Sunday, December 28, 2014

2015 has already begun, last chance to get extended 2014 reports,

Most of you know that 2014 year-end ratings came out at the beginning of this month, and if you've been reading this blog, you know that these ratings actually included matches played through just 11/9/14.  So what happens to the matches after 11/9?

They actually are part of the 2015 league year.  That year generally runs from November thru October, the specific end date depends on when Adult Nationals finish, and year-end ratings generally come out the Monday after Thanksgiving (to confuse things though, the USTA gives a year-end rating date of 12/31).  So there is no other place for the matches after 11/9 to go than into the 2015 calculations.

So if you've played matches after 11/9, you've actually started 2015 already!

So if you get an Estimated Dynamic NTRP Rating Report now, what do you get?  For the next couple days, you actually have a choice.  If you'd like your 2015 report showing just the matches since 11/9, I can do that.  But if you'd like an "extended" 2014 report, I can still do it thru 12/31.  This will show you your normal 2014 report, but just tack on the matches after 11/9 as if it was all part of the same year.

Not only does the extended report save you money by giving you two reports in one in a way, this sort of gives you a "what if" so you can see what your year-end rating might have been if those last few matches had been included in 2014.  And I've seen a few cases where someone's rating did go above/below a bump threshold after 11/9.

As always, if interested, contact me!

Wednesday, December 24, 2014

Another last minute Christmas gift option - Gift certificates for 2015 Estimated Dynamic NTRP Rating Reports

It is Christmas Eve and what to do for your tennis playing friend that told you about Estimated Dynamic NTRP Rating reports.  You know they already got their report for 2014, they showed it to you, so the Christmas stocking stuffer report won't do the trick, but how about a gift certificate for future reports?

I'm now offering gift certificates to be used during 2015.  See below for an example.  Like the stocking stuffer reports, they will also be provided in a printable PDF to put in that Christmas stocking.

And a gift certificate doesn't have to be for just one report.  Getting an initial report and then updates throughout the year is a great way to monitor a rating and progress towards a goal, so you can purchase an initial report and any number of updates.  As a bonus, if you purchase a certificate with an initial report and four or more updates (five or more total reports), the total price will be discounted 10%.

For those of you looking for a way to lower the cost of getting reports in 2015, yes, not only can you get gift certificates for your friends and teammates, but you can purchase one for yourself and take advantage of the 10% discount.

Contact me with any questions or to order any gift certificates.

Tuesday, December 23, 2014

Christmas reports now include a single page PDF, perfect for stocking stuffers for that tennis nut in your family

Estimated Dynamic NTRP Rating reports are a great Christmas gift for any of your USTA League playing family and friends, and I've improved the offering by now including a nicely formatted PDF with the summary, partner report, and match/dynamic ratings chart all fitting on a single page that can be conveniently printed and used as a stocking stuffer.  It doesn't include my explanations or other text, but I'll still include that in the e-mail that is sent.

Here is an example.

I'll be including the PDF for no extra charge the rest of the year upon request, so contact me and let me know if you want a report and if you want the "stocking stuffer" option!

Saturday, December 20, 2014

2014 USTA League year-end bump percentages for NTRP 4.5 broken out by section and gender - Interesting tennis league stats

Continuing on the quest to see how players were bumped up/down at different levels, here are charts showing how players at the 4.5 level were bumped up or down by section and gender in the 2014 year-end NTRP ratings.

Unlike what we saw for he 3.0, 3.5, and 4.0 levels, there is not the bias towards more bumps up than down with many sections have more players bumped down than up, particularly for the women.  There are a few exceptions though.

For the men, most sections had more bumps down than up or close to it.  Narrowly having more bump downs were Eastern, SoCal, and Texas, and Caribbean had a fair number more, but once again, the Pacific Northwest was hammered with nearly 25% of the 4.5s bumped up to 5.0 and less than 5% bumped down.

For the women, there were even fewer bumps up, just the Pacific Northwest and SoCal having more players going up than down.  Interestingly, Caribbean women received no bumps up, just bumps down.

Need a last minute Christmas gift for a USTA League tennis nut? An Estimated Dynamic NTRP Rating Report could be it!

Christmas is less than a week away and you still haven't gotten a gift for a teammate or that tennis nut family member.  The malls are crowded and picked over and buying online may not make it in time unless you pay extra for faster shipping.  What to do?

An Estimated Dynamic NTRP Rating Report could be the perfect gift.  Reports are generated and sent the same day (or first thing the following day if ordered very late) so it will get there in time, and it is a great way to let your tennis fanatic know how their matches really rated during the year and how close they really were to (not) being bumped up or down.

Contact me for more information or to request a report.

Saturday, December 13, 2014

New Estimated Dynamic NTRP Rating Report Chart

I've made a few tweaks to the chart included with an Estimated Dynamic NTRP Rating Report and an example is below.

Singles and doubles matches were already in a different color in the chart, but I've cleaned it up a bit and now also call out those matches that were in the playoffs so you can more easily see those results.

What do you think?

2014 USTA League year-end bump percentages for NTRP 3.0 broken out by section and gender - Interesting tennis league stats

Continuing on the quest to see how players were bumped up/down at different levels, here are charts showing how players at the 3.0 level were bumped up or down in the 2014 year-end NTRP ratings.

Given the bias towards bump ups we saw from the 3.5 level, it isn't a surprise to see more bumps up than down, but the 3.0s take it to another level.  This is probably to be expected as 3.0 is a level many new players start at and as they improve, they'll be bumped up.  So there are certainly a lot of new 3.5s to replace those that got bumped up to 4.0.

And while a few sections stand out being bumped up more than others, notably Caribbean and Southern California, there are fewer anomalies at this level.

Wednesday, December 10, 2014

The depths some players will go to to try to not get bumped up - The dark side of USTA League

I recently became aware of another seemingly blatant attempt by a few players to deliberately manipulate their ratings down to avoid being bumped up.  While not on the scale of the 2014 3.0 men's Nationals team I wrote about before as it wasn't an entire roster, it is arguably more egregious because it was by established players and the excuse of being new or not knowing simply doesn't fly.

This situation happened in my back yard in the Pacific Northwest and involved our One Doubles league.  This is a summer league we have that involves a team of at most 4 players playing a single doubles match for each team match.  You can think of it as our version of Tri-Level, we just split a Tri-Level team into individual teams.

This league is played in the summer, usually right after our local playoffs.  So teams have finished their "normal" seasons and are playing for fun or to get to play outside (most of our normal leagues are indoors) and while there are playoffs and you can advance to the Tri-Level Invitational, it isn't as widely played and generally not as serious as the 18+ and 40+ leagues.

The team in question had players from a local 4.0 playoff winner that was advancing on to Sectionals and ultimately Nationals, so had done very well.  It turns out that only 3 of the rostered players ended up playing in the One Doubles league, and they were all very high rated 4.0s.  One was an early start bump to 4.5 and my ratings had the other two very close.

If they were to do well in playoffs, they would likely be bumped up to 4.5 and in apparent fear of this, they went about using the One Doubles league to throw matches to manipulate their ratings down.  That is appalling enough, but the degree to which they did it takes things to the next level.  Remember now, this is normal doubles played against other 4.0s and these guys are higher rated 4.0s.

One player played 6 matches, losing every one, the closest score being a 6-1,6-3 loss and a full 8 of the 12 sets being 6-1 or 6-0 losses.  Another played 4 matches also losing every one, the closest score being 6-2,6-1.  The third played 2 matches winning a total of 6 games.  And even worse yet, some of the opponents were 3.5s playing up.

The ratings they generated from these matches were clear aberrations and not consistent with what they had done all year.  Were these players injured?  Did they lose their ability to play?  Not the case at all.  Each player subsequently did well at Sectionals and Nationals against stronger competition going 5-2, 5-3, and 5-3 there.  Was it because One Doubles was outside?  No, their Sectionals and Nationals were outside too.

So it seems pretty clear that all 6 of the matches played were deliberately thrown in an effort to manage their ratings down.  There are not strong enough words one can use to condemn this behavior and since this is a blog and will live on forever on the Web somewhere, I'll refrain from going overboard, but it is very very sad to see someone stoop to this level as it isn't fair on numerous levels.

First, they are trying to defraud the system and manipulate their ratings so they can play at a level they shouldn't be eligible to.  This goes against the spirit and rules of USTA League.  No, the USTA isn't perfect, but when players play normally, the system works pretty well.

Second, it isn't fair to their competition who signed up for a league to play competitive matches.  Playing a match where your opponent is likely or apparently missing shots deliberately or goofing off or whatever they did to lose so badly is not achieving this goal and diminishing the experience of playing in the league.

Third, it isn't fair to their opponents and given how the system works, all of their opponents' opponents, etc., as it is skewing ratings that could lead to other players being bumped up or down that shouldn't.  For example, a 3.5 playing these very strong 4.0s and getting a 6-1,6-1 win is going to have their rating incorrectly go up which may lead to them being bumped up when they shouldn't be.

So I'm sure some of you are wondering, were they successful?  The answer is thankfully no, they were all bumped up to 4.5 at year-end, their Sectionals and Nationals results were good and moved their ratings back up from where they were after the One Doubles league.  Additionally, I have not verified this, but the USTA may have thrown these matches out as while the scores are still shown on TennisLink, it also seems to indicate the League Coordinator changed them to default wins for the opponents.  So it appears someone caught on and the team was DQ'd from play.  I'd hope the results were ignored for ratings purposes too give how they would skew results for others.  But should more be done?

In my view, what should happen in clear and blatant cases like this is the team/players should be DQ'd, the matches not used for ratings, and the players should be given a ban from playing in any USTA sanctioned event.  There is simply no place for behavior like this.  While I'm sure this situation is very much an exception and not the rule, it is one of the things that gives USTA League a black eye and the revenue lost from a hand full of players being banned is more than offset by keeping players that leave because of behavior like this and the improved and fair competition you'll get when these types of shenanigans are weeded out.

I do not know one way or the other if there has been any ban or reprimand at all in addition to the DQ'ing of the matches and team from the league.  I would hope so and if so, kudos to the USTA.  If not, please document rules that allow bans to be imposed in these situations so these types of players can be gotten rid of.

2014 USTA League year-end bump percentages for NTRP 3.5 broken out by section and gender - Interesting tennis league stats

I wrote about how the NTRP 4.0 level players were bumped up or down in the 2014 year-end NTRP ratings, now it is time for the 3.5s.

We can see that the men were bumped up and lot more than down, and had three sections with bump rtes over 18% and several more close behind.

The women weren't quite as extreme, but still had several sections with bump up rates over 15%.

Did this may 3.5s really improve that much?  Or is the USTA trying to weaken the level by moving all the better 3.5s out?

Monday, December 8, 2014

2014 USTA League year-end bump percentages for NTRP 4.0 now broken out by gender - Interesting tennis league stats

I wrote about the overall bump percentages for 4.0s in USTA League play this year-end, but grouping men and women together doesn't tell the whole picture.  So below are separate charts for men and women 4.0s showing the percentage of 2013 year-end 4.0s that were bumped down to 3.5 or up to 4.5.

We see some differences between the men (left) and women (right), although the top-4 sections for bump ups for both men and women are PNW, NorCal, SoCal, and Texas.  The PNW 4.0 men were given the highest percentage of bump ups by a healthy margin but the SoCal women, and even the NorCal women beat out the PNW.

The Caribbean women had the largest percentage bumped down, but the men were bumped up more. Perhaps Caribbean should be the early favorites for 3.5 women Nationals next year.

Interestingly, the Northern men had a high percentage of bump ups even though neither 18+ nor 40+ made the semis at Nationals.  They did finish second in their flight in 18+ but third in 40+ so this number of bump ups and the low percentage of bump downs for them is odd.

Saturday, December 6, 2014

2014 USTA League year-end bump percentages for NTRP 4.0 - Interesting tennis league stats

Continuing on the break down of the 2014 year-end USTA League NTRP ratings bumps, below is a chart showing how the 4.0 level bumps occurred by section.

The Pacific Northwest led the way with the most bumps up, nearly 14% of 4.0s were bumped up, but that section also had over 7% bumped down which was not the least by any stretch, so there was movement both directions.

Southern and Northern California were next and interestingly, NorCal had even more bump downs as a percentage than PNW.  Texas was next with the 4th most bumps up.

These three sections did do well at 4.0 Nationals so it sort of makes sense.  But Caribbean had a large number of bump ups overall but was well back at the 4.0 level with more bump downs than up.

Friday, December 5, 2014

2014 USTA League year-end bump percentages by level - Interesting tennis league stats

I've written about the year-end bump percentages by state and section, so the next logic step is to look at them by NTRP level.

It is probably no surprise that the lower levels have more bump ups than down, and the higher levels have more down than up.  It looks like 4.0 is where it is most balanced, about the same number of bump ups as down.

How did this compare with last year?

There were quite a few more bumps this year than last.  Last year, 5.01% of 4.0s were bumped down and 4.83% were bumped up.  This year, 7.8% down and 8.37% up.  So a greater number moved, and the bias went from slightly down to slightly up.

More to come.

The mythical double bump not so mythical anymore

In USTA League play, the double bump used to be rare.  It hardly ever occurred and when it did, it was newsworthy.  In 2014 though, while not common place by any stretch, it did happen a lot more than in the past.

I wrote a few months ago that double bumps were very rare, just 25 my count at the end of 2013, but there were a whopping 160 in 2014.  In the grand scheme of things with a few hundred thousand players, neither number is very large, but it is over 7 times larger this year than last.

Whether this was a one time event as adjustments were made in some sections this year or we will see this more often will be determined.  But it is certainly interesting to see happen.

Thursday, December 4, 2014

A side effect of all the 2014 USTA League bump ups - Early start teams are decimated

With the large number of bump ups that occurred as part of the 2014 year-end ratings, an interesting side effect has occurred.  Early start teams are being decimated, particularly in those sections where the ratings increases were especially aggressive.

As background, many sections have 2015 leagues that start during 2014.  This is often due to court availability or other scheduling reasons or wanting to spread leagues out across the year and not try to have them all starting at the same time in January-March of 2015.  Since it is a 2015 league though that will have Nationals in the Fall of 2015, it doesn't really make sense to allow someone to play with their 2013 year-end rating, but since the league starts before 2014 year-end ratings are out, what to do?

Early start ratings are the solution.  They are basically a snapshot of a player's dynamic rating as of a cut-off date that is a month or two before the early start leagues start.  In most sections this falls in the July/August timeframe for leagues starting in September and October.  This will usually identify players that are going to be bumped up or down at year-end and gets them playing at the right level for 2015.

However, early start ratings are not perfect, and sometimes players will be bumped up or down in year-end differently than they were for early start.  In the case where a player wasn't bumped for early start but is at year-end, what happens to that early start team then?

Not every section is the same.  Some (case A) force the player to play at their new level immediately.  Others (case B) let them finish the regular season at their rostered level but have to play at the new level beyond that.  A few (case C) let the player continue playing at their rostered level the entire season through Nationals.

Which of these is the most equitable is debatable, but what is clear is that with the larger number of players being bumped up, those teams that have the case A or B rules could be significantly affected.  Players are either not eligible at all, or must now play with a lower rated partner which can make it difficult for a team to have enough valid players to play a match.

But even the case C rule teams can be affected.  While players are allowed to play at their rostered level, there is a clause still promotes them to their year-end level on the team if their rating reaches the "clearly above level" (strike level).  This is arguably fair to the team's opponents, why should they have to play a "4.0" that is now a strong 4.5, but just like the case A and B teams, this can wreak havoc on a roster.

For example, I just learned about an 8.0 Mixed team that had four of their 4.0 men be clearly above level and told they have to play as 4.5s now.  The problem is there are only two 3.5 women on the team that can even play with the glut of 4.5 men.  Further, there is now only one 4.0 man left to play with the glut of 4.0 women.  This team has effectively been force to play their two 3.5 women and one 4.0 man every match.  And with the regular season over this weekend, there is no opportunity to add any players to be able to field a team more easily.  They were playoff bound but may have to pull out.

I don't know the right answer.  It doesn't seem fair to opponents to have to play out of level players.  But it also doesn't seem fair to decimate a team that played by the rules and fielded a team using the published early start ratings.  And it is hard to tell a section that they have to wait for year-end ratings and can't have early start leagues.

What do you think?

Wednesday, December 3, 2014

2014 USTA League year-end bump percentages by section - Interesting tennis league stats

I wrote about the 2014 USTA League year-end bump percentages by state a day ago.  It seemed natural to show the break down by section as well.

Caribbean and Pacific Northwest dominate again as expected, but some of the individual states get lost in their section so don't stand out as they did before.  This results in Southern California, Texas, and Northern California stepping up to fill out the top-5.  Southern is close behind and interestingly New England is next.

What sections had the least bumps?  Midwest, Intermountain and Missouri Valley lead the way.

So no big surprises given the by state data, but it is good to see it by section.

A 55+ team report customer won USTA League Nationals - Congratulations

I just belatedly noticed/learned that a 55 & over team I helped won their Nationals.  Congratulations!

Tuesday, December 2, 2014

2014 USTA League year-end bump percentages by state - Interesting tennis league stats

USTA League year-end ratings are out for 2014 and consensus is that there have been a lot more bumps up than down.  As fun as it is to look at the strange situations that we hear about or hear the complaints from players being bumped up, that doesn't necessarily represent the majority, so naturally, I had to see if the data supported this.

Without further adieu, below is a chart showing the percentage of players bump up and down by state.  Apologies for it being so tall, but with 50 states plus D.C. and a total across the league, it was unavoidable.

Update: I added a second chart sorted by bump up percentage at the bottom.

Nationally, there were more than twice as many bumps up than down, 14.3% to 6.4%.  So the feeling that there were more bumps up than down was more than true.  Now, this isn't uncommon.  Last year there were 8.2% up and 4.1% down, but this years ratio is higher, 2.24 vs 1.99, but the sheer number of players bumped one way or the other is higher this year with nearly 20% of players being bumped this year vs just over 12% last year.

So is the perception that some sections were hit harder than others true?  The chart below is by state so isn't a perfect tool to determine that, but we can get a great idea.

If we ignore North Dakota (very few players so a statistical anomaly), the state with the highest bump up percentage and highest ratio is Puerto Rico.  Ok ok, it isn't a state, but work with me.  PR had just over 22% of their players bump up and just 5.4% bumped down.

But PR didn't have the highest ratio.  That award goes to Washington which had 21.5% bumped up and 4.8% down, a ratio of 4.5 compared to 4.1 for PR.

Next was Oregon with 20.3%/6.1%/3.31 and Utah with 18.5%/4.5%/4.2.

Given how the sections did at 2014 Nationals, it probably isn't surprising to see WA and OR so high as the Pacific Northwest section "won" the 18+ division points title, tied for the 40+ lead, and finished 4th overall.  So it appears the USTA really did focus bumps on at least this section that did very well at Nationals.

But what about the others?  Texas won the overall points title and they were just 11th on the list with 15.9%/6.5%/2.5.  These numbers are above the national average but just barely so they were not singled out like PNW was.

The Northern and Southern California also did well at Nationals and California was 12th on the list with 15.8%/6.3%/2.5, very similar to Texas and not singled out.

A couple states from Southern were higher on the list, Louisiana with 17.9%/5.8%/3.1 and Tennessee with 17.6%/5.5%/3.2 and Southern also did very well at Nationals.

On the other end of the spectrum, one state did have more bump downs than up.  Ignoring Alaska (another anomaly), it was Delaware with 9.3%/9.5%/0.99.  Others with low bump up percentages include Wisconsin (7.8%/7.5%/1.03), Nebraska (8.8%/6.9%/1.27), Vermont (9.4%/6.1%/1.54), and Colorado (10.7%/7.5%/1.42).

It is interesting that Colorado was so low while Utah was so high given that they are both in Intermountain, so whatever the USTA did was not necessarily applied to entire sections.

In any case, it does appear that the USTA has in general gone out of their way to bump more players up than in the past, and to do so somewhat in sections that did well in Nationals, although PNW and Caribbean were hit harder than other sections.

Here is a chart sorted by bump up percentage, highest first.

What? No way! Why? How can I appeal my USTA League year-end rating?

With USTA League year-end ratings coming out late last night, I imagine the remarks made in the post are probably pretty common today.

It appears the USTA let playoff results carry more weight this year and had how a section's team did at Nationals affect the other players from that section a lot more than in the past.  This has resulted in sections that did well at Nationals having a lot more bump ups than normal or expected while others that didn't do well maintained or even had some players not bumped up that might have been otherwise.

Reports are that at least for the levels that won Nationals, sections like NorCal, SoCal, Texas, and PNW have been bumped up a lot, while others like Northern and Missouri Valley have been spared.

My guess is that the USTA is trying to do a better job leveling the playing field at Nationals and "punishing" those that did well by bumping more players up.  The theory is that 2015 Nationals should have more parity.  We will see how it works.

I will be doing more objective analysis of the ratings soon, but wanted to share some of the subjective observations I've made or been made aware of so far.

People that were bumped up (or down) and don't understand can always get an Estimated Dynamic NTRP Rating Report that will help explain how it all works.  And if you are interested in appealing, check out the appeal rules from last year (I think they are still the same) to see if you might be eligible to do so.

Note that if you are familiar with seeing "B" ratings, while those calculations are still done, you won't see that designation any more.

Keep the comments coming on this blog or on Facebook or Twitter.

What happened to USTA NTRP benchmark or "B" ratings?

With the release of the 2014 USTA League year-end ratings, I don't believe that you will see any more "B" ratings for ratings dated 12/31/2014.  This was announced in a newsletter I received awhile ago, although it was perhaps just a PNW newsletter and not sent to all USTA league members.

Here is what was specifically written:
Rating Type “B”: This rating type will no longer be published but process will remain unchanged.  Championship players will not be allowed to appeal their ratings down for one year after playing in Championship matches but will be allowed to appeal their ratings up if they otherwise qualify.
 So benchmark calculations are still done, they've just elected to not publish who is/isn't a B.  And they've also made it so benchmarks can be eligible to appeal up.

I'm guessing this was done because people were confused by who was/wasn't a "B" or what it meant.  So they just removed it so it isn't published.

Monday, December 1, 2014

Initial observations from 2014 USTA League year-end ratings

USTA League year-end ratings are out and here are some initial observations.

There seems to be a definite upward bias with more players being bumped up than expected, and certainly a lot more bump ups than bump downs.  Some of these bump ups are happening in sections that did well at Nationals, but I'm hearing of bump ups in other sections too.

I don't know if there was any sort of rebalancing the levels or redefinition of levels/thresholds, but it seems like it as match results alone don't appear to support some of the bumps that have taken place.  And there being so many bump ups with so few bump downs seems to support that.

What are some of the strange bump ups or downs that you have seen?

2014 USTA League year-end ratings are out - No more "B" ratings

USTA League 2014 year-end ratings have been or are being published on TennisLink.

My initial checks are that my estimates and reports are accurate for the vast majority of players, but I've missed on a few that are really baffling.

The most baffling is a 4.0 that didn't play up and went 10-2 but had two losses in playoffs that got bumped up, no surprise, but to 5.0!

I'll share more strange ones as I see them.

But one thing I should have blogged about and didn't is that while benchmark calculations are still being done, players are not going to have the "B" designation anymore.  I'm not sure why the USTA changed this, but they did and it was in a newsletter I saw a few months ago.