Sunday, February 24, 2008


Last weekend at a practice for my 3.5 Mens USTA team (affiliated with the Gay Club) I chatted with an acquaintance who had been practicing on adjacent courts with another, non-gay, team.

My friend was unfamiliar with his team and expressed surprise at their conversation, about recruiting players who were underrated by NTRP (National Tennis Rating Program, maintained by USTA) so that they could dominate in the upcoming league play.

My friend, besides dissing people who purposely play down said, “ It’s kind of like cheating, Willie.”

I wholeheartedly agreed. And issued a summary statement with my usual subtlety and care:

The promise of NTRP is a fraud.

After my friend departed I continued the theme with others sitting on the bench, one of whom made a face and pointed in front of his chest at a guy sitting outside the fence, a prominent member of the Gay Club, who is a full time employee of USTA NorCal.

He had gotten a little drift so I boldly, of course, repeated my charge, and received well thought out defenses. My claim was based on eyewitnessing many many matches between players with the same NTRP rating that were not competitive, that is, there was little doubt who would win.

BTW: “NTRP” is a computer program maintained by national USTA that tracks each player’s match results and, based on those results, assigns each player a numeric rating from 1.0 (raw beginner) to 7.0 (grand slam seeds). For USTA team competition, the ratings are divided into half-point steps. So, I used to be a 3.0 player. I improved, so now the computer rates me 3.5.

So I did a little research.

Here’s what I consider to be the “promise of NTRP” taken from USTA NorCal’s website describing league play:

Teams are set up according to NTRP rating, so you're always sure that you'll be facing off against someone of comparable ability. That keeps the competition lively and the atmosphere social.

A reasonable person might conclude that he or she, if the system is working correctly, would have a substantial chance of winning (or of losing) any match played. Anyone who has played USTA league tennis knows this is not true.

So I found this in some FAQs regarding NTRP on the national USTA website:

7. What does it mean to play "competitively" with another player?

A "competitive" match is one in which the outcome is unpredictable (scores such s 6-4, 6-4 or closer). When one player consistently wins with only the occasional loss of a few games, the match is not "competitive." Properly rated, players within .2 of each other should be competitive in playing ability.

This explains why they don’t publish each player’s exact rating, say, to two decimal places. If they did this, a player could look at a prospective opponent’s exact rating, compare it to his or her own, and determine whether or not USTA thinks the match will be “competitive.”

The NTRP computer calculates ratings to multiple decimal points, but reveals the ratings only in half steps. They say I’m a 3.5 but they don’t say where I am within the 3.5 range.

The following charts show how NTRP, by its own definition, produces many “uncompetitive” matches.

The large box shows the full range of 3.5 players. The brackets indicate the range of competitive opponents .2 in red above the star (an individual’s rating), and .2 in green below the star.

Opponents within the red bracket will probably beat you, but you have a fighting chance. You will probably beat opponents within the green bracket, but they have a reasonable chance. Players with identical ratings should win half the time and lose half the time.

The black arrows above and below the colored brackets indicate matches that USTA considers “non-competitive.”

In the first illustration, you (the star) are in the middle of the 3.5 NTRP range. Only 20% of your matches will be non-competitive, 10% at the top and 10% at the bottom.

However, if, as in the second illustration, you are rated at the top of the range, 60% of your matches will feature non-competitive opponents.

Likewise if you are rated at the bottom of your range, except you’re gonna be a big loser.

We could get into more detail (if I weren’t so lazy) about singles vs doubles etc, but at the very least we see that the promise made by USTA NorCal is not sincere, and doesn’t paint a true picture of what a newcomer will discover.

----- o -----


Herman Chin said...

Oh No, USTA League play isn't fair and competitive?
I'll need more details than you've outlined here.

sfwillie said...

Dear Herman,

This was an attempt to explain a common (to USTA league players)observation: two opponents with the same computer rating playing a match in which the outcome is not in doubt--one opponent is obviously better than the other.

If you haven't shared this observation than the rest is irrelevant.

How can this happen when both opponents are rated, say, 3.5?

I thought one explanation was lack of enforcement by USTA officials.

By looking at the websites I saw that the cause is granularity, (3.5 is too big a grouping) to "be sure" that all opponents are of the same ability level.

USTA speak with forked tongue about this, per the two quotes.

I think USTA views leagues as a cash cow, and I suspect that sectional officials secretly support sandbaggers and score-managers because they want to win at Nationals.

If USTA published two-decimal-point ratings, all would be clear.