# Dynamic NTRP ratings

Discussion in 'Adult League & Tournament Talk' started by Orange, Apr 10, 2012.

1. ### OrangeRookie

Joined:
Sep 14, 2011
Messages:
209
I have read all the official USTA statements on the USTA website, but still have questions, and would appreciate any help you can give me!

1. When two self-rated players play each other in the first match for each, how is the dynamic rating calculated? For example, if two players have self-rated at 2.5, is the initial "current dynamic rating" for each player considered to be 2.50? If the score is quite lopsided, does the dynamic rating of the winner move much higher, and the dynamic rating of the loser move below 2.50?

2. Do two self-rated doubles partners at 2.5 likewise start out at 2.50 each for the purposes of dynamic ratings?

3. Does the fact that the USTA weights the four most recent matches most heavily mean that self-rated players essentially get four "free" (non-strikeable) matches?

4. If a player becomes a benchmark player due to losing in the playoffs of a 3.0 singles league and is moved down to 2.5 as a result, is her dynamic rating probably along the lines of 2.95 to start the year?

5. If the 2.5 benchmark player wants to move back up to 3.0 in the ratings the following year, what is the best way to accomplish this? So far, she has a 2.5 singles win, a 3.0 singles win, and a 2.5 singles loss (in a competitive match against an undefeated self-rated player).

Thanks in advance for any insight you can give me!

2. ### J_R_BHall of Fame

Joined:
Jan 27, 2010
Messages:
3,086
Location:
Newtown, PA
1. No. If no one in a match has a dynamic rating yet, then no dynamic ratings are generated for that match.

2. No. They get a dynamic rating from their first match against a dynamically rated team. They will get the same dynamic rating for that match if both are S-rated with no dynamic rating prior to that match.

3. No. Every time a player plays against someone with a dynamic rating, s/he gets a dynamic rating for that match. If the dynamic rating for that match is above the strike threshold, it is a strike. For the first some number (2 or 3) of matches, the ratings are generated individually (as in, just the performance in that match). After that, the rating is an average of the last some number (4 you said?) of matches.

4. No. A 2.95 dynamic rating is the top of the 3.0 range (3.01 starts the 3.5 rating). If a player is bumped down, you can guess that his/her rating is usually near the top of the lower level (~2.45 in this case), but that is not necessarily true. It is what the formula says it is.

5. Play up is the easiest way to get bumped up. Keep playing those 3.0 matches and she'll be back at 3.0 next year.

3. ### OrangeRookie

Joined:
Sep 14, 2011
Messages:
209
Thank you very much for your quick and detailed response! Thanks also for explaining the range for 2.5 and 3.0 ratings, which I had seen but forgotten, and thanks for answering the essence of my question even though I had the numbers wrong.

A. If one doubles player has a dynamic rating and the other three in the match do not, is a dynamic rating generated for the match? (I would guess not).
B. Is the information on dynamic ratings you provided published somewhere (so I could read it and therefore understand it better), or must one just learn from experience (or from asking questions on Talk Tennis)?

4. ### J_R_BHall of Fame

Joined:
Jan 27, 2010
Messages:
3,086
Location:
Newtown, PA
A. I would guess not as well, but I don't know that for sure.

B. I read all of this is online somewhere, but it's not all in one place (except here, LOL). Try searching on the USTA site or just Googling whatever you are interested in.

5. ### Tennis PhanaticNew User

Joined:
Apr 21, 2009
Messages:
11
Another thing to keep in mind is everyone starts with a rating of "0.0". It doesn't matter what you self-rate. Your self-rating has no impact on your DNTRP, it only establishes what level(s) you can play.

6. ### OrangeRookie

Joined:
Sep 14, 2011
Messages:
209
That's another useful piece of information. One team in my 2.5 league has several self-rated 2.0 players. That rating is essentially meaningless, then, in a league where the lowest possible team is a 2.5 team.

This information serves to explain why the top teams in our leagues have self-rated undefeated singles players who've won by huge margins. As 80% of the players in our league are self-rated, it is possible to play a long time without ever getting a dynamic rating, as long as one happens to play only players who themselves have no dynamic rating. As a result, those undefeated players can continue to win by huge margins without being disqualified.

At least two players on my team have no dynamic rating after playing three matches.

Joined:
Feb 24, 2010
Messages:
52
Location:
NJ
8. ### giseppiNew User

Joined:
Jul 18, 2010
Messages:
82
Is there a way to find out what your dynamic rating is?

9. ### dizzlmcwizzlHall of Fame

Joined:
Mar 24, 2010
Messages:
2,200
Location:
DE
no ... not unless you are in tight with a section coordinator

10. ### J_R_BHall of Fame

Joined:
Jan 27, 2010
Messages:
3,086
Location:
Newtown, PA
That is correct. However, the unrated S-rates have to keep playing themselves in an insulated group. As soon as one plays a C or B rated player, then he/she will have a dynamic rating, and whomever that person plays after that will also. Basically, once the dynamic ratings get seeded, they spread quickly, but the computer does need SOMETHING to work off of...

11. ### OrangeRookie

Joined:
Sep 14, 2011
Messages:
209
This all makes more sense with your explanations! The USTA website doesn't address how the system works for self-rated players who are playing their first several matches.

The USTA NTRP FAQs state:
I. Until today, my partner didn't have a dynamic rating. If she started at 0.0, as one poster said she would, and my dynamic rating was at 2.4 (hypothetically; of course I don't know), will a 2.4 differential really be maintained?

II. If we had won in a blowout, could that have pushed me above 2.5 because of the differential being maintained? That is, she would then have a win, so she surely couldn't really be 0.0; if she's considered at least a 1.0 after one win, wouldn't maintaining the differential put me at 3.4 for that match (a ridiculous result)?

III. If we had lost in a blowout, could she have a dynamic rating of 0.0 after that match?

I am truly trying to understand the process, so I continue to appreciate any information people can provide!

12. ### J_R_BHall of Fame

Joined:
Jan 27, 2010
Messages:
3,086
Location:
Newtown, PA
No, it's not really 0.0, it's blank (or null or whatever) before you get a rating. I think in that situation, your team gets a rating based on your rating vs the opponents and then your partner gets assigned that rating as well, but I'm not 100% sure about that. The "maintain the differential" part of the calculation applies to the situation where both players already have a dynamic rating.

13. ### AELTCNew User

Joined:
Feb 24, 2010
Messages:
52
Location:
NJ
Above all, remember that by design no one will ever understand the DNTRP system. The USTA intentionally withholds information about how DNTRPs are calculated.

I. I don't think a differential is maintained against someone who has no rating; that doesn't seem sensible. The system will probably compute what the new partner's DNTRP must have been to produce that match result, and start them off there.

II. No, see part I.

III. No, a result of 0-6, 0-6 is not used to compute DNTRP. That result doesn't provide information about how much better your opponents are. For example, if a 3.01 gets bageled by a player, all you know is that the other player is much better. He or she could be rated 5.0, or 3.6ish.

If the result is 6-0, 6-1, I think the system does have enough info to perform a computation, and I don't think a player would stay at 0.0. So basically, 0.0 is just a way to describe a self-rate with no match experience against rated players.

14. ### Local GirlNew User

Joined:
Jan 27, 2012
Messages:
36
This is from the USTA website http://www.usta.com/Play-Tennis/USTA-League/Information/Dynamic_ratings/

How Dynamic Ratings are calculated for a specific match
In matches where all players have previous ratings the procedure is as follows:

1. The system looks up the current dynamic rating of all the players in the match.

2. The system looks up from a table, the likely score of the match based on the current dynamics of the players.

3. The system compares the likely match score with the actual match score. For example, if one player or team has a tenth of a point higher rating than the opponent, the likely score is 6-4, 6-4.

•If the winning team wins by a larger than expected margin, each player’s ratings is increased based on the margin of victory and the losing player’s rating is decreased by the same amount.
•If the winning team wins by less than the expected margin, their ratings will actually decrease and the losing team’s ratings will increase.
•Likewise, the “wrong” team may win which causes their rating to increase markedly and the rating of the team which was favored would decrease by the same amount.

4. The rating obtained for each player in Step #3 is averaged with a maximum of their previous three dynamic ratings and that number becomes their new current dynamic rating. (Indirectly this connects the current dynamic to all previous matches but weights the four most recent matches more heavily.) The reason for this averaging is to even out the ratings in cases where some unusual situation causes an atypical result.

Each player rating is maintained in the system to the nearest hundredth of a point.

The difference in ratings of the members of a doubles team is held constant in a calculation of an individual match. If the two players are three hundredths (.03) of a point apart going into the match then they are three hundredths (.03) apart after the calculation in Step #4. However, once that number is averaged with the three previous dynamic ratings (Step #5) that difference may change. This is how we measure the performance of players as they change partners.

15. ### OrangeRookie

Joined:
Sep 14, 2011
Messages:
209
AELTC wrote in post #13 that a result of 0-6, 0-6 is not used to compute DNTRP.

ABTennis wrote on this forum in April 2011:
The only way I can reconcile these two statements is by concluding either that 6-0, 6-0 losses aren't counted but 6-0, 6-0 wins are, or that 6-0, 6-0 wins can result in strikes but do not otherwise affect DNTRP. Does anybody know which is true?

Joined:
Jan 27, 2012
Messages:
36
17. ### Alchemy-ZHall of Fame

Joined:
Dec 14, 2010
Messages:
1,776
Location:
Augusta, GA
From what I can tell the computer algorithm factors in Irony for each player.

Any player playing up...hitting the courts weekly...getting coaches....trying to improve (don't get bumped)

Any player that has a 50/50 record that only shows up to play matches and afterwards has a few beers and likes the level of play they are at (get's bumped)

18. ### gmatheisHall of Fame

Joined:
Aug 10, 2010
Messages:
1,631
DNRTP and DQ are different.

I don't believe anyone said that 6-0,6-0 will not be used to generate a strike, all they said was it will not be used to give you a dynamic rating.

If you are a self rated 3.5 and you play a computer rated 3.5 who's dynamic rating is 3.41 and you beat him 6-0,6-0 you would generate a strike because even a 3.50 should not beat a 3.41 6-0,6-0. BUT it will not generate any dynamic rating for you because it doesnt know if you are a 4.0 or a 6.0 it just knows you are more than a 3.5

19. ### schmkeHall of Fame

Joined:
Jan 20, 2010
Messages:
2,094
I don't know that it is accurate to say no one will ever understand it. I think many of us understand it, and I've even implemented my understanding to compute an "Estimated DNTRP" that has proven fairly accurate, we just may not have some of the specifics that allow an exact reproduction of the DNTRP.

One area I'd be interesting in knowing more specifics if anyone knows is around self rates and how quickly they get a DNTRP. I know and have implemented what is described in this thread, that a player has no rating until they play an opponent or opponents that have DNTRP and then they get an initial match rating computed.

My questions then are:

- Does this first match rating become their DNTRP immediately? Or must more than 1 match be played? I've seen mentions of there needing to be 2 or 3 matches played first but haven't seen an authoritative answer. And note that this impacts how quickly the DNTRP spreads amongst a group of self rated players once one plays a rated opponent.

- Assuming several matches need to be played before a first DNTRP is calculated, I assume the initial DNTRP is calculated from the (2 or 3) match ratings just by averaging those together. Is this accurate? If so, is the resulting DNTRP then the only one in the history at that point, such that the next DNTRP would be the average of the next match rating and just this one DNTRP?

The above reflects what I think is likely the case, although I can see a few other reasonable approaches, but any information or confirmation is welcome.

20. ### OrangeRookie

Joined:
Sep 14, 2011
Messages:
209
I would also like to know the answer to schmke's questions!

If the first match rating becomes the player's CNTRP immediately, a good player could receive a strike by beating by a wide margin someone who had beaten only one other player who had a C or B rating.

The player with a C rating could have declined in ability, having gotten her C rating based on playing in 2010 (as I did) and could have been injured or not played any tennis in the interim.

Supporting the idea that the first match rating doesn't become the player's DNTRP immediately is the existence of a self-rated player in my area who was 7-0 for the season, winning four of the matches 12-0, 12-2, 12-3 and 12-4 (and splitting sets to win in the others). It would seem that she would have received at least three strikes but for the fact that her opponents didn't yet have dynamic ratings.

21. ### schmkeHall of Fame

Joined:
Jan 20, 2010
Messages:
2,094
I'm pretty sure the first rating doesn't result in a Dynamic NTRP, and instead it takes 3 matches before that happens. And to count as one of those 3, the player has to be the only self rated player in the match, or both opponents must have a Dynamic NTRP.

22. ### OrangeRookie

Joined:
Sep 14, 2011
Messages:
209
I haven't been able to find any definitive information stating that a player must play 3 matches against players with dynamic NTRP ratings in order to get a dynamic rating herself; that is contrary to what J_R_B posted in post #2, above.

I have searched the web, including the USTA website, for definitive answers, and cannot find them. In fact, the most extensive explanations of the system have come from Talk Tennis.

23. ### schmkeHall of Fame

Joined:
Jan 20, 2010
Messages:
2,094
I don't think what I wrote is in conflict with what J_R_B posted. I think we agree that some initial number of valid (against players with dynamic ratings) matches (I believe 3) computes just an individual/match rating. It is only after this number of valid matches have been played that an actual/first dynamic rating is calculated by averaging them, and then the process goes on averaging the most recent match rating with the (up to) 3 prior dynamic ratings.

I believe that the initial (3) match ratings don't constitute a dynamic rating as I've seen situations where a self rated player plays 1 or 2 matches but still goes into the following year as a self rated player. Now, this really means they didn't have enough matches to get a C rating, not necessarily that they never had a dynamic rating that would be used in computing opponents match ratings, but I think that is likely the case.

24. ### SweetH2ORookie

Joined:
May 11, 2010
Messages:
378
Location:
Powder Springs, GA
http://www.atlanta.usta.com/Dynamic_NTRP/Home/
"Dynamic ratings require that all returning league players are entered into the computer with their most current hundredth-of-a-point computer rating at the beginning of the league year. New players to the program will generate a rating beginning with the first match in which they compete against a player who has a computer rating. Visual rating assignment will no longer be an integral part of the rating system."

...and from
http://www.atlanta.usta.com/Dynamic_NTRP/Dynamic_NTRPFAQ/
"When calculating dynamic ratings, does the computer treat doubles partners differently?

Whatever mathematical difference that existed between partners at the beginning of a match is maintained at the end of the match. For example if partners’ ratings were two-tenths of a point apart at the beginning of a match (3.3 and 3.1), then they will remain two-tenths of a point apart at the end of the match (3.5 and 3.3)."

So doubles partners, one with a dynamic rating and one without, could not be used to establish a dynamic rating for their opponents. Essentially, if one of the partners does not have a dynamic rating, the partnership would not be considered to have a dynamic rating either. But that's my interpretation of the above.

25. ### OrangeRookie

Joined:
Sep 14, 2011
Messages:
209
Thanks for this information! I'm trying to piece together the rules, so every clue helps.

I agree that each opponent in a doubles match must have a dynamic rating in order to generate a dynamic rating for the team playing them. I think that part is clear.

What is unclear to me is whether a player has a dynamic rating that can result in a strike against another player after playing only one match. In my example of the 7-0 player, above, all her matches were singles matches. It seems that she would have been disqualified if her opponents had gotten a dynamic rating after playing only one other person with a dynamic rating.

This quote:
can be read one of two ways. Either it means that the first match against a player with a dynamic rating gives this player a dynamic rating, OR it means that the first match against such a player begins the process of generating a dynamic rating, as opposed to the former process that started with a pro's evaluation.

26. ### J_R_BHall of Fame

Joined:
Jan 27, 2010
Messages:
3,086
Location:
Newtown, PA
You're confusing sseveral different usages of the "dynamic rating". Every match has a rating plus every player with at least one match against another rated player has a dynamic rating, and then, if you have 3 matches during the year, you get a dynamic rating at the end of the year, too.

For the first three matches, the dynamic rating for the match is stand alone and not an average, but the player's DNTRP is still an average of the individual match ratings. After you have 3 matches, then this distinction disappears and both the rating for the match and the overall player rating are averages. The difference in usage is that the match ratings are used to determine strikes. The player rating is used to evaluate a player's performace against expectations, given the differential to the opponent's rating.

27. ### Chelsie1Rookie

Joined:
Sep 27, 2011
Messages:
115
Estimated DNTRP

You may obtain an estimation of your DNTRP by visiting:

computerratings@techrunning.com

28. ### beernutzHall of Fame

Joined:
Aug 16, 2005
Messages:
4,577
Location:
Make TT better, use the Ignore List!

Joined:
Jan 20, 2010
Messages:
2,094
30. ### Angle QueenProfessional

Joined:
Jul 27, 2010
Messages:
904
Location:
On the deuce side, looking to come in
Gosh, schmke, I've taken a look at your blog and it's impressive. There are lots of geeks on this forum (my female self included) but I'm super impressed. Parts of me want to ask you where I stand on USTA's ladder, rest of me says...eh...let it ride. It is what it is....and it's just a number. I've been playing long enough in the same area to know where I stand amongst The Players.

But, seriously, I do wish USTA would clarify some of the stuff floating around (esp as it relates to self-rates).

Wish I knew, or cared enough, about the pigskin game to really get into the rest of your stuff. I'll casually watch the Pros, cheer SEC fare...and have season tickets to our local FCS contender (and 2008 National Champions!!). If I ever got into a Fantasy League...I'd be in trouble.