Can UTR be trusted to operate and accurate system?

[LOBALOT]

You can win every consolation back draw , and maybe end up with 400 points. That’s not even enough for top 50 in my state.

It gets to a point where you just win level 5 and 3 tournaments to have a higher ranking The kids at the top can’t game the system.

What is harder? To win a level 3 tournament playing daily for 6 days? Mentally. Physically.

Or playing one off utr matches against girls and adults who are playing more or less for fun.

Maybe it is more designed for utr events as playing someone close to your utr is more heavily weighted.

That was one example. There are many more. But it seems like you and your 12 year old have it figured out.

From what I see here I am sure he is bound for a successful time in juniors.

 

[jmnk]

Well UTR thinks they are fake anyway because of 2+ UTR gap rule. Those might as well be fake.

why are you intentionally misrepresenting facts? Two of those resent Donald's matches were against players within 2 UTR points.

 

[travlerajm]

why are you intentionally misrepresenting facts? Two of those resent Donald's matches were against players within 2 UTR points.

So you’re saying his record in the last month justifies his rise from ranking 900 in the world to 650?

 

[jmnk]

How about Jack Sock, 100% reliable UTR 15 singles?

If Sock is still really a 15, then I’m going to try out the pizza training regime too.

same story as with Kyrgios. to repeat "UTR page shows you he has no results for last 12 months, has no 3-months trend. His UTR chart is no longer shown because he does not have enough matches over last 12 months. What else do you want UTR to say to indicate that 'his UTR as shown reflects approximate ratings from the days he used to play - but there's no telling anymore what his current level is'."

really - what would you suggest UTR does?

Donald Young has not played in ATP sanctioned event since Oct 2023. Per your logic it seems you think his UTR should be what - zero? 7? 10? But then he played a non-ATP event in May 2024 and he beat 12.31 UTR player 61 61 and 12.63 UTR 64 62 - so it seems that his UTR ranking of ~13.50 still is fairly accurate. In his case.

 

[travlerajm]

same story as with Kyrgios. to repeat "UTR page shows you he has no results for last 12 months, has no 3-months trend. His UTR chart is no longer shown because he does not have enough matches over last 12 months. What else do you want UTR to say to indicate that 'his UTR as shown reflects approximate ratings from the days he used to play - but there's no telling anymore what his current level is'."

really - what would you suggest UTR does?

Donald Young has not played in ATP sanctioned event since Oct 2023. Per your logic it seems you think his UTR should be what - zero? 7? 10? But then he played a non-ATP event in May 2024 and he beat 12.31 UTR player 61 61 and 12.63 UTR 64 62 - so it seems that his UTR ranking of ~13.50 still is fairly accurate. In his case.

That’s a lot of words to defend Jack sock’s 100% reliability rating.

 

[Tennis2349]

same story as with Kyrgios. to repeat "UTR page shows you he has no results for last 12 months, has no 3-months trend. His UTR chart is no longer shown because he does not have enough matches over last 12 months. What else do you want UTR to say to indicate that 'his UTR as shown reflects approximate ratings from the days he used to play - but there's no telling anymore what his current level is'."

really - what would you suggest UTR does?

Donald Young has not played in ATP sanctioned event since Oct 2023. Per your logic it seems you think his UTR should be what - zero? 7? 10? But then he played a non-ATP event in May 2024 and he beat 12.31 UTR player 61 61 and 12.63 UTR 64 62 - so it seems that his UTR ranking of ~13.50 still is fairly accurate. In his case.

If you go to Nicks page he somehow went up a full .25 in the last 2 months. (10 months after he last played). Why?

My entire point has been that this also happens to juniors. Do not play at all and raise an entire UTR point in 5-7 months. As you see it raised Kyrgios .25 in 2 months. Always 100 percent certain. Pretty much going up at the maximum rate.

Would you agree that going up 1.5 to 2 utr in a year is difficult for juniors that are training daily?

I used to think so, but after looking at several kids that have not played for months they all end up at the same place. Going up.

 

[travlerajm]

I see that Donald Young, 100% reliability UTR 13.52, lost 4 games to a 100% reliability UTR 8 last month.

The TTW faithful tell me that it’s not possible for a UTR 13.5 to lose that many games to someone 5.5 UTR units below him. How is this possible?

 

[Sea70]

Of your sons last 30 matches(within the last year), how many players has he lost to with a higher Utr than his?

How many players has he won against players with a lower Utr than his?


If your kid stays in competitive tennis, I think you’ll see in a few years that though Utr is far from perfect,
it’s the best thing out there by far on tracking your son’s improvement.

 

[tennis4me]

I see that Donald Young, 100% reliability UTR 13.52, lost 4 games to a 100% reliability UTR 8 last month.

The TTW faithful tell me that it’s not possible for a UTR 13.5 to lose that many games to someone 5.5 UTR units below him. How is this possible?

Donald Young is tanking the match so he can play 4.5-5.0 USTA league?

The easiest way to figure out UTR's algorithm is probably to work there as a summer intern

 

[jmnk]

I see that Donald Young, 100% reliability UTR 13.52, lost 4 games to a 100% reliability UTR 8 last month.

The TTW faithful tell me that it’s not possible for a UTR 13.5 to lose that many games to someone 5.5 UTR units below him. How is this possible?

Thank you for providing a perfect example of why a match between players with such UTR disparity should be (and is) disregarded. Exactly because the score is completely meaningless.

 

[travlerajm]

Thank you for providing a perfect example of why a match between players with such UTR disparity should be (and is) disregarded. Exactly because the score is completely meaningless.

But your assumption that it is meaningless is fallacious.

I have played more than 100 practice sets versus players with past UTR at least 4+ UTR above my level. It’s 100x easier to win 4+ games against an “ex” UTR 13 than it is against a current UTR 13.

 

[Tennis2349]

Of your sons last 30 matches(within the last year), how many players has he lost to with a higher Utr than his?

How many players has he won against players with a lower Utr than his?


If your kid stays in competitive tennis, I think you’ll see in a few years that though Utr is far from perfect,
it’s the best thing out there by far on tracking your son’s improvement.

Of course. The utr isn’t totally random.

The tournament this weekend I see many upsets of a point of 2 points. I just feel people should know how they are weighted, and they should count and not be excluded. Especially considering you might not get too many chances to beat kids with a higher utr. (1.5 to 2 points) Not many 8 utr 13 year olds in your area.

Plenty of older 8 utrs, but again, I don’t feel a 13 yr old 8 is as good as a 17 year old 8. Even though the app says they should be

So today, when the 6.25 beat an 8, when will he even play one again? There aren’t tournament every weekend. Maybe in a few months? Next year? To say that “doesn’t count” or “doesn’t count much” is ridiculous.

 

[jmnk]

Of course. The utr isn’t totally random.

The tournament this weekend I see many upsets of a point of 2 points. I just feel people should know how they are weighted, and they should count and not be excluded. Especially considering you might not get too many chances to beat kids with a higher utr. (1.5 to 2 points) Not many 8 utr 13 year olds in your area.

Plenty of older 8 utrs, but again, I don’t feel a 13 yr old 8 is as good as a 17 year old 8. Even though the app says they should be

So today, when the 6.25 beat an 8, when will he even play one again? There aren’t tournament every weekend. Maybe in a few months? Next year? To say that “doesn’t count” or “doesn’t count much” is ridiculous.

What tournament is that? Can you give me some clues so I can find it on UTR site?

 

[jmnk]

But your assumption that it is meaningless is fallacious.

I have played more than 100 practice sets versus players with past UTR at least 4+ UTR above my level. It’s 100x easier to win 4+ games against an “ex” UTR 13 than it is against a current UTR 13.

I do not understand what that means. What are "players with past UTR at least 4+ UTR above my level."? Like players that used to be UTR 13 20 years ago but now can barely walk and are therefore UTR 7? What does it have to do with anything what UTR a player was 'in the past'?

 

[Tennis2349]

What tournament is that? Can you give me some clues so I can find it on UTR site?

It’s a junior level 5 USTA tournament. It’s not on the utr site as of yet.

That’s my point of view. So if you train April and May for the beginning summer cluster of local level 3 and 5 tournaments, pull off a couple upsets, you won’t get credit (or much) in the algorithm.

On top of that, the 6 utr who lost badly to the other 6 utr kid who is beating 8 utr’s will take a big hit to his utr. It supposedly measures matches within range higher.

Can you tell me how much a match is weighted in the algorithm if you beat someone 2 utr above? 1.5? 1.25? 1?

Nobody knows.

People who blindly support utr will say “well maybe the better player had a bad day! He was sick! The lesser player must have had bad results before!” Or if all else fails “it doesn’t matter for 6-8 utr anyway! Overtime it all equals out! It looks close to me!”

 

[tennis4me]

For fun, go here and ask your question to UTR's chat bot that's mostly trained with knowledge from UTR employees. Their disclaimer is not all answers are endorsed by UTR. So maybe those in the know of how to ask a chat bot can somehow coax the bot to spit out the algorithm or indirectly discloses it.

https://support.universaltennis.com.../9000151894-how-is-the-utr-rating-calculated-

Here's a start from a prompt about "inactive players" (rather generic answers that most of you probably already know)

Inactive players may see fluctuations in their UTR due to various factors such as time degradation, changes in the ratings of their past opponents, and the recalibration of the algorithm. As the system is dynamic and recalibrates daily, even without playing new matches, there can be small fluctuations in a player's UTR. These fluctuations are more common for players who have recently joined the system and as more matches are added to their record, their UTR becomes more stable.

 

[travlerajm]

For fun, go here and ask your question to UTR's chat bot that's mostly trained with knowledge from UTR employees. Their disclaimer is not all answers are endorsed by UTR. So maybe those in the know of how to ask a chat bot can somehow coax the bot to spit out the algorithm or indirectly discloses it.

https://support.universaltennis.com.../9000151894-how-is-the-utr-rating-calculated-

Here's a start from a prompt about "inactive players" (rather generic answers that most of you probably already know)

When I emailed UTR my query about how my rating could be lower than every single opponent rating when I won all my matches, the responses were so nonsensical that it felt like I was dealing with a poorly programmed AI chatbot.

 

[travlerajm]

I do not understand what that means. What are "players with past UTR at least 4+ UTR above my level."? Like players that used to be UTR 13 20 years ago but now can barely walk and are therefore UTR 7? What does it have to do with anything what UTR a player was 'in the past'?

DY is an ex UTR 15 tennis player who is now a pickleball player.

 

[jmnk]

It’s a junior level 5 USTA tournament. It’s not on the utr site as of yet.

That’s my point of view. So if you train April and May for the beginning summer cluster of local level 3 and 5 tournaments, pull off a couple upsets, you won’t get credit (or much) in the algorithm.

On top of that, the 6 utr who lost badly to the other 6 utr kid who is beating 8 utr’s will take a big hit to his utr. It supposedly measures matches within range higher.

Can you tell me how much a match is weighted in the algorithm if you beat someone 2 utr above? 1.5? 1.25? 1?

Nobody knows.

People who blindly support utr will say “well maybe the better player had a bad day! He was sick! The lesser player must have had bad results before!” Or if all else fails “it doesn’t matter for 6-8 utr anyway! Overtime it all equals out! It looks close to me!”

Could you give me some clues what level 5 usta tournament it is? So I can see it on USTA site?

 

[tennis4me]

When I emailed UTR my query about how my rating could be lower than every single opponent rating when I won all my matches, the responses were so nonsensical that it felt like I was dealing with a poorly programmed AI chatbot.

In that case, I think the chatbot needs to train the human there.

 

[Tennis2349]

Out of these three players who would you assume has the higher utr or is the better player? Last 10 matches for each. Assuming old matches don’t have as much relevance. All matches within past 3 months. Each player 100 percent accuracy . Separated wins and losses to make it easier to read.

Player a
Wins :
5.58. 12-2
7.3. 12-3
8.03. 13-8
7.6. 8-7
7.04. 8-2
6.5. 8-1


Losses:

9.2. 12-0
7.75. 12-4
8.2. 8-6
6.88. 13-8

Player b
Wins:
5.76 12-1
5.84 8-2
3.67 8-0
6.9!13-10
5.96 4.8
4.7 12-2

Losses:

8.9 12-1
6.7 12.3
5.93 8-6
5.57 12-11

Player c
Wins:
5.78 12-3
4.62 12-3
6.5 11-8

Losses:
7.9 12-7
7.1 12-3
8.4 12-2
7.8 12-3
7.35 12-4
7.04 8-5
5.5 8-5

 

[travlerajm]

Out of these three players who would you assume has the higher utr or is the better player? Last 10 matches for each. Assuming old matches don’t have as much relevance. All matches within past 3 months. Each player 100 percent accuracy . Separated wins and losses to make it easier to read.

Player a
Wins :
5.58. 12-2
7.3. 12-3
8.03. 13-8
7.6. 8-7
7.04. 8-2
6.5. 8-1


Losses:

9.2. 12-0
7.75. 12-4
8.2. 8-6
6.88. 13-8

Player b
Wins:
5.76 12-1
5.84 8-2
3.67 8-0
6.9!13-10
5.96 4.8
4.7 12-2

Losses:

8.9 12-1
6.7 12.3
5.93 8-6
5.57 12-11

Player c
Wins:
5.78 12-3
4.62 12-3
6.5 11-8

Losses:
7.9 12-7
7.1 12-3
8.4 12-2
7.8 12-3
7.35 12-4
7.04 8-5
5.5 8-5

Player A should be a UTR 7.5.

Player B should be a UTR 6.

Player C should be a UTR 6.

 

[Tennis2349]

Player A should be a UTR 7.5.

Player B should be a UTR 6.

Player C should be a UTR 6.

That seems more accurate to me. And there is nothing further back in the past for these players that really stands out.

The entire point of the number is “100 percent accuracy at this point in time”. So it should only look at numbers and disregard everything else.

Player a: 6.70. Same age as player c. Just turned 13. Didn’t chase rankings in 12’s and played locally against older kids.

Player b 7.30. 14.5 years old. Assuming his utr raises naturally for getting older. Losing to players far below doesn’t hurt his utr.

Player c: 6.85. Just moved up from 12’s 4 months ago where he had a higher utr beating up on 10-11 year olds. Still moving up but slowly.

So is it “sort of accurate”? Sure.

But when programs or academies ask your utr they aren’t giving it this much thought and digging into every result. The number is blindly trusted.

Being a half point higher seems like a much stronger player when you are only at 6 or 7 utr. Player c seems slightly better than player a. Unless you look at actual results, instead of how the algorithm works to have a final number.

Player A has highest USTA ranking and highest WTN number. And highest on tennis recruiting

 

[travlerajm]

That seems more accurate to me. And there is nothing further back in the past for these players that really stands out.

The entire point of the number is “100 percent accuracy at this point in time”. So it should only look at numbers and disregard everything else.

Player a: 6.70. Same age as player c. Just turned 13. Didn’t chase rankings in 12’s and played locally against older kids.

Player b 7.30. 14.5 years old. Assuming his utr raises naturally for getting older. Losing to players far below doesn’t hurt his utr.

Player c: 6.85. Just moved up from 12’s 4 months ago where he had a higher utr beating up on 10-11 year olds. Still moving up but slowly.

So is it “sort of accurate”? Sure.

But when programs or academies ask your utr they aren’t giving it this much thought and digging into every result. The number is blindly trusted.

Being a half point higher seems like a much stronger player when you are only at 6 or 7 utr. Player c seems slightly better than player a. Unless you look at actual results, instead of how the algorithm works to have a final number.

Player A has highest USTA ranking and highest WTN number.

this could really use a 60 minutes segment to expose UTR.

 

[Tennis2349]

this could really use a 60 minutes segment to expose UTR.

Exactly. And how long will it take utr to properly sort these 3 players? That player a is probably atleast a half point stronger?

3 months? 6 months? A year? I’ll keep checking on this example and see.

It certainly is not even close to 100 percent accurate now.

 

[Tennis2349]

Atleast USTA, tennis recruiting network, and WTN have the order correct.

 

[FuzzyYellowBalls]

I still say you tow should just email each other or call each other lol

 

[travlerajm]

Atleast USTA, tennis recruiting network, and WTN have the order correct.

And the funny thing is, as you pointed out, UTR would also have the order correct if they just used their own data correctly in the algorithm without outsmarting themselves.

 

[travlerajm]

I still say you tow should just email each other or call each other lol

Sorry if this discussion is jeopardizing your UTR promotion paycheck.

 

[Tennis2349]

And the funny thing is, as you pointed out, UTR would also have the order correct if they just used their own data correctly in the algorithm without outsmarting themselves.

And I am not even mentioning problems when factoring adults vs kids, girls vs boys, or geographic location. These kids are all basically the same age, in the same area, playing the same events.

 

[FuzzyYellowBalls]

Sorry if this discussion is jeopardizing your UTR promotion paycheck.

And I am not even mentioning problems when factoring adults vs kids, girls vs boys, or geographic location. These kids are all basically the same age, in the same area, playing the same events.

I think you guys work more at troubleshooting UTR than the actual UTR employees, might put together a presentation and offer your services as consultants.

 

[travlerajm]

I think you guys work more at troubleshooting UTR than the actual UTR employees,

this is obvious.

 

[Tennis2349]

this is obvious.

Then I think it’s gets into further problems ..

If underrated player A beats another 8, again the algorithm says “it doesn’t count much because of such a big utr difference”

Where in reality player A’s form is 7.5. Not 6.7.

Meanwhile I will bet player B keeps slowly rising no matter how badly he plays. So to have player A surpass player b may take 6 months. To surpass by half a point much longer.

 

[jmnk]

What tournament is that? Can you give me some clues so I can find it on UTR site?

It’s a junior level 5 USTA tournament. It’s not on the utr site as of yet.

Could you give me some clues what level 5 usta tournament it is? So I can see it on USTA site?

Could you provide an actual example of a player that :
Had 100% reliable utr ranking in 6 or more utr range
Then did not play for 6 months,
And then his utr ranking increased by 1 or more points?

@Tennis2349 - so it is safe to assume you will never actually provide any examples that could be checked?

 

[jmnk]

Thank you for providing a perfect example of why a match between players with such UTR disparity should be (and is) disregarded. Exactly because the score is completely meaningless.

But your assumption that it is meaningless is fallacious.

I have played more than 100 practice sets versus players with past UTR at least 4+ UTR above my level. It’s 100x easier to win 4+ games against an “ex” UTR 13 than it is against a current UTR 13.

I do not understand what that means. What are "players with past UTR at least 4+ UTR above my level."? Like players that used to be UTR 13 20 years ago but now can barely walk and are therefore UTR 7? What does it have to do with anything what UTR a player was 'in the past'?

DY is an ex UTR 15 tennis player who is now a pickleball player.

@travlerajm - I think you need to talk to me like I'm 5 years old. I have no idea anymore what you are arguing. Or why are you claiming that Donald Young is now a pickleball player - he may very well be but based on his recent results it does appear that his UTR ranking is more/less correct.

 

[travlerajm]

@travlerajm - I think you need to talk to me like I'm 5 years old. I have no idea anymore what you are arguing. Or why are you claiming that Donald Young is now a pickleball player - he may very well be but based on his recent results it does appear that his UTR ranking is more/less correct.

Full disclosure, I collaborated briefly with DY a little over a year ago, after he reached out to me, when he was still seriously interested in resurrecting his tennis career and was still competing in challenger events. His UTR history from last month’s local tourney does not agree with his rating. Find any other UTR 13.5 and the scores against similarly rated lower level opponents will be much more one-sided.

 

[Tennis2349]

@Tennis2349 - so it is safe to assume you will never actually provide any examples that could be checked?

It took a while to record all of that. I’m not posting kids names on the site. Why would I?

What do you mean “be checked”? I am making it all up?

Are you doubting the validity? Any thoughts on the actual post with the match results? If not skip the thread.

 

[Tennis2349]

In the above utr match plays examples…

In real life, player A has been ahead of B and C for a while. Now more results are in that I posted above..

How long before player a surpasses b and c?

Does A have to continually be 8 utr’s to pass them? What if A has mediocre results like B and C moving forward. And they all have mediocre results. Does player a just stay half point behind?

If the system were accurate in this case, it would be corrected much quicker. There are more than enough results. With player a is atleast equal to player b and moves up, or the other two move down…How can it be 100 percent accurate if it is that far off, and will remain to be even with so many results?

 

[jmnk]

It took a while to record all of that. I’m not posting kids names on the site. Why would I?

What do you mean “be checked”? I am making it all up?

Are you doubting the validity? Any thoughts on the actual post with the match results? If not skip the thread.

you do not need to post any names. You can safely post a link to a public USTA tournament the results of which you are questioning.

 

[Tennis2349]

you do not need to post any names. You can safely post a link to a public USTA tournament the results of which you are questioning.

The last ten matches for the 3 different kids are from several different tournaments.

Again, I already did the work. Any thoughts? I’m not questioning any results… they are posted correctly in the utr site.

 

[Tennis2349]

Why would it move so slowly and incrementally? It’s also harming the kids who lose to player a.

 

[jmnk]

The last ten matches for the 3 different kids are from several different tournaments.

Again, I already did the work. Any thoughts? I’m not questioning any results… they are posted correctly in the utr site.

yes, the thought is that if you provided some clues so one could find the complete players history the current rankings in question could have been likely easily explained. Instead you are posting some results that are not really incorrect but they lack larger context (i.e. complete history over last 30 matches). This is a known technique - trav used to do that all the time when trying to point holes in UTR. Almost in all cases the correctness of the algorithm was easily shown once _all the results_ were known.

 

[Tennis2349]

yes, the thought is that if you provided some clues so one could find the complete players history the current rankings in question could have been likely easily explained. Instead you are posting some results that are not really incorrect but they lack larger context (i.e. complete history over last 30 matches). This is a known technique - trav used to do that all the time when trying to point holes in UTR. Almost in all cases the correctness of the algorithm was easily shown once _all the results_ were known.

I disagree because that is totally irrelevant. Guessing or trying to explain why it is wrong doesn’t make it right. It’s clearly incorrect. I can post the last 60 matches, and no matter what those results were, it’s still wrong today and has been, and will be.

UTR touts itself as 100 percent accurate TODAY.

In my example it probably wasn’t accurate 2 months ago, today, and most likely won’t be 3 months from now. Utr even states that most recent matches have the most weight.

When do you think the algorithm will correct itself amongst these 3 players?

Too add more context:

Best victories.

Player a: 8.03

Player b: 7.5

Player c: 7.2

I already explained why in my
Opinion it is so wrong.

Player b goes up as he is in the system the longest, and player c was beating 10 to 11 year olds. No matter what the reason, wrong is wrong.

If player A lost more games against older kids in the past that had a lower UTR, maybe that would explain the number. But it’s still wrong if the overall objective is to pick out the strongest player.

 

[Tennis2349]

So I asked the utr bot about if you beat a 2 plus utr. 2 different answers.

If the weaker player wins a match against a higher-rated opponent, this result will count towards both players' ratings. For example, if a player with a UTR of 5.20 wins against a player with a UTR of 7.61, this match will be included in the ratings calculation.

When asked again, Then it said:

Even if you win a match against an opponent with a UTR rating difference of more than 2.00, the match may not count towards your UTR rating due to the significant rating gap exceeding the exclusion threshold set by the algorithm.

 

[travlerajm]

So I asked the utr bot about if you beat a 2 plus utr. 2 different answers.

If the weaker player wins a match against a higher-rated opponent, this result will count towards both players' ratings. For example, if a player with a UTR of 5.20 wins against a player with a UTR of 7.61, this match will be included in the ratings calculation.

When asked again, Then it said:

Even if you win a match against an opponent with a UTR rating difference of more than 2.00, the match may not count towards your UTR rating due to the significant rating gap exceeding the exclusion threshold set by the algorithm.

That chatbot responds like the UTR rep that emailed me back with nonsensical conflicting answers.

 

[Tennis2349]

That chatbot responds like the UTR rep that emailed me back with nonsensical conflicting answers.

I asked why it gave me two different answers…

Then it said

I apologize for the confusion in my previous responses. To clarify, matches with a UTR rating difference of more than 2.00 are excluded by the algorithm and do not count towards the players' UTR ratings, regardless of the match outcome.

Which is different than what it says on the utr website. The website clearly says it will count if the lower player wins.

Utr bot also said matches with a difference of 1 point do not count as much either lol. What a joke

 

[Tennis2349]

As a parent who has watched this number for years, something has changed. I feel it used to be more accurate. Even USTA matches that used to be inputted the next day now might not be at all or days later. Sometimes tournaments are entered twice.

The problem UTR has is that no juniors want to play their events. People want USTA POINTS, and the USTA has a competing system. Some USTA events seed players by WtN, which seems more accurate. I used to feel the opposite.

When kids are a 1-3 utr the utr events might be chosen. But nobody that is serious is wasting time at UTR events. There is no comparison between a day of utr matches vs a level 5 or 3 tournament.


I can’t really trust it when the site has totally different answers than the staff or chat bot gives you.

Saying it disregards completely a win over someone 2 utr ahead is absolutely ridiculous. As is saying 1 point ahead is given less weight.

Maybe it needs to do that to pretend the algorithm is still “accurate”.

Sort of “yes you moved up magically without playing and are a 7 utr. So when you go play and lose to 5 and 6 utr we won’t count it! See you really are a 7! And that 5 really is a 5! If you tie another inflated 7 we will count that!”

 

[Tennis2349]

What would you make of this?

I put all 3 players WTN into this formula and these were the results.

Player a: 6.91. (6.7 actual)

Player b:6.47 utr (7.3 actual)

Player c: 6.31 utr ( 6.85 actual)


UTR = 15.3567 – 0.3539 * WTN

So the WTN conversion formula gets the order correct of strength between the players. (I know this as they actually play together as well as the results) Yet the algorithm has b and c with an inflated utr , and player a with a lower utr.

 

[Tennis2349]

The utility of utr is to find a competitive match between two strangers. Nothing more.

If it excludes wins over 2 plus utr, and even lessens wins over 1 plus utr , it obviously can’t be accurate. The utr website itself is embarrassed to even admit this.

It’s accurate to the point where if you meet a stranger that is a 10, and you are a 3, you know it won’t be worthwhile playing with them.

The problem lies within how it is used and blindly trusted to be accurate. People think it means if you are a 6, and the other guy is a 6.7 it means he should win. Nothing could be further from the truth. Having an extra decimal point is even more comical. Maybe that tricks people into thinking it is “that accurate”.

And it makes coaches lives easier by deciding on which court and which group kids should play in. They all say UTR is garbage, but order their programs by UTR so parents don’t bug them.

 

[FuzzyYellowBalls]

I know of only one other poster in TTF who likes to talk to themselves on posts, has anyone ever seen @travlerajm and @Tennis2349 in the same room together? The similarities are increasing.