good loss” or how good the opponents were.

 

[Moon Shooter]

How do we know what TR is “supposed to” do?

It’s clear that TR force fits ratings at the end of the season by adjusting ratings to match the usta computer rating band. I see this as simply an annual recalibration.

But other than that, it seems to be its own independent algorithm.

Yes it completely recalibrates it ratings to fit usta bumps which I consider a pretty big deal.
It also never includes matches that are not USTA specific matches.

But it also refuses to rate even USTA matches except in the odd ways USTA does. They refuse to rate mixed doubles and combine them with non-mixed ratings. Even with mixed ratings it also follows the odd usta rule of not including combo mixed matches in your mixed rating. I mean think about that. It counts all your 6.0, 7.0, 8.0 and 9.0 mixed matches but not the 6.5, 7.5, and 8.5! Is it really doing it's own thing and just coincidently happened to decide to to that?

So for example I played 5 mixed doubles matches in a 6.5 league in 2021. But then when I started a 7.0 mixed league in January of 2022 none of those matches counted for my rating and I was treated as a brand new self rate for my partner and opponents. And then I added 3 7.0 mixed matches but that mixed rating had no impact on my adult rated game.

If it was trying to mirror the usta dynamic ratings and become better at predicting bumps, then we would see an effort to reduce the number of out-of-band mismatches, with the number continually decreasing each year. But it doesn’t seem that that is happening. I conclude that whoever runs TR doesn’t care about that.

I think given it completely recalibrates to usta and how closely it mirrors very odd decisions of USTA in what games are rated and how, it is obvious they are trying to at least give the appearance of tracking USTA.

"We have taken the knowledge we have learned and have developed our own estimated computer ratings."

The title says "Statistical Analysis and Estimated Tennis Ratings to the 10,000th of a Point"

TennisRecord.com - Estimated Tennis Ratings and Statistics

Research estimated ratings and stats for adult and junior tennis players.

www.tennisrecord.com

Our own "estimated" computer ratings? The site says it is an estimated computer rating. If it wasn't trying to estimate the USTA it would just say it *is* a computer rating. WTN *is* a computer rating. NTRP and UTR *are* computer ratings. They are not trying to "estimate" some other computer rating. They are giving you a computer rating.

I think there can be some ambiguity there. I mean they do say it is our "own" estimated computer rating. Which is true just like Schmke has his own estimated computer rating. Moreover some of these others ratings might say their number is an estimate of your skill level. But there is a difference between saying it is an "estimate of skill" and "estimated tennis rating." I seems to me they are primarily "estimating your rating" and only secondarily "estimating your skill".

Now if they wanted to say they are estimating your skill then they would measure the predictive value of the ratings they give and then adjust them to increase the predictive value of their rating. But instead they just keep the two different ratings. So someone that is a 3.0 self rate but only played mixed and has a dynamic mixed rating that puts them in the 4.0 category is still listed as a 3.0 player. Why? Because that is what USTA says and that is what they are trying to estimate.

There were a couple obvious algorithm glitches early on when TR first came out, but those bugs were quickly addressed, and now it has just been chugging along doing its thing.

Yes it is doing its thing and IMO it's thing is to try at least to some extent estimate USTA dynamic ratings.

I am not sure the glitches have been fixed. And I would not be at all surprised if the people running it have had relationships with people connected to usta ratings. I wouldn't be surprised if they didn't either. I just find it odd that they do not reveal their identities.

 

[Moon Shooter]

TR’s website clearly states that it is not trying to mimic USTA, and that it is it’s own rating. Also, being off by 0.3 is HUGE as that is over half a level, and is difference between top of level and bottom of level. And yes, captains use TR not only to create lineups but to determine whether you should be on a team in the first place. While it’s not perfect, until USTA publishes their ratings or until @schmke posts his for free, TR is all captains have to go on. They cannot simply tell, by an eyeball test or by a match they did not see, which partner was the reason a doubles pairing lost a match or whether a loss was a “good loss” or how good the opponents were.

I agree that TR is better than nothing for picking line ups. I think after WTN works out some issues it will be a better rating. But TR is helpful in that it gives performance ratings. I mean you can kind of guesstimate them with WTN given the match score and everyone's rating. But it is convenient that TR gives that. That way you can see how much someone's rating got deflated/inflated because a self rate was not properly rated or due to a team mate or opponent getting injured or exhausted mid match. TR's format makes it easy to see why someone might have an artificially low or high TR dynamic rating. WTN forces you to do more work.

Another obvious example is TR lists your team by dynamic rating! WTN you have to keep searching for each individual player.

 

[travlerajm]

Yes it completely recalibrates it ratings to fit usta bumps which I consider a pretty big deal.
It also never includes matches that are not USTA specific matches.

But it also refuses to rate even USTA matches except in the odd ways USTA does. They refuse to rate mixed doubles and combine them with non-mixed ratings. Even with mixed ratings it also follows the odd usta rule of not including combo mixed matches in your mixed rating. I mean think about that. It counts all your 6.0, 7.0, 8.0 and 9.0 mixed matches but not the 6.5, 7.5, and 8.5! Is it really doing it's own thing and just coincidently happened to decide to to that?

So for example I played 5 mixed doubles matches in a 6.5 league in 2021. But then when I started a 7.0 mixed league in January of 2022 none of those matches counted for my rating and I was treated as a brand new self rate for my partner and opponents. And then I added 3 7.0 mixed matches but that mixed rating had no impact on my adult rated game.




I think given it completely recalibrates to usta and how closely it mirrors very odd decisions of USTA in what games are rated and how, it is obvious they are trying to at least give the appearance of tracking USTA.

"We have taken the knowledge we have learned and have developed our own estimated computer ratings."

The title says "Statistical Analysis and Estimated Tennis Ratings to the 10,000th of a Point"

TennisRecord.com - Estimated Tennis Ratings and Statistics

Research estimated ratings and stats for adult and junior tennis players.

www.tennisrecord.com

Our own "estimated" computer ratings? The site says it is an estimated computer rating. If it wasn't trying to estimate the USTA it would just say it *is* a computer rating. WTN *is* a computer rating. NTRP and UTR *are* computer ratings. They are not trying to "estimate" some other computer rating. They are giving you a computer rating.

I think there can be some ambiguity there. I mean they do say it is our "own" estimated computer rating. Which is true just like Schmke has his own estimated computer rating. Moreover some of these others ratings might say their number is an estimate of your skill level. But there is a difference between saying it is an "estimate of skill" and "estimated tennis rating." I seems to me they are primarily "estimating your rating" and only secondarily "estimating your skill".

Now if they wanted to say they are estimating your skill then they would measure the predictive value of the ratings they give and then adjust them to increase the predictive value of their rating. But instead they just keep the two different ratings. So someone that is a 3.0 self rate but only played mixed and has a dynamic mixed rating that puts them in the 4.0 category is still listed as a 3.0 player. Why? Because that is what USTA says and that is what they are trying to estimate.



Yes it is doing its thing and IMO it's thing is to try at least to some extent estimate USTA dynamic ratings.

I am not sure the glitches have been fixed. And I would not be at all surprised if the people running it have had relationships with people connected to usta ratings. I wouldn't be surprised if they didn't either. I just find it odd that they do not reveal their identities.

By doing the annual calibration, it ensures that its ratings are staying roughly in the same range as the usta ntrp bands. Otherwise, it could gradually drift in one direction and end up with, for example, an average usta 4.0 being rated mid 4.5. The calibration ensures that an average usta 4.0 is also an average TR 4.0, statistically speaking.

 

[travlerajm]

Another thing that is useful about TR:

Even though singles and doubles are not tracked with separate averages, a captain can easily record the match ratings and then create his/her own averages for both. For example, player might be a TR 3.21, but averaging his last 10 singles results gives a 3.07, while his last 10 doubles results gives a 3.43. The captain would then know that this player’s TR rating is underrating him for doubles and overrating him for singles.

 

[schmke]

TR’s website clearly states that it is not trying to mimic USTA, and that it is it’s own rating.

This very well may be to avoid the wrath of the USTA.

 

[jm1980]

I agree that tr is transparent and that is good. I never understood why the companies think their own Elo or Glicko system is some super secret formula that no one else should ever know. These systems have been out for decades and their tweaks are no big deal. The marketing they do around this is just idiotic and makes them lose credibility. The service they do is in data management and entering the scores and giving other analysis tools. Those purposes would be better served if people understood how they get the number they do instead of having a number fall from the sky.

Usually the rationale behind keeping the formula as secret as possible, is to prevent players from manipulating their ratings one way or another

 

[travlerajm]

Usually the rationale behind keeping the formula as secret as possible, is to prevent players from manipulating their ratings one way or another

I’m currently trying to manipulate my rating and make it go up.

 

[Moveforwardalways]

I agree that TR is better than nothing for picking line ups. I think after WTN works out some issues it will be a better rating. But TR is helpful in that it gives performance ratings. I mean you can kind of guesstimate them with WTN given the match score and everyone's rating. But it is convenient that TR gives that. That way you can see how much someone's rating got deflated/inflated because a self rate was not properly rated or due to a team mate or opponent getting injured or exhausted mid match. TR's format makes it easy to see why someone might have an artificially low or high TR dynamic rating. WTN forces you to do more work.

Another obvious example is TR lists your team by dynamic rating! WTN you have to keep searching for each individual player.

I don’t know if WTN is granular enough. If you are a captain of a team of 4.0 players, are all the players’ WTN’s going to be different enough to know who is playing better? Then again, can you trust TR for that? Not sure.

 

[travlerajm]

I don’t know if WTN is granular enough. If you are a captain of a team of 4.0 players, are all the players’ WTN’s going to be different enough to know who is playing better? Then again, can you trust TR for that? Not sure.

TR seems to be a fairly reliable indicator of level as long as you break out singles, doubles, and mixed results. WTN still seems to be in the beta stage where it has some bugs that are spitting out unrealistic ratings for some players.

 

[time_fly]

I don’t know if WTN is granular enough. If you are a captain of a team of 4.0 players, are all the players’ WTN’s going to be different enough to know who is playing better? Then again, can you trust TR for that? Not sure.

I generally like how TR presents the data on players so I often use it for scouting and sometimes for lineup help. But when estimating the strength of a player I look at the number of matches played and win-loss ratio as well as looking at the pseudo-NTRP.

 

[Bodhi312]

They very well may now. What I shared occurred several years ago.

And what I've seen Heather say in other webinars, and shared with me directly when I've talked with her, is that she got complaints from players saying their captain didn't play them because TLS/TR said they were a weak 3.5 (say 3.10) when in fact Heather looked up their actual dynamic rating and it was a high 3.5 (say 3.40). She cited the unreliability of the ratings causing captains to make bad line-up decisions as one of the negative effects of captains using TLS/TR.

The person who runs USTA for Chicago is a friend.

She says that TR and USTA aren’t affiliated in any way. At best, TR can be used as an indicator. Also, she can’t see a given player’s dynamic rating (so she didn’t know who was going to be bumped up/down this past December or compare TR for accuracy).

There’s no contact information on the TR site (which seems a little odd).

 

[ShaunS]

How do we know what TR is “supposed to” do?

It’s clear that TR force fits ratings at the end of the season by adjusting ratings to match the usta computer rating band. I see this as simply an annual recalibration.

IMO that's illogical. TR sets out to estimate your NTRP rating, and that's obviously why they rate the likelihood of you getting bumped and do *corrections* to align with what USTA NTRP says. I'm not sure if they still do this last thing, but they even used to tell you if you'd be within the range of a USTA appeal.

The fact that they call it a rating estimation even suggests they're trying to determine your NTRP. Honestly, I had no idea anyone thought TR was doing something other than this.

TR’s website clearly states that it is not trying to mimic USTA, and that it is it’s own rating.

Where does it say that? All I actually see is this:

This site is not affiliated, endorsed, or associated by or with the USTA in any manner nor are the ratings affiliated with the NTRP rating system.

This is basic legalese to avoid conflict with the USTA, and it's definitely not saying they're doing their own rating. Only that there's no official relationship.

Otherwise, it could gradually drift in one direction and end up with, for example, an average usta 4.0 being rated mid 4.5. The calibration ensures that an average usta 4.0 is also an average TR 4.0, statistically speaking.

Again, the plausibility test here. Why would TR care if their ratings were different if they weren't trying to match the NTRP?

I'm beating a dead horse here, but one last question for the people who believe TR might not be trying to estimate NTRP. Are you on the mailing list for TennisRecord? The Q&A emails they send out might do more to shift your belief than just my words.

I generally like how TR presents the data on players so I often use it for scouting and sometimes for lineup help. But when estimating the strength of a player I look at the number of matches played and win-loss ratio as well as looking at the pseudo-NTRP.

100% agree. It's often a much better place to lookup data on players/teams than the nightmarish TennisLink site. You just want to be careful trusting their ratings because there's plenty of evidence (anecdotal and otherwise) that they struggle a lot.

I also enjoy the Advanced Player Stats page.

 

[Moveforwardalways]

No one would argue that TR doesn’t have the best ratings website right now. That’s beyond question. How accurate it is - that is not beyond question.

 

[FuzzyYellowBalls]

No one would argue that TR doesn’t have the best ratings website right now. That’s beyond question. How accurate it is - that is not beyond question.

I prefer UTR, my hard to deny argument: If an adult plays 3-100 or whatever UTR tournaments per year or does a UTR flex league those results won't be in TR.

A smaller point, no one tanks UTR on purpose unless they are just insane, there are plenty of USTA players I know about that tank USTA on purpose, but play for real in UTR consistently. Of course UTR captures the tanked USTA results, but still, at least when they play hard it's also in there.

Even smaller point, UTR captures recent college and high school results, so you can see the what the self-raters have been up to .

 

[southernncgal]

Does anyone have the chart of match expectations for a doubles match and/or how to calculate the match rating at the end of the match? TEnnis Record hasn't pulled over tourn data in three month this summer and our tournaments count towards ratings here in the South. Schmke doesn't do calculations for tournaments but if anyone knows how to do this or will do this for me, I'd be more than happy to pay you a good fee. I just need to know if it's worth traveling around to these tournaments, spending $, etc. so I can at least see if I am in the right direction for my goal. Where else in life do we set goals, have no idea where we stand all year long and then press a magic button at the end of the year to see if we achieved our goal, or not?!?! I'm just a simple 3.0 trying to get back to 3.5. I have zero tricks up my sleeve nor would i want to. Most women in my situation want to move up. I don't see any trying to manipulate the system. We just want to improve and get to be chosen by better teams and play more competitive tennis. HELP!!

 

[nyta2]

Does anyone have the chart of match expectations for a doubles match and/or how to calculate the match rating at the end of the match? TEnnis Record hasn't pulled over tourn data in three month this summer and our tournaments count towards ratings here in the South. Schmke doesn't do calculations for tournaments but if anyone knows how to do this or will do this for me, I'd be more than happy to pay you a good fee. I just need to know if it's worth traveling around to these tournaments, spending $, etc. so I can at least see if I am in the right direction for my goal. Where else in life do we set goals, have no idea where we stand all year long and then press a magic button at the end of the year to see if we achieved our goal, or not?!?! I'm just a simple 3.0 trying to get back to 3.5. I have zero tricks up my sleeve nor would i want to. Most women in my situation want to move up. I don't see any trying to manipulate the system. We just want to improve and get to be chosen by better teams and play more competitive tennis. HELP!!

imo, more useful goals to set for yourself on based on skill acquisition... vs. achieving a utr/ntrp (especially at the 3.0 level - major skill deficiencies)
examples:
* elim doubles in my next match by learning at heavy slice or kicker... how many spin serves in a row can you hit (ideally 10/10)
* develop a reliable groundstroke (topspin and slice) game (can i maintain a 25 hit (each person) cooperative rally?)
- vs. waist high balls
- vs. moon balls
- vs. slicers?
* at 3.0-3.5, do you even have volley&overhead skills?
* fitness... can you run a 8,7,6 min mile?
etc...
if you focus on specific skill acquisition... you'll catapult to 4.0 (ie. knows how to consistently hit all shots) while the rest are still in 3.0-3.5 purgatory.

 

[loveallcats]

Does anyone have the chart of match expectations for a doubles match and/or how to calculate the match rating at the end of the match? TEnnis Record hasn't pulled over tourn data in three month this summer and our tournaments count towards ratings here in the South. Schmke doesn't do calculations for tournaments but if anyone knows how to do this or will do this for me, I'd be more than happy to pay you a good fee. I just need to know if it's worth traveling around to these tournaments, spending $, etc. so I can at least see if I am in the right direction for my goal. Where else in life do we set goals, have no idea where we stand all year long and then press a magic button at the end of the year to see if we achieved our goal, or not?!?! I'm just a simple 3.0 trying to get back to 3.5. I have zero tricks up my sleeve nor would i want to. Most women in my situation want to move up. I don't see any trying to manipulate the system. We just want to improve and get to be chosen by better teams and play more competitive tennis. HELP!!

I'm not Shmeke or anywhere close to his awesomeness, but I've managed to pretty closely reproduce the tennis record results with my own Excel sheet, using the info about partner's and opponent's ratings as a starting point (so obv if that is wrong, my result will be wrong too). I've compared my results to tennis records (I run mine match by match before theirs come out) and it's pretty darn close.
I suspect you can create a similar calculation in Excel if you follow some of the assumptions that are out there:
*Between 3.0 and 3.5 expected score is 6-0, 6-0 therefore each game is worth approx 0.0417
*If playing doubles, your scores and opponents scores are averaged and then you figure out what the expected result is. Your match score is determined by how much you over or underperform against this.
*Your new dynamic rating is your newest match score averaged with your last three dynamic ratings.

 

[time_fly]

Most women in my situation want to move up.

Be careful of what you wish for. Getting bumped up often means going from being a key player on a winning team to a bench player on a winning team, or a lower courts starter on a less successful team.

Aside from that, work on your game and the rating will take care of itself.

 

[southernncgal]

I suspect that neither TR nor any other sensible algorithm would allow a shutout win to cause your (dynamic) rating to go down. TR may display a match rating for a 6-0, 6-0 win that is less than your pre-match dynamic rating, but then perhaps does not actually use that value in the calculation of your new dynamic rating in the next column?

In general, I've found that it's pretty easy to reproduce TR's calculation of the match rating, but less easy to recreate their match-by-match updates to the dynamic rating. Averaging the new match rating with the prior 3 dynamic ratings sometimes reproduces their result, but it's easy to find exceptions.

 

[southernncgal]

Curious as to how you can guesstimate a match rating? I hate that tennis record is not carrying over tourn data for months now. It helped me with their match ratings so I could accumulate all those results with our USTA season since tournaments count in the South Region towards final yearly NTRP ratings. For instance, My rating currently (according to TR but not including tournaments is 2.75. I played a player 3.19 and lost 6-3, 0-6 and 0-1 in tiebreaker. Any idea what match rating that would be? The next day i played a 3.39 player and lost 5-7, 0-6. My other touraments have been double and my partner always moves up more than I do when we win matches. So frutstrating USTA can't share your actual dynamic rating withe people, at least once a year so we know if we are anhywhere near or goal, or not! Thank for any insight! Laurie

 

[TennisOTM]

Curious as to how you can guesstimate a match rating? I hate that tennis record is not carrying over tourn data for months now. It helped me with their match ratings so I could accumulate all those results with our USTA season since tournaments count in the South Region towards final yearly NTRP ratings. For instance, My rating currently (according to TR but not including tournaments is 2.75. I played a player 3.19 and lost 6-3, 0-6 and 0-1 in tiebreaker. Any idea what match rating that would be? The next day i played a 3.39 player and lost 5-7, 0-6. My other touraments have been double and my partner always moves up more than I do when we win matches. So frutstrating USTA can't share your actual dynamic rating withe people, at least once a year so we know if we are anhywhere near or goal, or not! Thank for any insight! Laurie

For singles, your match rating is based on your opponent's pre-match rating and the score of the match only (your own pre-match rating doesn't matter). For the score it should only count the number of games, so your 6-3, 0-6, 0-1 result would be counted as 6-10 in games. If the game score counted that way had been tied then your match rating would be the same as your opponent's pre-match rating (3.19). But you lost more games, so your match score would be something less than 3.19. To get the "something" you can browse around TR to find a rated match where the game score was 6-10, and figure out the answer that way.

 

[silverwyvern4]

For singles, your match rating is based on your opponent's pre-match rating and the score of the match only (your own pre-match rating doesn't matter). For the score it should only count the number of games, so your 6-3, 0-6, 0-1 result would be counted as 6-10 in games. If the game score counted that way had been tied then your match rating would be the same as your opponent's pre-match rating (3.19). But you lost more games, so your match score would be something less than 3.19. To get the "something" you can browse around TR to find a rated match where the game score was 6-10, and figure out the answer that way.

Would a 6-4 0-0 timed match count the same as a 7-6 7-6 win?

 

[silverwyvern4]

Would a 6-4 0-0 timed match count the same as a 7-6 7-6 win or would it count like a 6-4 6-4 win?

 

[ChaelAZ]

Years back there was a lot of threads with specific information on the calculations of ratings, broke down by match differentials for game (from 0,0 to 6,6) tailing ratings, dubs, and more. That might have even been pre Schmke?

 

[TennisOTM]

Would a 6-4 0-0 timed match count the same as a 7-6 7-6 win or would it count like a 6-4 6-4 win?

I believe a higher % of games won gives you a higher match rating, so 60% is better than ~54% in your example. However the match rating then needs to get averaged in to update your overall rating, and that is more of a mystery how that works - it's possible that a timed match gets weighted less heavily when averaged in.

 

[southernncgal]

Years back there was a lot of threads with specific information on the calculations of ratings, broke down by match differentials for game (from 0,0 to 6,6) tailing ratings, dubs, and more. That might have even been pre Schmke?

That would be nice to find!

 

[southernncgal]

imo, more useful goals to set for yourself on based on skill acquisition... vs. achieving a utr/ntrp (especially at the 3.0 level - major skill deficiencies)
examples:
* elim doubles in my next match by learning at heavy slice or kicker... how many spin serves in a row can you hit (ideally 10/10)
* develop a reliable groundstroke (topspin and slice) game (can i maintain a 25 hit (each person) cooperative rally?)
- vs. waist high balls
- vs. moon balls
- vs. slicers?
* at 3.0-3.5, do you even have volley&overhead skills?
* fitness... can you run a 8,7,6 min mile?
etc...
if you focus on specific skill acquisition... you'll catapult to 4.0 (ie. knows how to consistently hit all shots) while the rest are still in 3.0-3.5 purgatory.

Sadly and embarrassing to admit, I once was a 4.0 (when I was 18) and played D3 tennis. I quit exactly then and picked up the racket again three years ago (33 year break) and I finally have my strokes back. So, yes, I have those shots (except spin serves prob 8/10 vs. 10/10). I have a very good deep topspin forehand and my greatest strenght is my slice backhand and drop shots. I was finally picked up to play with high 3.5s at my club. But only because a few played with me and told others my game was better than my rating. BUT, I can't close tournament matches against 3.5s. I lose in a lot of tiebreakers (like close to 1/2). I keep playing up though in tournaments and I keep trying. But, if I just knew how close I was, it would help me immensely. It's a lot of time and $ to travel to tournaments so I'd like to know I'm making progress. Appreciate your response, I'm not giving up

 

[southernncgal]

Be careful of what you wish for. Getting bumped up often means going from being a key player on a winning team to a bench player on a winning team, or a lower courts starter on a less successful team.

Aside from that, work on your game and the rating will take care of itself.

I'm okay with losing against better players. That's how you improve your game. Not my goal to play against 3.0s (which include 2.5s on a 3.0 team) and win everything. I don't improve that way.