Tuesday, October 29, 2013

Interesting USTA League Stats - Histogram of NTRP ratings down to the tenth

Next in my series of posts on interesting stats from the data I collect for the Estimated Dynamic NTRP Rating reports I generate, I took a look at how many players are at each rating level down to the tenth.

Why down to the tenth?  There was a discussion on Talk Tennis about the issues with the new "plus" leagues where it was hypothesized that there were clusters of players in certain ranges and how, naturally or otherwise, a slight movement of the range for each NTRP level might result in better alignment with those groups of players.  With my ratings, we should be able to chart the ratings rounded to the tenth and clearly see if and where there are any clusters.

Here is the chart:
From this, it doesn't appear there are any abnormal peaks or valleys that would indicate such clusters exist.  There is a peak at 3.2 a slight dip at 3.4, then another peak at 3.5, but this could potentially be explained by the fact that this chart includes both men and women, and perhaps women have their peak at 3.2 and men at 3.5 and when put together, you get the chart above.

So we don't see any odd clusters.  It could be that my ratings aren't accurate and so don't show them, but given my accuracy in predicting DQs and bumps, I think it is more likely that my ratings are pretty accurate and there aren't any significant clusters of players around a particular rating other than the normal large group of players at the NTRP 3.5 and 4.0 levels.

Sunday, October 27, 2013

A team report customer wins USTA League Nationals

I had a number of teams that purchased reports from me to aid them in scouting opponents as well as their own teams, several advanced in the playoffs, a few made it to Nationals, and today one of them won it all.  They've made nice comments about how much the reports helped them scout and plan, and I appreciate them, but the players still had to go out and win the matches and that is what they did, so kudos to the whole team!

Friday, October 25, 2013

The last 2013 USTA League Nationals are going on now

Ok, it isn't really the last, as there are some mixed and perhaps other leagues that have their Nationals in November, but for the 18 & over and 40 & over leagues, this is the last weekend of Nationals in what has been an October (and bit of September) packed with play.  With the introduction of the 40 & over league this year, there was a bunch more play.

In Tucson this weekend is the 18 & over 3.5 Nationals.  I have a team I've helped with reports playing there so will be following it closely.

Then in Indian Wells is the 40 & over 4.5+ Nationals.  I also have a team I've helped playing there so will follow as well.

Stay tuned for updates as I'll likely take a look at how my ratings do prediction matches.

Saturday, October 19, 2013

Another Estimated Dynamic NTRP Rating Report Customer Reaches USTA League Nationals Semis

Several more USTA League Nationals are taking place this weekend and I was pleased to see another one of my customers having made the semis.  Good luck guys!

Friday, October 18, 2013

Now available, Mixed Exclusive Estimated Dynamic NTRP Rating Reports for USTA League

Most people that play USTA League plan in a men's/women's league, but many also play in a mixed doubles league.   There are some that only play mixed though, and for these, the NTRP ratings are calculated a little differently.

If you play in at least three men's/women's matches, your rating will be calculated based on only those and your mixed results aren't factored in.  However, if you only play mixed, you will be given a Mixed Exclusive (M) rating.

A mixed exclusive rating is used only for mixed leagues, and if you have one and decide to play a men's/women's league, you have to self-rate for that league.

So, these M ratings are a bit unique as they would seemingly have to be calculated for everyone who plays mixed in case they don't play men's/women's, but then they aren't used.  In theory, a player could be a 4.0C (men's/women's rating) but a 3.5M based on their mixed matches if they hadn't  played men's/women's.

I've toyed with calculating M ratings for awhile now, but believe I have something working reasonably well now, so if you only play mixed, or play both mixed and men's/women's and just want to know what your M rating would be, I can now calculate these as part of my Estimated Dynamic NTRP Rating Reports.  If you are interested in a report, contact me.

Wednesday, October 16, 2013

When will USTA 2013 year-end NTRP ratings be released?

Most of the National Championships for USTA League tennis are being played this month, and with their completion, year-end ratings can be calculated.  Here is my understanding of the timeline.

  • September / October - Adult 18 & over and 40 & over Nationals take place.
  • End of October - The "USTA year" comes to a close and with that, the last matches to be included in year-end calculations have been played.
  • November - Year-end calculations are performed including benchmark calculations taking into account the Nationals that were just completed.
  • End of November / Early December - 2013 year-end ratings are published and available on TennisLink.
The exact "USTA year" may vary a bit and depend on when exactly the last Adult Nationals finish and when the last weekend of October or first weekend of November falls.  And the year-end ratings are usually released on the Monday after Thanksgiving, but I don't know that there is a guarantee of that.

Many players aren't going to Nationals and aren't playing in a fall league that counts towards their rating, so all the matches that count for them have been played.  If you are one of these and are anxious to know what your year-end rating may be, or just want to know more accurately where your rating falls within your level, an Estimated Dynamic NTRP Rating Report is a great way to find out.  The report will show you your rating and how it changed, match by match, and I'll also comment on the likelihood that you get bumped up or down.  Contact me if you are interested.

Update: I've confirmed the last matches for the 2013 league year were played on 10/27.  Click here for more information.

A report customer wins a National Championship

I did several reports for a player in my section earlier this year, and was pleased to find out this morning that her team won Nationals last weekend.  Congratulations!

Monday, October 14, 2013

Another report customer advances to USTA Nationals, this one a Mixed team

I've had the opportunity to help a number of teams make their way to Nationals by providing reports on team members and opponents.  I'm not kidding myself into thinking they get there because of the reports, they still have to go out and win the matches on the courts, but feedback has been consistently positive indicating that the reports helped them with line-ups.

I'm pleased to report that another team I did reports for has qualified for Nationals, this one an 18 & over mixed team.  Yes, that is right, I am now doing reports for mixed (more on that later).

Something that was interesting about looking at mixed was comparing their normal dynamic rating with what their mixed exclusive rating would be.  For many, they were remarkably similar, but there were a few where they were quite different, and it was good to get feedback that where the mixed rating was higher, that player was actually a better mixed player than not.  So I must be doing something right as the ratings seemed to reflect reality.

This team had to navigate 4 matches in Districts, and then a semi and final to win their Sectionals.  They indicated that the reports were "... incredibly helpful in setting lineups and understanding what our chances were."

As you might imagine, I'm now working on reports for their competition at Nationals.

Good luck team!

Friday, October 11, 2013

More Interesting USTA League Stats - Looking at gender and NTRP level

I received feedback on the last interesting stats blog post showing some interest in seeing a chart that combines gender and level.  For those that wanted that, here it is!

This gets a lot more complicated looking, so some explanation is in order.  And you can click on the image to see a larger rendering which may be easier to read.

Each pair of columns are for men and women in the indicated section.  Each column is showing the percentage of men/women at the level indicated by color.  The total height of the column is the percentage of players within that section for that gender.  Make sense?

I'm sure lots of observations can be made.  Feel free to leave a comment with yours.

More Interesting USTA League Stats - Do more women or men play USTA League?

Next in my series of posts on interesting stats from the data I collect for the Estimated Dynamic NTRP Rating reports I generate, I took a look at how many men and women play USTA League by section.

Here is the chart:

Perhaps surprisingly, it varies quite a bit from section to section.  Hawaii has the most men at 62% while New England has the most women at 69%.  Northern Cal and the Caribbean have high percentages of men too while the Southern section has a large percentage of women.

Is there an "equal opportunity" section that is 50/50?  It appears not, with the Pacific Northwest being the closest at 49% men and 51% women and Texas next at 48%/52%.

Now, I don't have absolutely every USTA match played in 2013 in my database, but I have a large enough sample (thousands of players in each section) that they are representative.

Thursday, October 10, 2013

Estimated Dynamic NTRP Ratings for Northern Virginia (NOVA) Now Posted

I've had a lot of report customers in Northern Virginia so I decided to post a list of my Estimated Dynamic NTRP Ratings through July 1st of this year for that area.  Check it out here.

Wednesday, October 9, 2013

It is time for USTA Nationals, and all the sandbagging accusations that come with it. Here is a solution.

It is October, and for those that follow USTA League you know that Nationals are taking place for the 18 & over and 40 & over men's and women's leagues.

A few of the levels have finished, and as is usually the case, particularly at the lower levels, the observations and accusations about players that appear to have skills well above the level they are playing at and are sandbagging are flying.

The scenario is this.  Players that haven't played USTA League before self-rate according to documented guidelines.  Unscrupulous captains and players may deliberately rate a level or two too low in order to have a competitive advantage.  There are provisions in place to automatically bump a player up a level if their results indicate they self-rated too low, but at the lower levels, there is a large buffer to allow a player to improve without getting disqualified at a given level.

The result is that you find several to many of these players at Nationals, someone that looks like a 3.5 or even perhaps a 4.0, playing at 3.0 for example.  One could argue that to get to Nationals and do well, you actually need to have several of these players on your roster and there is some truth to that.  If you are the best from your area, district, and section, you probably are playing at the very top of it not above the level in question.

Now, one might argue, and many have, that an easy solution to this is to not allow self-rated players to play in the playoffs or at least not allow them to play at Nationals or Sectionals.  There is some merit to this, but it would only be a partial solution as it only delays the problem and doesn't address a segment of the players at Nationals.

Go back to the scenario above.  If our player is really a 3.0 and self-rates at 2.5, they likely get bumped up to 3.0 at year end and have a 3.0C rating.  Now, in playing a year, they probably improved some and are able to play at a 3.5 level.  But they are a 3.0C so they sign up for a 3.0 league the next year, and with a little coordination join a team that has a few other players in their same situation.  They dominate their local league, start getting excited about the possibility of going to Nationals and get some lessons and their games improve to a near 4.0 level.

This team now shows up at Nationals and looks like a bunch of sandbaggers, just like the self-rated player would look, but they have followed all the rules and are C-rated players so can't be DQ'd and the proposed rule not allowing self-rates to go to Nationals doesn't affect them.

Second, players playing well above their level didn't necessarily self-rate too low.  Some players simply decide to improve their game and the yearly cycle of re-leveling players allows for someone to improve enough to be well above their level.

This second issue is just an unavoidable part of having tiers/levels of play and these players shouldn't be punished.  There will always be players at the top of the tier and even above it that are on their way to being bumped up.  The very nature of Nationals dictates that you have to have these players to get there, so we shouldn't expect 3.0 Nationals to look like a regular season 3.0 match in our local league play.  By definition almost, it is going to look like 3.5 players.

Now what I'd propose would help address the first issue of self-rated players self-rating too low would be two things.

First, while I can see the argument for not allowing them to go to playoffs, another solution would be to tighten up the buffer given to allow for natural improvement.  Simply lower the threshold for strikes and these players would get DQ'd and bumped up and be ineligible to play on the team.

Second, and more important, I would institute a rule that any self-rated player that gets DQ'd or bumped up in their first year should be treated as a self-rated player again the next year.  And if it happens in their first two years, they are still treated as self-rated in year three, etc.  I don't think you could extend any prohibition on going to Nationals to year two, but this would at least preclude them from sandbagging their first year to get a C rating and then being protected from DQs.

There is no panacea for this, having levels of play will always result in issues where the levels meet, but a few minor changes could reduce the opportunity sandbaggers have to cheat the system.