Fiend at Court Unplugged
The USTA has announced that starting in 2021, NTRP Adult tournaments will be divided into three age tiers. The stated reason for NTRP age divisions is that it will increase participation at tournaments. In fact, the planned elimination of Mixed Doubles from NTRP tournaments was justified by projecting that the additional age tiers will create so much participation that there will no longer be courts available to accommodate Mixed.
No one can look into a crystal ball for an accurate prediction on how this new initiative will work out. However, we can examine historical participation data to determine if there is any credibility to the claim or possibility that more age tiers will magically conjure up new players. As an engineer, I am wired to look at problems through the lens of cold hard data.
NTRP Tournament Participation Data
First, some boring details. I scraped my participation data off TennisLink. Getting the data off the site is relatively straight forward, but significant scrubbing is necessary to make it useful for large scale analysis. For this initial data set, I performed this manually. I have the intention to automate the scrubbing process in the future, but I needed to make sure I understood the nuances of the data set before developing the data transformation code. (I know, what a geek.)
As a starting point in TennisLink, I used the rankings list for each calendar year. In my home section of Texas, there is a complete rankings list produced each year that also happens to be an accurate and comprehensive list of all players who played tournaments in the given year. I was astonished to discover that every section does not publish the same comprehensive list. Additionally, some sections that do publish year end rankings are quite obviously only including a subset of the top players.
This brings me to the inadvertent discovery that I am 100% behind the part of the 2021 tournament initiative that will standardize participation data into a unified format. The TennisLink tournament participation records of some of the other sections is, quite simply, unusable. After struggling with that fact in this analysis, I am galvanized into the position that national standardization is clearly needed.
Since the inaugural 2018 NTRP National Championships were announced in the middle of the 2017 playing season, none of the tournaments in my home section of Texas were conducted with NTRP 50+ divisions in that year. That changed the following year in 2018 when a full slate of tournaments with NTRP 50+ divisions were conducted. This makes it possible to look at the participation numbers across those two years to determine what effect dividing the NTRP divisions by age had on participation.
Number Crunching
For my initial pass at the data, I decided to focus on Women’s 3.5 Singles. The rationale behind that selection is because 3.5 has the highest participation rate for Women’s Singles in the Texas section. When drawing inferences from data, larger data sets generally have less variance. Additionally, Singles was selected over Doubles to eliminate the influence of partner’s from this initial pass of analysis.
I really wanted to perform this analysis on the national data set, but was stymied by the lack of uniformity in the national data. As a footnote to this post, I included a table which contains the total participation rates across all the USTA sections for NTRP Women’s 3.5 Singles 2017 tournament participation. Texas leads the list with 194 players. Southern was second with 123 participants and Northern California clocked in at 68. The drop off in participation rates is fairly dramatic.
With the addition of Women’s NTRP 50+ in 2018, the total participation for Women’s 3.5 Singles dropped to 181 total players in Texas. At the same time, the new NTRP 50+ division at that level also attracted 33 unique participants. Bad data analysis would stop there, claiming a net rise from 194 to 214. However, it is not a rise in participation at all.
19 of the 33 players who played 50+ divisions also played the standard 18+ during the year. Thus this apparent gain in participation represents the player pool shifting around, not an increased number of players. The stacked bar chart below illustrates this with much greater clarity than I just articulated. Participation in Texas was flat in Women’s 3.5 singles.
I wanted to compare the effects of participation across the other sections, but was thwarted by the fact that neither Southern or Northern California, the next two sections in the Women’s NTRP 3.5 participation list, published a 3.5 50+ rankings list in 2018.
I can observe that Southern participation dropped from 123 to 111 players in the standard NTRP 3.5 division. In Northern California participation dropped from 68 to 60 players. Due to the lack of standardization of the data and rankings lists across sections, I don’t really have a good way of knowing whether there was some 50+ play that may have siphoned off players that was not reflected in the rankings.
Cold Hard Data
Based on this narrow analysis of Women’s 3.5 participation in Texas, the initiative that created an age-based NTRP division did not drive an increase in participation. In order to draw hard conclusions analysis across a broader set of participation data is in order. In the true spirit of this site, that is coming tomorrow.
I have analyzed the complete participation data for all women who played NTRP level divisions in a tennis tournament in both 2017 and 2018. This eliminates variance associated with players being bumped up or down between these two years. Additionally, this expanded aperture makes it possible to also examine whether the prospect of qualification for the NTRP National Championships drove increased participation in Texas.
Grab a blanket. The data gets cold in a hurry.
- Fiend at Court Request, July 21, 2020, (Official response to emailed questions.)
- 2021 Adult Tournament Changes, USTA National Webinar, undated.
- USTA Adult Tournament Changes for 2021, USTA National Website, viewed 7/25/2020
| Section | Total Players |
| Texas | 194 |
| Southern | 123 |
| Northern California | 68 |
| Florida | 44 |
| Pacific Northwest | 32 |
| New England | 23 |
| Eastern | 19 |
| Middle Atlantic | 12 |
| Hawaii | 12 |
| Southern California | 8 |
| Missouri Valley | 7 |
| Northern | 5 |
| Middle States | 3 |
| Midwest | 0 |
| Southwest | 0 |
| Carribbean | 0 |
