Simulating 2023 USTA League Nationals - 40 & Over 4.0 Women

The fifth week of 2023 USTA League Nationals is upon us, and with that, the simulations continue.

I've been doing these simulations since 2018 and have found that the four semi-finalists are usually in the top six to seven projected teams, and more often than you might think, I project the four semi-finalists perfectly.  There are surprises sometimes and a mid-pack team overachieves and advances, that is why the matches are played on the courts and not on a computer.  Still, it is interesting to do these simulations and see how the projections do.

Why do these simulations you ask?  The primary reason is that the format for USTA League Nationals is now a flight-less random round-robin where each team plays four other random opponents.  This introduces significant variations in schedule strength, the possibility of an undefeated team not making the top-4, and teams vying for the top-4 perhaps not having played head-to-head and unfortunate tie-breakers being used.  The simulations aim to educate folks on how it all works and look at what may happen.  Also see this write-up for some things to know about Nationals.

This year I'll be primarily just giving the data from the simulation and not writing a bunch of words.  Most of it is self-explanatory, but the favorites and contenders will be listed in alphabetical order.  The strengths are based on top-10 averages.  If you want more details, I can do my simulation report for you.

On to the projections from the simulation, continuing with the 40 & Over 4.0 women.

Teams: 17

Chance of 3 undefeated: 48%

Chance of 4 undefeated: 25%

Chance of 5 undefeated: 5%

Chance of tie for last spot / most likely size: 73% / 3

Strongest / weakest team strength: 4.01 / 3.77

Toughest / easiest schedule: 3.97     / 3.82

Favorites: Middle States, Midwest, Southern
Contenders: Eastern, Florida, Pacific Northwest

Fringe: Caribbean, Mid-Atlantic

A very broad range here across the team strength, a full 0.24, so there are some definite favorites and just five Favorites/Contenders.  I added a Fringe category as the drop off after those two is big.  Because of all this, a relatively high chance of five undefeated (5%) and very good chance of four (25%), so a loss early may really knock a team out.

I've also started looking at my Shenanigans Score, to see if there is any questionable roster construction, and there is one team with a high score, Southwest, at 0.94 due to a lot of 4.5s in the past, but they aren't listed even as a Fringe team.  Will they surprise because of past strong players?  Other teams are in the 0.6s and lower led by Hawaii and Missouri Valley, also not even Fringe teams.

For those interested, I offer a variety of reports to make Nationals more fun and help captains prepare.  I have a Simulation Report that has all of the details of the simulation including the average ratings for each team, each team's schedule strength, the most likely record for each team, and the chance of each possible record for each team.  I also offer reports to help teams scout opponents in more detail, both a Flight Report with full roster averages, top-8 averages and played by court averages for each team, as well as full Team Reports with detailed ratings for each rostered player and stats who who plays with who and on which court and how they do together.  Contact me if interested in any of these reports.

Simulating 2023 USTA League Nationals - 40 & Over 4.0 Men

The fifth week of 2023 USTA League Nationals is upon us, and with that, the simulations continue.

I've been doing these simulations since 2018 and have found that the four semi-finalists are usually in the top six to seven projected teams, and more often than you might think, I project the four semi-finalists perfectly.  There are surprises sometimes and a mid-pack team overachieves and advances, that is why the matches are played on the courts and not on a computer.  Still, it is interesting to do these simulations and see how the projections do.

Why do these simulations you ask?  The primary reason is that the format for USTA League Nationals is now a flight-less random round-robin where each team plays four other random opponents.  This introduces significant variations in schedule strength, the possibility of an undefeated team not making the top-4, and teams vying for the top-4 perhaps not having played head-to-head and unfortunate tie-breakers being used.  The simulations aim to educate folks on how it all works and look at what may happen.  Also see this write-up for some things to know about Nationals.

This year I'll be primarily just giving the data from the simulation and not writing a bunch of words.  Most of it is self-explanatory, but the favorites and contenders will be listed in alphabetical order.  The strengths are based on top-10 averages.  If you want more details, I can do my simulation report for you.

On to the projections from the simulation, continuing with the 40 & Over 4.0 men.

Teams: 17

Chance of 3 undefeated: 33%

Chance of 4 undefeated: 9%

Chance of 5 undefeated: <1%

Chance of tie for last spot / most likely size: 88% / 4

Strongest / weakest team strength: 4.10 / 3.88

Toughest / easiest schedule: 4.03 / 3.93

Favorites: Intermountain, SoCal, Texas
Contenders: Middle States, Midwest, NorCal, Southern

The range of strong to weak is pretty big, and the strongest teams got reasonably easy schedules, so three favorites and just four contenders.

I've also started looking at my Shenanigans Score, to see if there is any questionable roster construction, and there are some high scores!  Three teams, led by Texas at 1.1 are over 1.0, with Middle States and New England nearly matching them.  Each is largely driven by having players above 4.0 in the past.  This isn't uncommon in 40+, but each has double digits in this category.  Intermountain and Midwest are in the 0.8s and a full four more in the 0.7s.  Even if everything isn't by the book, the teams appear to be equally matched by this measure.

For those interested, I offer a variety of reports to make Nationals more fun and help captains prepare.  I have a Simulation Report that has all of the details of the simulation including the average ratings for each team, each team's schedule strength, the most likely record for each team, and the chance of each possible record for each team.  I also offer reports to help teams scout opponents in more detail, both a Flight Report with full roster averages, top-8 averages and played by court averages for each team, as well as full Team Reports with detailed ratings for each rostered player and stats who who plays with who and on which court and how they do together.  Contact me if interested in any of these reports.

How often are 2-2 ties happening at 2023 USTA League Nationals?

We've had three weekends of play at Nationals where the 40 & Over division has had events, and with the 4-court format used for 40 & Over, that means 2-2 "ties" are happening.  There are established tie-breakers when the team match ends in a 2-2 tie so that all works, but I still think it is silly that the 4-court format exists and thankfully it appears to be going away next year!

But since it is used this year, it is always interesting to look to see how many matches end in 2-2 ties, and how far down the tie-breaker rules matches go to be decided.

As background, see my prior analysis on 2021 and 2022 matches using this format where we see a pretty consistent trend of around 26% of team matches ending in 2-2 ties, around 10% of all of them being tied on sets, and around 1% of all matches being tied on games and being decided by the court one winner.

So far, the 3.0, 3.5, and 4.5 levels of 40 & Over have played with the 4.0 level to be played this coming weekend.

Overall there have been 226 team matches played, and 62 (27%) have ended in 2-2 ties.  Of those 62, 43 were decided by the sets lost tie-breaker leaving 19 (8.4%) to be decided by the games lost tie-breaker, and 18 of those were, with just one (0.44%) having to be decided by the winner of court one doubles.  These numbers are generally consistent with past years, although the number having to be decided by the court one doubles winner is a little lower than past years.

One level/gender stands out as the one with the most ties this year and that was the 4.5 women round-robin where 14 of the 34 team matches (41%) were tied 2-2, five were tied on sets (15%) and the one tied on games  (3%) happened here.

Interestingly none of the 2-2 ties for the 4.5 women were in the semis/3rd-4th/final, but the other levels/genders had some:

• 3.0 women - The final was a 2-2 tie decided on sets.
• 3.0 men - Each semi and the final were 2-2 ties!  Each one decided on sets.
• 3.5 women - One semi, the 3rd-4th, and final were 2-2 ties, the final decided on sets.
• 3.5 men - Both semis and the 3rd-4th were 2-2 ties.
• 4.5 men - 3rd-4th and the final were 2-2 ties, the final decided by 2 games!

 That is 12 of 24 or 50%, so the matches are more closely matched than regular season for sure, but also round-robin matches it appears.

We'll see how this weekend's matches go!

What do you think, are you sorry to see the 4-court format go?  Did you like the excitement of counting sets and games? 🤣

How to fix the USTA League Nationals format so an undefeated team isn't sent home

The USTA has been using an un-flighted round-robin format at Nationals for a number of years now, and for regular readers of my blog, you know I've written about a number of things that are issues with the format.  There are some positives of the format too of course, but it would be great if some steps were taken to mitigate the issues.

One of the big issues is that the format inherently allows for there to be more than four 4-0 teams, but just four teams advance to the semis, so it is possible for a 4-0 team to be sent home.  All teams should know this going in and take steps to give themselves the best chance at winning the standings tie-breaker between the 4-0 teams, but it would still be really tough if a team goes 4-0 and doesn't advance.

While the chances of it happening are usually very very small, it is always possible and in fact has happened at least twice.  Earlier this year at the 55 & Over National Invitational, the 7.0 flight had five 4-0 teams and Eastern was tied with Southern and Midwest at 4-0 and 8-4 on courts, but lost more sets and missed out on the semis.

Then this weekend at 40 & Over 3.0 Nationals it happened again with Texas being 4-0 and 12-4 on courts, but PNW at 4-0 and 13-3 on courts beat them out for 4th.

And just a few weeks ago it very nearly happened and came down to the last match where Southern lost a 2-2 match where winning a match tie-break or a few more games would have resulted in five 4-0 teams.

You might say, it has only happened twice, what is the big deal?  But tell that to that Eastern and Texas teams it happened to.  And it has very nearly happened several other times and only upsets prevented it from happening.  And it has happened numerous times at Sectionals events (also here and here), although this is largely on the section for implementing the format poorly.

What is the solution?  There are several that could be considered.

One would be to go back to four flights where just the flight winners advance.  By definition, you can only have one undefeated team in a full round-robin flight, but doing this would lose all the benefits of the un-flighted round-robin format, namely not having flights with different numbers of teams, all teams getting an equal number of matches (four), and a team that loses early still having a real chance to advance.  I don't think I'd recommend doing this.

Two would be to introduce quarterfinals so you could take the top-8 teams in the standings.  There is no way you can have nine 4-0 teams with the format, so this would guarantee an undefeated team gets to advance.  It would also address some of the inequities with the standings tie-breakers that are used, as they really don't work well when the teams being compared may not have played any common opponents can have significantly different strengths of schedule.

The challenge with this is it is adding a match to the schedule meaning teams would have to play three matches on one of the days, or the final is played on a 4th day (Monday) of the event.  Now, I'm guessing teams would prefer to play an extra match or stay an extra day to avoid the scenario of an undefeated team being sent home.  And many teams may stay through Monday and not rush out on Sunday anyway.

Three would be a variation of quarter finals where instead of taking a full 8 teams, there is a provision to have a play-in game between the 4th and 5th 4-0 teams if necessary to determine the 4th seed to go to the semis.  This would only be done if required, so would only add a match on the off chance this occurs, and while it would put these teams at a disadvantage having to play an extra match, I think it is still more equitable than sending a 4-0 team home.  Teams would know in advance they don't want to be in the 4th/5th game and try to take steps to avoid it.

A fourth option would be to redo the schedule after three matches in order to ensure undefeated teams play each other so it is decided on the court.  For example, in the case that happened this weekend, with five 3-0 teams and none of them playing each other, the schedule gets modified so at least the 4th and 5th teams in the standings play each other so you are guaranteed no more than four 4-0 teams.  Logistically this might be hard to do, but it would address the issue without adding days or matches.  One could also debate the fairness of only doing this for the 4th/5th teams and letting the other 4-0 teams play potentially weaker opponents so perhaps the redoing of the schedule need to include all the teams.

A fifth option would be to construct the schedule in a way to minimize the chance of it happening.  Today, the schedule is seemingly random which is great in that there is no bias, but given the large variation in team strength at Nationals it means some teams can end up with easy schedules and others with hard schedules which is not terribly fair when the teams are ultimately compared in standings using wins, courts, sets, games.  And in fact, we can see that this weekend where the five 4-0 teams happened, my simulation said there was a very good chance of it happening for this very reason, five strong teams got some of the easiest schedules, which greatly increased the chances of it happening.

If instead of a random schedule the teams were seeded or a strength identified like my simulations do, and then the round-robin schedule done in a way to have as equal schedule strength as possible, the chances of more than four 4-0 teams would go way down.  There would still be a chance of course, but if this was done then perhaps just the third play-in option could be used in the rare case it happens.

Sixth and last, like the fifth option this doesn't completely solve it but would reduce the chances of it happening, and that is to implement some of my suggestions to get rid of wildly out of level teams at Nationals.  As noted in the fifth option, the chances of more than four 4-0 teams is high because some teams are head and shoulders better than others, and if they get favorable schedules it is likely they all win.  Well, why are some teams head and shoulders better?  It is a combination of things that my suggestions would address by lowering strike thresholds, making more players subject to strikes, preventing tanking, and more.  If the teams are simply more similar, it becomes less likely a lot of teams go 4-0.

What would I do?  I personally think several of my suggestions in the sixth option need to be done regardless and that would help.  That alone may help make fairer schedules, but I think my fifth option needs to be done too to make the schedules more even.  After that, I'd vote for having quarterfinals, but would also support doing the play-in game.

What do you think?  What is the solution to this problem?  Or do you think it isn't a problem?

2023 USTA League Nationals Week 4 Recap - A self-rated team wins, a 4-0 team misses the semis

Week 4 of USTA League Nationals is complete, here is a recap.  You can also see my recap after day 2 here.

This was the busiest week so far with 8 events going on.  And there were some notable events!

The most notable was that we had a team that went 4-0, didn't lose a match, and didn't advance to the semis!  This happened because five teams in the 40 & Over 3.0 men's event went 4-0 and only four can go to the semis.  The original simulation said it was a real possibility, then after the first day the chances went way up, and then it happened.  The team that got the short-end of the stick was Texas who finished in 5th.  In the semis we had two 2-2 ties, errr wins, with PNW beating Middle States due to one fewer set lost, and Southern beat NorCal also with one fewer set lost.  The final went to PNW in yet another 2-2 tie/win where they also lost one fewer set.

The 40 & Over 3.0 women was not quite as exciting until the final as Texas beat SoCal 4-0 and PNW beat NorCal 4-0, but in the final we had another 2-2 tie/win with Texas losing one fewer set.

The 18 & Over 4.0 women had its own form of excitement as Texas beat New England in one semi and Florida beat Mid-Atlantic in the other.  The final had Texas win 4-1.  What was the excitement?  Texas had a small roster, just 10, but the interesting thing is that 9 of those were self-rated.  Self-rated players at Nationals are not uncommon, but large numbers of them are usually at lower levels, not 4.0.  Everything could have been by the book and fair, but a team with that many self-rates winning Nationals makes you scratch your head.

The 18 & Over 4.0 men was relatively normal with Middle States beating Eastern and Southern taking out Florida to meet in the final where Southern won it all 3-2.  What is semi interesting is Southern has a fairly high shenanigans score indicating they had a lot of self-rates and/or players who had been 4.5s in the past.

[Update: Upon reviewing the Southern team, while the score was reasonably high, I saw no indication of any tanking, and self-rates didn't appear to manage scores, i.e. no shenanigans.  It appears to be a team that did it by the book.]

The 55 & Over events got going this weekend the 6.0 women saw MOValley beat SoCal and Northern beat Midwest, before MOValley won the final.

The 6.0 men was interesting in that they show SoCal in 4th, but 5th place Midwest in the semis.  Looking at the matches SoCal defaulted a court every match so must have been ruled ineligible to advance for that reason.  Normally teams must bring the minimum number of players, but I've heard of extenuating circumstances causing the shortage of players being a reason a team is allowed to play but they aren't eligible to advance.  The teams that did play had Midwest beat Eastern and Northern beat NorCal before Northern won it all.

The 8.0 women saw Texas beat Middle States and Florida beat Midwest before Florida won it all, all the matches 2-1 results.

Last, the 8.0 men saw Mid-Atlantic beat Southern and SoCal beat Florida with SoCal winning it all, again all the results were 2-1 so very close.

Congratulations to all the new champs!

What ties did we have for week 4 of 2023 USTA League Nationals? A 4-0 team sent home!

The round-robin for week 4 of 2023 USTA League Nationals is complete and we have eight sets of semi-finalists.

The 18 & Over 4.0 women had two 4-0 teams in Texas and Florida, then a 5-way tie at 3-1 for two spots.  Mid-Atlantic and New England went 14-6 and 13-7 on courts which was better than Southern and Middle States both at 12-8.  I noted in the preview/simulation that Texas had 9 self-rates and they came through to be the #1 seed perhaps bringing into question the validity of those self-ratings.

The 18 & Over 4.0 men also had two 4-0 teams in Middle States and Southern, then a 4-way tie at 3-1 with Florida and Eastern getting the spots with better court records than SoCal and Hawaii.

The 40 & Over 3.0 women also had two 4-0 teams in Texas and NorCal, then a 4-way tie at 3-1 with PNW and SoCal having better court records over Florida and Midwest.

The 40 & Over 3.0 men is where things got interesting with five 4-0 teams and Texas being sent home without a loss with the worse court record despite having a tougher schedule than those who did advance.  The teams going on are Middle States, NorCal, Southern, and PNW.  See this for some suggestions on how to fix this.

The 55 & Over 6.0 women continued the trend of two 4-0 teams in MOValley and Midwest, then a 5-way tie at 3-1.  Northern had the best court record, but then SoCal and NorCal were both 9-3 on courts and SoCal got the spot having lost fewer sets.

The 55 & Over 6.0 men is an odd one as yet again there were two 4-0 teams in Eastern and NorCal, then a 4-way tie at 3-1 with Northern and SoCal listed as in 4th due to better court records than Midwest and PNW, but SoCal is not listed in the semis draw!  Instead Midwest is listed.  Looking at SoCal's matches, I see they defaulted a court every match (and still were in 4th!) which Nationals tries to avoid, so I have to think something happened due to that.

The 55 & Over 8.0 women also had two 4-0 teams!  Here it was Texas and Midwest, then Florida and Middle States were 3-1 with better court records than SoCal and Southern.

The 55 & Over 8.0 men broke the trend and had three 4-0 teams in Southern, SoCal, and Florida.  Four teams were tied at 3-1 for one spot, and Mid-Atlantic got it over NorCal despite both being 3-1 / 8-4 / 18-10.  This tie-breaker came down to games lost where Mid-Atlantic was better.

Congratulations to all the teams, good luck on Sunday

Well, it happened - An undefeated team misses the semis at USTA League Nationals!

Ever since the new format for USTA League Nationals came out, I've written about the possibility of more than four undefeated teams leading to a team without a loss missing out on the semis and being sent home.  It very nearly happened two weeks ago, an upset in the last match prevented it, and it has nearly happened several times at prior year's Nationals.

And in fact it has happened at a "National" event in the past, it was just 55 & Over Mixed 7.0 in April of 2022 at what is technically an "Invitational", but it uses the same format of Nationals so it effectively is a Nationals.  At that event, Eastern was 4-0 and tied on courts with two teams ahead of them, but lost more sets so lost the tie-breaker.

Well, it happened again, and this time at "real" Nationals, the 40 & Over 3.0 men's event.

I wrote in my simulation of the event that there was perhaps the largest chance of it occurring we'd seen, 16%, and after one day my updated simulation said the chances were now 34% as the five 2-0 teams didn't play each other on Saturday, so if they all just won, it would happen.

Each team won their 3rd match and each one faced an 0-3 or 1-2 team in their final match, so there was a strong probability it would happen at this point and it did.

The final standings were:

Middle States - 4-0 / 15-1
NorCal - 4-0 / 14-2
Southern - 4-0 / 14-2
PNW - 4-0 / 13-3
Texas - 4-0 / 12-4

The tie-breakers were clear, the 4-0 teams that were 13-3 or better on courts advanced, which mean Texas who won each of their matches 3-1 would not go the semis.

There are a few unfortunate things here.  The main one is that the format allows for a team to go to Nationals and not lose a team match, but not advance.  That has to be tough for the team and I hope they knew it was a possibility and this wasn't a surprise.

But another is that the criteria used to determine which team doesn't advance isn't exactly equitable.  Yes, Texas had the worst individual court record of the five teams, but the format used for Nationals doesn't have teams play the same opponents, and it is even possible that the teams tied at 4-0 don't play any common opponents.

This means one team could have a really tough schedule and eke by going 4-0 while another has a really easy schedule on their way to 4-0.  The team with the tough schedule will likely have more close wins and a worse individual court record, and as a result fare worse in the tie-breaker and not advance.

In this case, Texas played 1-3 Southwest, 2-2 Intermountain, 1-3 New England, and 1-3 Midwest, a combined 5-11.

PNW just ahead of them played 1-3 New England, 3-1 Caribbean, 2-2 Florida, and 1-3 Eastern, a combined 7-9, perhaps a tougher schedule.

But Southern got 0-4 Hawaii, 1-3 Midwest, 0-4 SoCal, and 1-3 Southwest, a combined 2-14!  This is clearly an easier schedule than Texas and likely contributed to Southern's individual court record being 14-2 to Texas' 12-4.

And NorCal got 0-4 SoCal, 2-2 Florida, 0-4 MOvalley, and 0-4 Hawaii, also a combined 2-14, a very weak schedule.

Middle States opponents were 4-12, so a bit tougher than Southern's and NorCal's, but not exactly tough.

My preview highlighted the schedule strengths to be:

• Middle States - 3.05
• NorCal - 3.05
• Southern - 3.08
• PNW - 3.05
• Texas - 3.07

So NorCal, Middle States, and PNW had easier schedules than Texas.

If we look at who actually played, the played schedule strengths were:

• Middle States - 2.99
• NorCal - 2.99
• Southern - 3.03
• PNW - 3.05
• Texas - 3..05

Here Texas had the toughest schedule, NorCal and Middle States the easiest.  So there is something to the theory that Texas may have missed out on the semis simply because they had the toughest schedule.

This makes it especially tough for a team to lose out by one court when they had one of the two toughest schedules and the other teams they were in a tie with had easier schedule.

With a random round-robin schedule, it is hard to create similar schedules, but I'm not sure any attempt is made to make more of an equitable schedule and perhaps there should be.

To be fair, I've written about this for years and I have to hope everyone involved knew of this possibility going in, so everyone knew the rules and the possibility of this and what they may need to do to win the tie-breaker.  If so, them is the rules everyone is playing by so you have to accept it.

Stay tuned, I'll be posting my suggestions on how to fix this soon.  But what do you think?  Is it a problem that a 4-0 team can miss out on the semis?  Are the unequal schedules an issue?  Or everyone knows the rules going in so it is what it is?

Update: See my suggestions on how to avoid this happening again.