New format for USTA League Nationals for 2018 - What might happen?

Zman said:
It's actually possible for *seven* teams to go undefeated at this year's 40+ 4.0 men's nationals, namely Eastern, Intermountain, *******, Northern, Southern California, Southwest, and Texas.
Click to expand...

Is the word "Midwe$t" viewed as profanity in this part of the internet? Does anyone know why it gets replaced by ******* ?
 
Zman said:
It's actually possible for *seven* teams to go undefeated at this year's 40+ 4.0 men's nationals, namely Eastern, Intermountain, *******, Northern, Southern California, Southwest, and Texas. Then courts/sets/games lost would determine who advanced to the semifinals and which three undefeated teams were eliminated, even though (for instance) Eastern and SoCal have no common opponents so it isn't clear how much their numbers of courts/sets/games lost reflect their own abilities compared to how much they reflect their opponents' abilities, or whether those opponents were playing their strongest players or just giving everyone on the roster a chance to play, etc. That seems a little like breaking the tie based on who had a better record in their (completely different) local leagues. Bizarre.

If a team has three completely unbeatable doubles pairs, but can't win a singles court, then it would be a lock to win every match it plays all season no matter who it played against, so that I for one would view the team as being the best team in the country. But with the new nationals format, such a team would very likely be eliminated if more than four teams had 4-0 records after Saturday at nationals.
Click to expand...
Well said and what I've pointed out in my numerous posts on my blog.

I had not looked at the specifics of the "random" for 40+ 4.0 though, I'd only done my "equitable" random assignment in my simulation where 5 teams could finish undefeated. With 17 teams, I don't think there is a way to guarantee there can't be more than 4 without changing things after 2-3 matches are played to have the undefeated teams play each other. And if the USTA created a "random" where 7 can finish undefeated, the chances of it happening are a whole lot more likely than my analysis said (about 2% of the time, so once every 2 years or so).
 
Zman said:
It's actually possible for *seven* teams to go undefeated at this year's 40+ 4.0 men's nationals, namely Eastern, Intermountain, *******, Northern, Southern California, Southwest, and Texas. Then courts/sets/games lost would determine who advanced to the semifinals and which three undefeated teams were eliminated, even though (for instance) Eastern and SoCal have no common opponents so it isn't clear how much their numbers of courts/sets/games lost reflect their own abilities compared to how much they reflect their opponents' abilities, or whether those opponents were playing their strongest players or just giving everyone on the roster a chance to play, etc. That seems a little like breaking the tie based on who had a better record in their (completely different) local leagues. Bizarre.

If a team has three completely unbeatable doubles pairs, but can't win a singles court, then it would be a lock to win every match it plays all season no matter who it played against, so that I for one would view the team as being the best team in the country. But with the new nationals format, such a team would very likely be eliminated if more than four teams had 4-0 records after Saturday at nationals.
Click to expand...
You need to rerun the analysis assuming that Middle States will be undefeated.
 
schmke said:
With 17 teams, I don't think there is a way to guarantee there can't be more than 4 [undefeated teams after each team plays 4 matches] without changing things after 2-3 matches are played to have the undefeated teams play each other.
Click to expand...

Actually the way they ran nationals in previous years already achieved that, with one five-team round-robin and three four-team round-robins -- all you need to do now is add a fourth match for each of the 12 teams involved in the four-team round-robins (which of course will be against a team from a different four-team round-robin). There are also various other ways to achieve the same goal; one common feature of all of them is that they all have several batches of three teams A,B,C where there are matches between each of A,B and A,C and B,C. If you make the schedule at random then you won't get such batches of three teams, and it will always be possible to have more than 4 undefeated teams.

I suspect that this fourth match for each team will often be bad for the integrity of the competition, since some teams who are out of contention after three matches will put out lineups for their fourth match consisting of weaker players who came on the trip. I don't fault those teams for doing that, but clearly it's not right if one still-in-contention team gets to play against a lineup of non-nationals-caliber players in their fourth match, while another in-contention team has to play against a nationals-caliber lineup. I would be happier if the previous system were used, with one five-team round-robin and three four-team round-robins, but then have additional consolation matches afterwards for the teams that didn't win their round-robin but would like to play more tennis. I think one should minimize the role played by out-of-contention teams in deciding which in-contention teams advance.

But regardless, using the old tournament format and then adding 6 more matches to give each team a total of 4 on the first two days of nationals would be better than the current format, since at least it would ensure that every undefeated team will advance to the semifinals.
 
Zman said:
Actually the way they ran nationals in previous years already achieved that, with one five-team round-robin and three four-team round-robins -- all you need to do now is add a fourth match for each of the 12 teams involved in the four-team round-robins (which of course will be against a team from a different four-team round-robin). There are also various other ways to achieve the same goal; one common feature of all of them is that they all have several batches of three teams A,B,C where there are matches between each of A,B and A,C and B,C. If you make the schedule at random then you won't get such batches of three teams, and it will always be possible to have more than 4 undefeated teams.
Click to expand...
Oh, I agree completely. I think the quote you took from my post was a little out of context. It was about the random round-robin format. What you are describing is not a random round robin, but creating a format that just adds a fourth match for the teams in 4 team flights. You could achieve something similar by doing the random round-robin for 3 matches and then taking the winning teams and putting them into a championship bracket while having the rest play a fourth match against each other.

Zman said:
I suspect that this fourth match for each team will often be bad for the integrity of the competition, since some teams who are out of contention after three matches will put out lineups for their fourth match consisting of weaker players who came on the trip. I don't fault those teams for doing that, but clearly it's not right if one still-in-contention team gets to play against a lineup of non-nationals-caliber players in their fourth match, while another in-contention team has to play against a nationals-caliber lineup. I would be happier if the previous system were used, with one five-team round-robin and three four-team round-robins, but then have additional consolation matches afterwards for the teams that didn't win their round-robin but would like to play more tennis. I think one should minimize the role played by out-of-contention teams in deciding which in-contention teams advance.
Click to expand...
Agree with this too and it is one of the points I've made. Since the courts/sets/games tie-breakers will become so important now, who gets to play their fourth match against a team out of contention just mailing it in becomes a big deal now.

Zman said:
But regardless, using the old tournament format and then adding 6 more matches to give each team a total of 4 on the first two days of nationals would be better than the current format, since at least it would ensure that every undefeated team will advance to the semifinals.
Click to expand...
Yep.
 
It appears some flights/schedules have changed. From just looking at the names of the schedule PDFs, 18 & Over 2.5, 3.5, and 5.0+ have changed. I'll try to compare old/new to see what specifically changed, but for the 3.5 men at least I think who some teams play changed.
 
With actual teams/schedules showing up on TennisLink, I'm ramping up my simulations and will be writing more on my blog soon. I did just write about some observations I've seen so far from looking at several levels and while it is a very slim chance, I'm seeing as many as 6 teams being undefeated for several events, so the USTA didn't even arrange the schedule in a way to ensure no more than 5 could be undefeated (which would be possible to do). Read my blog for some other initial observations.

But I'm also open to which age/gender/level to simulate and write about first, any votes here?
 
schmke said:
With actual teams/schedules showing up on TennisLink, I'm ramping up my simulations and will be writing more on my blog soon. I did just write about some observations I've seen so far from looking at several levels and while it is a very slim chance, I'm seeing as many as 6 teams being undefeated for several events, so the USTA didn't even arrange the schedule in a way to ensure no more than 5 could be undefeated (which would be possible to do). Read my blog for some other initial observations.

But I'm also open to which age/gender/level to simulate and write about first, any votes here?
Click to expand...

18+ 4.5 men
 
schmke said:
With actual teams/schedules showing up on TennisLink, I'm ramping up my simulations and will be writing more on my blog soon. I did just write about some observations I've seen so far from looking at several levels and while it is a very slim chance, I'm seeing as many as 6 teams being undefeated for several events, so the USTA didn't even arrange the schedule in a way to ensure no more than 5 could be undefeated (which would be possible to do). Read my blog for some other initial observations.

But I'm also open to which age/gender/level to simulate and write about first, any votes here?
Click to expand...

Why are these idiots allowed to run the USTA when I could do a vastly better job?

J
 
schmke said:
With actual teams/schedules showing up on TennisLink, I'm ramping up my simulations and will be writing more on my blog soon. I did just write about some observations I've seen so far from looking at several levels and while it is a very slim chance, I'm seeing as many as 6 teams being undefeated for several events, so the USTA didn't even arrange the schedule in a way to ensure no more than 5 could be undefeated (which would be possible to do). Read my blog for some other initial observations.

But I'm also open to which age/gender/level to simulate and write about first, any votes here?
Click to expand...


Surely you will factor current at level players into your Nationals calculations since there are those bad eggs who were bumped post-sectionals. :rolleyes:
 
schmke said:
With actual teams/schedules showing up on TennisLink, I'm ramping up my simulations and will be writing more on my blog soon.

But I'm also open to which age/gender/level to simulate and write about first, any votes here?
Click to expand...
Yippee! My biased vote is for 18+ 3.5M. :)
 
BeyondTheTape said:
Surely you will factor current at level players into your Nationals calculations since there are those bad eggs who were bumped post-sectionals. :rolleyes:
Click to expand...
Of course! I omit those players that are ineligible to play at Nationals from the average I calculate for each team.
 
Watch out for Middle States in 4.0 40+. Our section has been runner up 3 times in the last 5 years. This is the year we bring home the gold.
 
schmke said:
With actual teams/schedules showing up on TennisLink, I'm ramping up my simulations and will be writing more on my blog soon. I did just write about some observations I've seen so far from looking at several levels and while it is a very slim chance, I'm seeing as many as 6 teams being undefeated for several events, so the USTA didn't even arrange the schedule in a way to ensure no more than 5 could be undefeated (which would be possible to do). Read my blog for some other initial observations.

But I'm also open to which age/gender/level to simulate and write about first, any votes here?
Click to expand...

How about 4.0 18 +?
 
J_R_B said:
Watch out for Middle States in 4.0 40+. Our section has been runner up 3 times in the last 5 years. This is the year we bring home the gold.
Click to expand...
I have not written up the full simulation, but it looks like Middle States is the 3rd favorite to make the semis, so they have a shot!
 
I just wrote up the first couple of simulations for the 18 & Over 4.5 men and women. Go see my blog for both write-ups, plus I wrote up a summary of my methodology for doing the simulations. I'm not allowed to post the links, but someone else can if they are so inclined.
 
schmke said:
I just wrote up the first couple of simulations for the 18 & Over 4.5 men and women. Go see my blog for both write-ups, plus I wrote up a summary of my methodology for doing the simulations. I'm not allowed to post the links, but someone else can if they are so inclined.
Click to expand...
The 4.5 All TT team would like to fill the 17th spot in your simulations! Looks like the team would have made the top 7 in your average team ratings.
 
Last edited:
Shankadelic said:
The 4.5 All TT team would like to fill the 17th spot in your simulations! Looks like the team would have made been in the top 7 in your average team ratings.
Click to expand...
Actually, the 4.5 All TT Team would have been in a tie for 4th! As long as they didn't get a bad draw, they could have been a favorite to make the semis.
 
BeyondTheTape said:
hey... @rod99

Men's 4.5 18+ Link

@schmke i appreciate your analysis, looking at head to head schedules, figuring out which section has the most "generous" luck of the draw.
Click to expand...
I just wrote on the 18 & Over 3.5 men. The 3.5 women will be done as soon as the USTA corrects an error on their schedule (they had SoCal playing 5 matches including Texas twice, and NorCal just playing 3).
 
schmke said:
I just wrote on the 18 & Over 3.5 men. The 3.5 women will be done as soon as the USTA corrects an error on their schedule (they had SoCal playing 5 matches including Texas twice, and NorCal just playing 3).
Click to expand...

3.5 M 18& Over Link

Seeing a trend in 3.5 and 4.5 for Southern Men.... the section isn't necessarily within the top 5/6 teams at Nationals in terms of top 10 overall ratings but has been given favorable blind draws.
 
Las Vegas had a really good showing at Sectionals and several local teams qualified for Nationals over the past few weekends:
Intermountain:
Women's 3.5 40+ Team
Men's 3.0 40+
Women's 2.5 18+
Men's 3.0 18+
Women's 3.5 18+ (not the same group as the 40+)
Men's 4.5 18+

I know (and have lost to severeal and beat exactly 1) a ton of these players ... hope they do well!
 
OnTheLine said:
Las Vegas had a really good showing at Sectionals and several local teams qualified for Nationals over the past few weekends:
Intermountain:
Women's 3.5 40+ Team
Men's 3.0 40+
Women's 2.5 18+
Men's 3.0 18+
Women's 3.5 18+ (not the same group as the 40+)
Men's 4.5 18+

I know (and have lost to severeal and beat exactly 1) a ton of these players ... hope they do well!
Click to expand...
Congratulations. That is pretty impressive for Nevada to pull down that many Sectionals titles. Utah and Colorado can't be too happy ...
 
schmke said:
NorCal, SoCal, what is the difference to someone back East ... :O

It should be a simple fix and I'm sure it will get corrected, I'll probably just assume I'm right on how they'll fix it and do the simulation later today.
Click to expand...
And I just went ahead and posted it.
 
schmke said:
And I just went ahead and posted it.
Click to expand...
Going back to observations on the new format, I was just looking at strength of schedule for the 3.5 women, and one team that has a low/mid top-10 average, has to play the #2, #3, #4, and #7 top-10 average teams. Needless to say, the simulation doesn't give them much chance to advance to the semis.

But this speaks to the randomness of a random round-robin. Some teams have very easy schedules, some very hard, so luck of the draw is huge since everyone is being compared with each other and not just those in a flight where everyone plays each other.
 
schmke said:
Going back to observations on the new format, I was just looking at strength of schedule for the 3.5 women, and one team that has a low/mid top-10 average, has to play the #2, #3, #4, and #7 top-10 average teams. Needless to say, the simulation doesn't give them much chance to advance to the semis.

But this speaks to the randomness of a random round-robin. Some teams have very easy schedules, some very hard, so luck of the draw is huge since everyone is being compared with each other and not just those in a flight where everyone plays each other.
Click to expand...

Is there a fair way to seed so you wouldn't end up with a soccer like pool of death?

J
 
J011yroger said:
Is there a fair way to seed so you wouldn't end up with a soccer like pool of death?

J
Click to expand...
Without resorting to groups or flights, I think it is awfully hard to guarantee every team has a "fair" schedule. You are almost certain to have something of a range or an outlier on one end and/or the other.

But now you have me thinking ...
 
J_R_B said:
Watch out for Middle States in 4.0 40+. Our section has been runner up 3 times in the last 5 years. This is the year we bring home the gold.
Click to expand...
They aren't the overall favorite, but have a very good shot at the semis, and then anything can happen.
 
J011yroger said:
Yes, because nobody else at Nationals is underrated...

J
Click to expand...
Exactly. There is no way I can equitably account for under/over rated players on any roster, and if one has some ringers it is likely at least some others do too, so I go with top-10 averages for each team, and then introduce random variability each iteration to try to account for players doing better/worse than the numbers say or having good/bad days.

Also, I have no way of knowing who will/won't show up for a team, but I do try to identify those on the roster that are ineligible (DQ, not enough matches) and exclude them from the averages I calculate.
 
You must log in or register to reply here.

Similar threads

A
New Nationals Format - thoughts?
2
Replies
62
Views
4K
schmke
schmke
schmke
First victim of the questionable tie-breakers at Nationals
Replies
24
Views
3K
J011yroger
J011yroger
schmke
2023 Nationals Week 2 Simulations Posted - 18+ 4.5
Replies
44
Views
3K
schmke
schmke
schmke
A team could have to beat the same team twice at Nationals!
Replies
16
Views
2K
schmke
schmke
schmke
2023 Nationals Week 4 Simulations
2
Replies
63
Views
5K
CiscoPC600
CiscoPC600
Back
Top