Wednesday, July 30, 2014

When should I default or retire from a USTA League match?

Let me start this post by saying I do not advocate defaulting or retiring from matches unless there is a valid reason.  And in the case of defaults, they should be communicated ahead of time so as to avoid inconveniencing players that have to arrange to make a match and potentially drive significant distance and don't want to show up to find out they have no match.

That said, there are implications on one's NTRP rating from retirements and defaults that is is good to be aware of.  So here goes.

I think everyone knows, and common sense dictates, that if a match is a complete default, it doesn't do anything for your rating one way or the other.  In fact, most defaults only have the "winning" team entered so there isn't even an opponent to calculate a rating against.  The defaults are entered as 6-0, 6-0 scores for sets and games won/lost purposes in team standings and for tie-breakers, but the score doesn't matter to the players' ratings.

But what about scenarios where a team or player show up late and tardiness penalties kick in?  In this case, the match is not defaulted but games/sets may be forfeited if a team is late but not so late to make the entire match a default.  This has the potential for impacting ratings in perhaps an unfair way as, to my knowledge, these scores are entered normally and there is nothing in the system to indicate how many games were actually played vs how many were from a tardiness penalty.

Thus, by being late and starting a match down 1-0, you have possibly hurt your rating as you've given up a game you might not have otherwise.  Worse, and I recently had a report customer where this happened, you might have to forfeit a set 6-0 and start the match at that point.  That will significantly affect a player's rating regardless of whether they win or lose.  Certainly one doesn't want to default the match if it could be won, and as I noted above it isn't really fair to all players involved to not get to play, especially if all the players are there, but if the match is likely a loss anyway and you don't mind the 6-0,6-0 result for the team standings, defaulting the match instead of playing it would likely be better for your NTRP rating.

As far as retirements go, my understanding is that there is a point in a match where enough of it is played and it becomes a match that is counted, but prior to that point if a retirement were to occur it wouldn't be counted.  I haven't seen an authoritative source and have heard differing views on what this point is, but believe it is most likely one set or 10 games.

So again, I would not advocate retiring in a match unless there is a real reason to do so, but if you are injured and clearly won't be able to continue and win the match and the injury is causing you to play well below your normal capability, if you don't want to adversely impact your rating you are better to retire before the first set is complete than to complete it and then retire.

Now, one might argue that retirements shouldn't be counted at all as they might not properly reflect the capabilities of each player.  But I think the idea is that a retirement occurs usually due to an injury that occurs during the match and if it is bad enough to be a retirement, it happens immediately.  In this case, the games played to that point were fairly played and it makes sense to include them.

The first issue with tardiness seems like a pretty serious flaw in the system.  Now, tardiness penalties are probably not that common, and when they occur it is probably just a game or two so it probably isn't a big deal.  But the full set penalty certainly impacts ratings in an unfair way it would seem.

What do you think?

Monday, July 28, 2014

USTA League playoffs continue, paths to advance, more report customers do so

This is a busy time of the year for USTA League play as rounds of playoffs continue to be played and more teams advance.  This weekend found some districts in the Midwest playing having their playoffs in various divisions and levels, and and the Southern section playing their doing the same.  And I'm sure there was other playoff activity.  And it is always fun to hear from report customers after playoffs to hear of their success, either individually or as a team, and their excitement to hear about how their matches have affected their NTRP dynamic rating.

Something that has always been interesting to me is how all the different sections and districts are organized. and how many rounds of playoffs a team may have to go through.

The shortest path is usually in a section like Hawaii, where simply beating out a few teams in local league play can be all that is required to advance.

In my section, the Pacific Northwest, we typically have local league play followed by local playoffs which serve as our Districts effectively, and then we advance to Sectionals.

In the South, states are aligned with districts and a team may have local playoffs to get to districts (states) and then on to Sectionals.

In the Midwest, the districts are smaller so a team may play Districts which gets them to States (Regionals), and then to Sectionals.

So it varies, and I guess the teams with a longer path just get the opportunity to play more, and isn't that the goal, to play more tennis?

Friday, July 25, 2014

More Interesting USTA League Stats - Middle States section strength of tennis players by NTRP level and district

Continuing in the series ofinteresting stats blog entries, here is how the Middle States districts compare.

The chart below shows the average NTRP rating by level for each district in the Middle States section of the USTA. Ratings used are the Estimated Dynamic NTRP Ratings used in calculating Individual and Team reports.

Here is a static image.


You can hover over each bar in the interactive chart and see what the average rating for that level in that district is.

Wednesday, July 23, 2014

More Interesting USTA League Stats - Inter-Mountain section strength of tennis players by NTRP level and district

Continuing in the series ofinteresting stats blog entries, here is how the Inter-Mountain districts compare.

The chart below shows the average NTRP rating by level for each district in the Inter-Mountain section of the USTA. Ratings used are the Estimated Dynamic NTRP Ratings used in calculating Individual and Team reports.

Here is a static image.

You can hover over each bar in the interactive chart and see what the average rating for that level in that district is.

Tuesday, July 22, 2014

More Interesting USTA League Stats - Midwest section strength of tennis players by NTRP level and district

Continuing in the series ofinteresting stats blog entries, here is how the Midwest districts compare.

The chart below shows the average NTRP rating by level for each district in the Midwest section of the USTA. Ratings used are the Estimated Dynamic NTRP Ratings used in calculating Individual and Team reports.

You can hover over each bar in the interactive chart and see what the average rating for that level in that district is.
Should the interactive chart above not render correctly, here is a static image.

Another scenario where a Computer (C) rated player had a successful grievance filed against them

I wrote last week about when a USTA NTRP C rating isn't really a C rating and I heard about another scenario today.

This player had played at a small college but self rated as a 2.5, got DQ'd to a 3.0, and had a year-end rating of 3.5, then while playing as a 3.5 had a grievance filed against them.  The USTA ignored the C rating and instead looked at what they should have self-rated at and DQ'd them and promoted them to be a 4.5.

So if you think you can slip one by and ignore your experience and the self-rate guidelines, watch out, you could have a grievance filed against you even after your get the coveted C rating.

Sunday, July 20, 2014

More Interesting USTA League Stats - Pacific Northwest section strength of tennis players by NTRP level and district

Sectionals in the Pacific Northwest section are coming up in August, so this interesting stats blog entry will tell us a bit about how the districts compare.

The chart below shows the average NTRP rating by level for each district in the Pacific Northwest section of the USTA. Ratings used are the Estimated Dynamic NTRP Ratings used in calculating Individual and Team reports.

You can hover over each bar in the interactive chart and see what the average rating for that level in that district is.
Looking at averages for players playing in each district, Northern Oregon and Northwest Washington lead the way for 3.0s, Southern Oregon for the 3.5s and 4.0s, and Alaska for the 4.5s.

Should the interactive chart above not render correctly, here is a static image.

More Interesting USTA League Stats - Southern section strength of tennis players by NTRP level and district

I asked on facebook what section people wanted to see interesting stats on next, and Southern won out, so here you go.

The chart below shows the average NTRP rating by level for each district in the Southern section of the USTA. Ratings used are the Estimated Dynamic NTRP Ratings used in calculating Individual and Team reports.

You can hover over each bar in the interactive chart and see what the average rating for that level in that district (state) is.
When you look at it at the district level, there isn't a huge amount of variance. But Kentucky has the highest average for 3.0s, Kentucky, Mississippi, and South Carolina have the highest for 3.5s, North and South Carolina have the highest 4.0s, and Arkansas the highest 4.5s.

Should the interactive chart above not render correctly, here is a static image.

Wednesday, July 16, 2014

Texas USTA League early start ratings are out

The Pacific Northwest section released their early start ratings a few weeks ago, and now Texas has released theirs.

While I did not go through and do any predictions for early start bumps like I did for the PNW, I am already getting feedback from some Texas players I did reports for that my predictions were correct.

And I just went and checked 25 players that I just did reports for in the past few weeks and was correct on all but two, both of them were bumped up to 4.5 and I had them at 3.99. So not off more than a few hundredths on those and correct on all the others.  And if you give me a little leeway, even just a 0.02 margin of error, I would have gone 25 for 25.  I'll take that.

Now, early start bumps can appeal their rating, and if you do or know of anyone that has, I'd be interested to know it as that is a way to get an idea of how close to the threshold a rating was and that helps me improve my accuracy.

Where are people that get Estimated Dynamic NTRP Rating Reports from? Everywhere USTA League is played!

This is a busy time of year in USTA League as regular seasons are wrapping up and/or local or district playoffs are being played.  This has resulted in a lot of people being interested in individual and team Estimated Dynamic NTRP Rating Reports which is great and keeps me busy.

Where are the players from that get reports?  In the past couple weeks alone, I've done reports for players in over half the states in the country.  This includes:
  • Alabama
  • California
  • Colorado
  • Connecticut
  • Florida
  • Georgia
  • Illinois
  • Louisiana
  • Maryland
  • Michigan
  • Minnesota
  • Mississippi
  • New Jersey
  • New York
  • North Carolina
  • Ohio
  • Oklahoma
  • Oregon
  • Pennsylvania
  • Tennessee
  • Texas
  • Utah
  • Vermont
  • Virginia
  • Washington
  • Washington D.C.
If you too are interested in finding out more about your NTRP rating, contact me about getting a report.

Monday, July 14, 2014

More Interesting USTA League Stats - Strength of tennis players by NTRP level and area - Texas 4.0s

Continuing in my series of posts on interesting USTA League stats, here is the average ratings for 4.0s by area in Texas.  I wrote yesterday about Texas 3.5s.


A few areas stand out with much higher than average averages, Midland, NoHo, and Waco.  Based on this, we might expect that those areas will do well at Sectionals, but again, these are averages and not necessarily an indication of relative strength of the teams that make it there.

Sunday, July 13, 2014

More Interesting USTA League Stats - Strength of tennis players by NTRP level and area - Texas 3.5s

In my series of posts on interesting USTA League stats, I recently wrote about average ratings for a given level in different areas within a section.  I first took a look at Mid-Atlantic 3.5s and 4.0s.  Now for a look at Texas 3.5s.


A couple areas stand out, Midland and somewhat, Waco.  That may be surprising, but the big cities like Dallas and Houston likely have a broader range of players and so they have lower rated ones that pull their average down.

Stay tuned for more.

When is a USTA NTRP computer (C) rating not really a C rating?

A lot of people that play USTA are aware of the general rules regarding self-rated (S) players being subject to three strike disqualifications.  This is where a new player goes through the self-rating process to get assigned to an NTRP level but, whether innocently or deliberately, they self-rate too low and end up being promoted up a level mid-year and results to date are usually reversed as well.  This process is in place to try to catch players that self-rate too low as it is unfair to the established players at that level.

What not so many may be aware of is that there is also a process where a grievance can be filed against a player asserting that they self-rated incorrectly and should be playing at a higher level.  Specifically, USTA documents state:
Any league captain, coordinator or member of a championship committee has the right to file a NTRP Grievance against a player and/or captain. The grievance will claim that the player and/or captain committed or condoned obtaining a Self Rate or possesses a rating which is at a lower NTRP level but has demonstrated the ability to compete at a higher NTRP level.
They also mention:
  1. The Grievance must be against a player of the following rating types: This list includes S, M, T, Dynamic Early Start League and Medical type ratings. A player with a valid NTRP Computer (C) rating, a Benchmark (B) rating, or a granted Automatic Appeal of a Computer (C) rating is not subject to a NTRP grievance.
  2. The complaint must state that the player has self-rated or possesses a rating which is too low and may also include information that a team captain or other person conspired to obtain that incorrect rating.
    I have bolded and underlined the important part for the purposes of the following discussion.

    I was recently made aware of an interesting NTRP rating grievance where a C rated player had a grievance filed against them and it was upheld!  This would seem to be in direct conflict with the grievance protocol documented above.

    The situation is a high schooler new to the game self-rates as a 3.0 a few years ago but only plays in leagues that don't count towards his rating so he doesn't get a C rating.  His game improves and he ends up appealing up and/or self-rating again at 3.5 in a following year and does play Adult and gets a 3.5C rating at the end of last year.

    What he fails to disclose or use when self-rating again however is that he did end up playing in high school nor does he mention that he walked on at a small Division II school and played there with meager success.

    He plays this year with his 3.5C and does very well winning all of his matches, but some in match tie-breaks and only one that would really be considered a lopsided win.  Nevertheless, someone files a grievance against him citing his playing history in high school and college that would arguably require he self-rate at 5.0.  The committee agrees and he gets DQ'd up to 5.0!

    So apparently, at least in this section/district, a C rated player can have a grievance filed against them.

    Should the player have self-rated at a higher level when they redid it after walking on at a division II college?  Yes, and there are provisions for doing so at the right level and then appealing down.  But is he playing at a 5.0 level?  Almost certainly not, my ratings have him as a likely bump up to 4.0 at year-end, but nothing egregious seems to going on here.

    I understand the grievance is being fought, so perhaps this isn't over yet, and if I hear more I will share it.  But beware C rated players, you may not be as golden as you thought.

    Saturday, July 12, 2014

    More Interesting USTA League Stats - Strength of players by NTRP level and area - Mid Atlantic 3.5s

    I wrote yesterday about the average rating by area in the Mid-Atlantic section for NTRP 4.0 level players.  Today, I add the chart for NTRP 3.5 level players.


    We see again that some areas have players that are significantly stronger/weaker than others.  But are they the same as for the 4.0s?  For reference here is that chart again.


    They aren't necessarily the same.  Charlottesville for example has among the strongest 4.0s but is weaker for 3.5s.  And going the other way, Forty West had the lowest average of 4.0s but is mid-pack for 3.5s.

    Another interesting observation I did not make yesterday is that you'd expect the "average" 4.0 to be at 3.75, but the majority of the area did not have averages around that, several being well below.  This is likely just a reflection of the median player in USTA being closer to a 3.5, so you are going to find more weak 4.0s than strong 4.0s and that will skew the average down from the mid-point.  This is confirmed by the median 3.5 being 3.25 and the averages being a whole lot closer to being right around that number.

    Note also that these averages don't mean players that live in the respective areas are stronger/weaker, rather it is just that players that play in those areas.  Many players in Mid-Atlantic play in multiple areas and so a lot show up in multiple of the averages.  So this rather may just be an indication of which areas draw the stronger players to play.  It doesn't mean there can't be strong teams in the other areas, just that there aren't as many higher rated players.

    So if you are trying to make Districts or Sectionals, you may want to field a team in one of the areas with the lower average as the road to get there may be easier.

    More to come.

    Friday, July 11, 2014

    More Interesting USTA League Stats - Strength of players by NTRP level and area - Mid Atlantic 4.0s

    I was recently asked if a perception that players from a certain area were stronger than others was true or not.  I thought it was a great question and went about seeing what I could determine from the data I have from estimating NTRP ratings.

    What I've come up with is something like the below.  This happens to be for the Mid-Atlantic section, specially looking at the average rating of the 4.0s that played in each area.


    Now it is important to note that in the Mid-Atlantic section, a lot of players will play in multiple areas.  So those players are being counted in the average for multiple areas.

    I don't know if this confirms or goes against the conventional wisdom or perception, perhaps folks will comment on that.  But I think it is interesting how much variation there is, from 3.63 in Forty West  to 3.78 in Talbot County.  Is this much difference just what you'd normally see given enough areas?  Do better players seek each other out in Talbot but shy away from Forty West?

    It would be interesting to see if the same areas are stronger/weaker at the adjacent NTRP levels too and if so, to think about why that might be.  Stay tuned.

    Update: Post on Mid-Atlantic 3.5s now available.

    Sunday, July 6, 2014

    Have you received a three strike DQ'd in USTA League tennis? Get a free report that explains why

    I made a free report for three strike disqualified players offer last year and had quite a few people take me up on it.  I just had someone ask about it so figured I'd make the offer again.

    So, if you've been three strike disqualified during the 2014 USTA League season and can provide me your DQ letter that includes the matches that were strikes, I will generate an Estimated Dynamic NTRP Rating Report for you that will help explain the DQ to you.

    Why do I do this?  First, it is always interesting to look at DQ's as there are some strange ones out there, but I can usually explain them.  Second, looking at DQ's and knowing which matches are strikes helps me validate my ratings and algorithm so I can improve the accuracy for everyone.

    So if you'd been DQ'd, let me know!

    Wednesday, July 2, 2014

    Pacific Northwest USTA League Early Start NTRP ratings are out

    The Pacific Northwest section of the USTA has published their Early Start League Ratings.  I posted a short list as a sampling of players that I expected to be bumped up, and I was correct on 19 of the 20 I listed.  The one I missed was a self-rated player that had only played a few matches and these are always a little tricky to calculate.  Still 95% and missing on only one isn't bad and is a testament to the accuracy of my ratings.  I'll take it.

    I will try to do a more thorough analysis over the next few days and report back.

    Poll: Do you want your USTA NTRP rating to go up, down, or stay the same?

    It is that time of year when local league play in the main USTA Adult Leagues may be wrapping up meaning that for some, the matches that contribute to their NTRP rating have been played.  As a result, people start to think about if they will be bumped up or down for next year.

    In some sections like the Pacific Northwest, we have 2015 leagues that start early, as soon as late summer in 2014.  In order for these leagues to try to have players play at the right level, early start ratings are released, basically a snapshot of where everyone's dynamic rating stands, that are used to determine the level players must play at for these early start leagues.  These early start ratings are usually a very good indicator of whether a player will be bumped up or down at year-end as well.

    A question I get asked is why people get Estimated Dynamic NTRP Rating Reports from me.  The simple answer is that they want to track progress of their rating towards a goal or just see details of how individual matches are rated and what direction their rating is going.

    Now most people want their rating to go up, but not all.  So I thought it would be interesting to put a poll up to let all my readers weigh in.  See it in the sidebar to the right.

    So, whether you have early start ratings and will get a glimpse into your rating, or have to wait for year-end, what are you wanting your rating to do?  Vote in the poll to the right and lets see if USTA players want to improve or just stack teams for runs at Nationals :)

    Pacific Northwest 2014 USTA Early Start Ratings Predictions - Northwest Washington

    Early start ratings for the Pacific Northwest are due out tomorrow, and it is always fun to see if my ratings agree.  As a reminder, these ratings are used to determine the level a player can play at for the 2015 leagues that start play early, e.g. in the late summer/fall of 2014.

    In the spirit of that fun, here are some players that my ratings say should be bumped up on the early start list for the Northwest Washington district.

    NameCurrentBump
    Elisabeth Beebe3.5Up
    Jill Hawkins3.0Up
    Barbara Smith3.5Up
    Badrinath Vengalathur3.5Up
    Brad Bennett3.0Up
    Bruce Bethards3.5Up
    Terry Ward3.0Up
    Dan Wu2.5Up
    Traci Lynch3.0Up
    Michael LaMoria3.0Up
    Mark Ball4.0Up
    Paula Holtzclaw3.0Up
    Sana Watterson3.5Up
    Petra Schuler4.0Up
    Patty Terhar2.5Up
    Maria Recchi3.0Up
    Susan Shinoda3.5Up
    Mara Vinnedge3.5Up
    Mark Metcalf2.5Up

    I'll let you know how I do predicting tomorrow!

    Tuesday, July 1, 2014

    Update: Pacific Northwest USTA League Early Start Ratings Coming Soon for 2015 leagues!

    I wrote a week ago or so about early start ratings in the Pacific Northwest section of the USTA and said that matches through the Northwest Washington 40 & over local playoffs would be included.  I was wrong!  The matches played at the 65 & over Sectionals held this past weekend in Sunriver will also be included.

    Stay tuned for more about early start ratings.