Aug 11 2013

Open Source the NTRP

History | League Tennis | NTRP | Rules | Tencap Tennis | USTA       Clif Render      

As an active adult who loves playing sports, I am continually amazed at how well the USTA Adult (Age 18 and over) League System works. The USTA organizes and manages league play for (at last count) over 800,000 adults around the country. That's a pretty big endeavor. The fact that the system works at all is amazing! It really is a testament to the hard work of thousands of individual volunteers, pros, coaches and staff members around the country. If USTA League Tennis was only a localized endeavor, manpower alone would be enough. But, since the sport also involves District, Sectional, and National levels of play, a little electronic assistance is an absolute necessity.

Enter the NTRP. The NTRP (The National Tennis Rating Program) is the rating system that the USTA devised to group league players into common ability levels in order to make league play more fair and competitive. It is this system that is the backbone of what makes USTA League play work nationwide. Think about it. Without a common rating system, Players from different areas of the country wouldn't be able to play each other on even footing. It would be like tournaments only having an "Open" Division rather than subdividing players into common levels of play. How many of us would even consider playing if we knew that we had no chance of winning? But it's not just about winning, of course. It's also about seeing improvement - watching your rating go from 3.0 to 3.5 and beyond. How would you be able to quantifiably measure improvement if we didn't have the NTRP? What would P.J. Simmons call his blog? I think not.

No, the NTRP (or, at least, some automated ratings tracking system) is an absolute necessity. Of course, the NTRP was developed way back in 1979 and hasn't changed a whole lot since then so many would argue that it isn't the best option out there anymore . Some would say that the Tencap Tennis Rating System is a better option, and they might be right. That, though, is a thought for another day. For now, it's enough to say that the NTRP Rating System is the system that we have and it does it's job.

In case you don't know, the NTRP, in general, breaks people up into numerical categories categories ranging from 1.5 for "never heard of tennis or played any sport" beginners to 7.0 world-class professionals going in .5 point increments. For all intents and purposes, the main ratings that you see in league play are from from 2.5 through 4.5. The full breakdown of what constitutes each rating can be found on the USTA's About the NTRP website.

Early on in the life of the NTRP, a starting rating would be given to a player by a teaching pro or some other certified USTA official. Certified officials would also watch players during league and championship play and players that appeared to be rated too low were "bumped up" to a higher rating. This was in many ways, a very subjective system prone to human error and bias. Also, this type of system doesn't account for a player just having a really great day. Every now and then, I get into the zone and just have a fantastic day. Whether I am actually playing better than my opponents on those days or whether my opponents just keep hitting everything into my wheel house, I really don't know. All I do know is that everything just works. It's a rare and beautiful thing. Days like that are the wrong days to have a USTA observer nearby. Needless to say, this system left lots to be desired.

In 2000, an experimental system was approved for development that utilized an automatic computer based ratings system called "Dynamic NTRP". By 2003, the entire country was on the new system. This new system was a huge improvement in many ways. Now your rating would be dynamically updated by computer based on your match play. The system would not be a purely subjective one any longer. Your rating was now a much more accurate measure of the results of your actual game play over time. Another major change in the process was that you no longer needed a teaching pro or USTA Official to give you your initial rating. A new "self-rate" process was introduced that allowed you to determine your own rating based upon a set of questions and answers. This process is still in use today. Sample questions are things like, "Were you competitive in High School in any sport," "Have you ever played in a recreational Tennis league," and so on. This system allowed for more people to get on the court than ever before.

And such is the current state of things. The Dynamic NTRP system is still up and running. League Tennis has become more and more competitive over the years with more players involved than ever before. As participation numbers grow steadily toward the one million mark, many people have begun to complain about the current system. They complain that the system allows people to lie during the self-rate process so that they can dominate at the lower levels and stand a better chance of winning and making it to Nationals. They complain that people are able to game the system by throwing matches that don't matter in order to keep themselves playing below their true skill level. The complaint that I hear most often, however, is that the system just doesn't make sense. One player of moderate ability with only a decent record will get bumped up while another player with exceptional skills and a great record will remain at their old level. I hear this complaint nearly every season, and I have to admit, I do wonder at times if there's not some truth to it.

Here's the thing, though. The USTA does listen to these kinds of complaints and has done a number of things to try to make the system fairer. For one thing, the USTA is very careful to hide player's true dynamic rating number in order to keep people from being able to know which matches they need to lose in order to keep their ratings low. The rating that you see is only accurate to the nearest .5 points while the actual system goes down to the hundredths place and really is a closely guarded secret. Additionally, they have created new age limited leagues so that older players won't have to play against younger players if they don't want to. Introduced in 2012 were "40 and over" and "55 and over" leagues in place of "50 and over" and "60 and over" leagues. I understand the intention with lowering the adding age group restrictions, but I'm not sure that it makes good sense in the long run. Why not subdivide the ratings a bit more? Increase it to a 100 point system as Tencap does? I really don't know. Maybe there are really great reasons to do it this way, but I don't see them. This, however, isn't the biggest problem with the current system.

The biggest problem with the current system is its mystery. The current Dynamic NTRP system is something of a black box that very few people seem to understand. If you're familiar with the term "Open Source," then you'll understand immediately when I tell you that this thing is 100% "Closed Source". Well, maybe that's not actually true. We do know a few things about the system. For one, your rating does not change based on whether or not you win or lose. Rather, its changes are based on how much you won or lost by and to whom. A rating example provided by the USTA explains, "A typical match result for a player, for example, with a 3.01 rating versus a 3.49 player, both of whom are 3.5s, would be 6-0, 6-0 in favor of the higher rated player." If the score of the match between these two players is anything other than what is expected, the lower ranked player's rating increases while the higher ranked player's rank decreases. The amount of the increase or decrease is a secret, of course, but we can assume that it is some function of the difference between the ratings of the two players. In the case of a doubles pair, the amount of the increase or decrease is divided proportionally between the two of them based on the difference between the pair's ratings at the start of the match.

While I would love to see the NTRP system go "Open Source", there is no doubt that there would be many individuals who would try to game the system. Those types of individuals will work to increase their chances of winning by any (and all) means necessary. So, which is better: A mystery that some people try to abuse and other people complain about or a known formula that some people abuse and other people complain about? I think that I'd prefer to see the USTA error on the side of transparency, myself.

Increasing the number of rating categories would be helpful, too. Currently 85% of all USTA League players are crammed into the 3.0, 3.5, and 4.0 levels. You can see this data graphically represented on Kevin Schmidt's Computer Ratings blog. Kevin's blog also has lots of other good insights on NTRP distributions by section, comparisons between sections, and general wisdom regarding the NTRP rating system. If you're looking for a detailed (but certainly unofficial) listing of players and their estimated ranking to the hundredth's place in your area, check out the information on

In my opinion, the USTA really needs to figure out how to produce a better distribution of players across at least the most densely populated levels. I won't go so far as to say that the current system is broken, but I will say that it is dangerously close to being so. My warning to the USTA: Tweak now or forever hold your peace.

Comments (77) -

Kim Kim says:

This is a very interesting look at an issue that a lot of people don't care to understand.  I find it funny that players are willing to throw a match just to maintain a lower NTRP rating since the whole point of NTRP is to come up with a system where players can play competitive matches against opponents who are at their level.  Here in Houston, the whole self-rating thing is pretty ridiculous - totally skewed.  But, as you point out, its currently the best system we have and its better than nothing.

Clif Render Clif Render says:

Isn't it amazing! The mind of the ultra-competitive sees things very differently from the way we do. When winning is the primary goal, it only makes sense to use absolutely every advantage at your disposal to make it happen. If that includes taking advantage of flaws in the system to keep yourself in a position where you can win more often then you take advantage of flaws in the system in order to be able to win more often. I certainly follow the logic and I can't argue with the fact that it does work but It would take a lot of the fun out of the game for me. Oh well. Guess I'm not going pro, after all...

Kevin Kevin says:

Nice summary of how the system works.

While it isn't open source, I have effectively reverse engineered the system and IMHO understand it quite well.  See my blog at for my estimated ratings and other thoughts about league tennis and the Dynamic NTRP system.

Clif Render Clif Render says:

Very Impressive stuff, Kevin! I've tweaked my post to include some of the information from your site as well as some links to direct folks to the source. Thanks for sharing!

Add comment

  • Comment
  • Preview