Monday, May 25, 2020

Poll: What should the USTA do about 2020 year-end NTRP ratings?

COVID-19 has made this a strange year in many ways and USTA League play is no exception.

USTA League Nationals has been canceled, and with that, some sections have decided to cancel Sectionals and other playoffs leading up to those.  With varying levels of playoffs being played, the need to play regular season leagues varies too.

In some areas, play started and a regular season was mostly but not completely finished.  Some of these areas are planning to resume play and complete the season, but others have already said it is canceled.

Other areas may have completed a regular season, but did not get to playoffs yet.

And many areas, for one or more leagues, didn't even get started yet, and it is questionable if matches will ever be played.  Some areas are planning to begin play when allowed for at least a few leagues

Those that do start or resume play may do so with revised formats, some dropping to 3 courts to make more available for more matches, others choosing to play 5 courts instead of 4 (as was the new rule for 40 & Over this year) since it won't be leading up to Nationals.

Many players are just wanting to be able to play again.  I know in my area clubs are just beginning to re-open with restricted rules, but players are flocking to public courts seeking out an opportunity to play.  Will these players want to get back to playing USTA League?  Or with the season disrupted to they not feel the need?

Regardless, there have been and will be more USTA League matches played this year, and as such a big question becomes, what will the USTA do with 2020 year-end NTRP ratings?

Some players played a fair number of league matches as originally planned, and as such have played enough to ordinarily get a new year-end rating.  Those areas that resume/begin play will only add to this total.  So even without Nationals or other playoffs leading up to that, there is sufficient match play to calculate ratings.

But, year-end ratings normally factor in Nationals and how sections and specific players did there as part of their benchmark calculations.  These are an important part of ratings, to help balance things across all of the sections to try to make sure that a 3.5 in Mid-Atlantic is relatively the same as a 3.5 in Northern California.  In the absence of Nationals, what should be done?

One approach would be to just throw 2020 out as if it never happened, and have players just keep their 2019 (or prior if they didn't have a 2019) year-end rating and use it for 2021.  This has some merit, teams were formed and never had a chance to play or finish a season, this approach would let those teams stay as-is for 2021 and get their chance then.

Another approach is to go ahead and calculate 2020 year-end ratings for those that played enough matches with an alternate approach to doing so.  This could simply be using a player's final dynamic rating, or some other factors could be included in the absence of Nationals.

The problem with the first approach is that player's abilities do change year to year as players practice and improve, or as they get older and skills diminish.  Throwing out 2020 completely means there is no chance for any of these changes to be reflected.  It would seem that if a player played the minimum 3 matches to get a new year-end rating, they should get one so if their game improved or declined, that can be reflected and they play at the appropriate level for 2021.

This is especially true for self-rated players.  If they get the requisite 3 matches in, why should they have to stay self-rated for 2021 leagues?  About 35% of all self-rates in a given year are bumped up or down at year-end, so it would seem to be a big miss if all of these players remained self-rated.  You would have a significant number of players who should have been bumped up and get to play too low a level, or another significant group that lost badly in 2020 and are stuck having to play at too high a level.

Appeal players are also a group that one could argue should get a new year-end rating.  These players had a higher or lower rating but were close enough to appeal.  If their rating has leapt up or down from the matches they did play in 2020, that would seem to be a pretty strong data point saying the appeal was perhaps not appropriate and they really should be at the higher or lower level.

Personally, I think there should be 2020 year-end ratings and all players that played 3 matches should get one.  I think there are more problems from players playing with ratings that are based on results from 18+ months ago than there are for 2020 teams not having had their chance to play a full season.

Now, the dark side of league play is that some players that know their remaining matches will count will use them to tank matches to try and offset good results from earlier in the year to stay at their level, or try to get bumped down.  If this is a significant concern, the USTA could elect to only use matches played prior to the suspension of play.  This still seems more reasonable to me than throwing out the whole year.

Note that even if all matches played are used, and certainly if only those played prior to the suspension count, there will be a lot of players that don't end up playing 3 matches and have their 2019 rating carry over anyway.  It just seems wrong to not use matches played in good faith for those that did play, especially for self-rated players.

For reference, in the 2019 rating year, there were just under 134K team matches played in 18/40 & Over leagues.  So far in 2020, just under 35K have been played.  Thirty-five thousand is a significant number, but about 25% of what might have been played had 2020 been a normal year.  This means, a lot of players will stay the same for 2020 anyway, no need to throw all of 2020 out to accomplish that.

But what do you think?  To get your feedback, I've set up a poll with a number of options, please vote as I'm genuinely curious what people think in general.


Should the USTA publish 2020 year-end NTRP ratings?

Yes - Just use year-end dynamic rating if played 3 or more matches
Yes - Figure out a "Nationals-less" benchmark calculation
Yes - But give more lenient appeal thresholds
Yes - But only use matches played prior to the suspension of play
Yes/No - Only do it for self-rates and appeals, not for everyone else
Yes/No - Only do it for self-rates, not for everyone else
No - Publish no new ratings, everyone carries over from 2019
Other
Please Specify:
Created with Quiz Maker


5 comments:

  1. Another consideration that probably cannot fully be accounted for... many people probably stopped playing completely for several months, so there may be a noticeable drop in their level of play. The break in play probably won't affect everyone by the same amount, so there will probably be a "recalibration" period where the dynamic ratings don't necessarily reflect reality.

    ReplyDelete
    Replies
    1. There could be some truth to that, but that's always a factor, too. I doubt even if a player hasn't picked up a racquet for 2 months that their ability has decreased much, if any. If they've decided to play league again, I'd bet they'd practice at least 2-3x before their first match and shake off the rust quite a bit. But if not, so be it. But, that's the tricky nature of the usta ratings system. Even if everyone plays honorably and doesn't sandbag, there's always going to be a few wacky ratings. The usta rewards players at certain ratings like 3.49, 3.99, 4.49 to be at the top of their level, while ratings like 3.55, 4.05, 4.55 for example are highly penalized.

      There's need to be new ratings in some way. At the very least, the matches before suspension of play need to count. But, even after play resumes, I think they should count, too, unless sandbagging is egregious and widespread, then may need to be modified.

      Delete
    2. Good points Michael. I don't think a significant number of people are going to show up to play a league match cold turkey any more than they normally would. I think the exceptions that may fall into this camp are fewer than the number of players that would fairly get a new level from new ratings.

      Delete
    3. I think you are underestimating the casual nature of most recreational league play. At least in my area, I would say the vast majority of the players do not practice outside of league matches. Granted, those are not the people we would worry about sandbagging... But it still could have an effect on the overallqoverall of play.

      Mind you, this problem is nothing specific to the NTRP system... UTR et al. would have the same issues.

      Delete
  2. I would say the exact opposite. I know there's some players who might only play league matches for the most part, but the vast majority(90-95+%) that I know are definitely playing much more or at least some outside of league matches, especially the players advancing to playoffs, sectionals, nationals, etc., throughout the season. The casual players that don't care about winning much and only play league matches are the types of players we really don't have to worry about ratings with.

    These players are just showing up for matches and never practicing, then they're almost always rusty. I wouldn't think a little bit longer of a break would diminish people's level of play much. I'd imagine that the places in the country that will be open for usta will have courts available for use somewhere for at least a week or 2 before any league matches begin. Players will have time to practice and play before. If this is wrong and courts are only available the day of the first league matches, then maybe the first 1-2 matches could feature rusty players, but after that, not really. Also, if everyone is rusty, then it's basically irrelevant as same for everyone.

    ReplyDelete