These are some pretty well stated arguments. At least you've come up with some ideas rather than just complaining like 90% of people.
That being said let's argue a bit:
So two counter-arguments:
#1 well, there are many folks complaining that including results from mixed or even doubles matches makes the rating unreliable. Which is why (I think) mixed results are not counted if one plays enough non-mixed matches.
#2 statistical evidence shows that many, many folks play like 3-5 matches _per year_. I believe
@schmke provided some data in the past that showed that median number of matches is 4 per year. If you do not include non-advancing matches then you will end up with so few matches that the ranking is going to be not accurate at all.
ok, I can see that. Now - it seems that for many folks hope of playing for championship with friends is what makes them join USTA league in the first place. It is going to be pretty hard to convince such a player 'well, you self rate this year, you play but can't join us if we advance, but _next year_ you will be all right. Unless of course people on the team change, get bumped, etc - in which case there's no _next year_'. So sure, that would make it perhaps 'more fair' at a cost of reducing the number of people that play USTA. Which is kind of contrary to what USTA tries to promote.
what would such anti-tanking code be?
I'm not sure if I completely follow here. First, there's that issue of folks playing ~5 matches per year so having ranking based on 8 matches is a tall order. Second, a self-rated player is subject to strikes so effectively his self-rating is evaluated already. If he is not DQ'ed - doesn't that mean that he did not rated too low?
If a player plays two mixed matches, and a combo - a league with no national championship I believe - do you really care what his rating is? I mean we are talking mixed and combo, essentially a social gathering with some physical activity.
Do you truly think that rating based on combo results is reflecting one's true ability?
can't argue too much here, I have no idea what are the grounds for non-medical appeals. I would have no problems with not allowing those. Anyone knows what the reasons for non-medical appeals are?
sure. Although one can probably see that due to climate/weather differences between sections the number of leagues/matches vary greatly - so perhaps some rules need to reflect those differences.
My counter-arguments to your comments:
1)Doesn't the USTA generate an "M" rating for those players who play 3 mixed matches? If the USTA thinks their algorithm works for Mixed Only matches, why should you doubt it can be applied to all matches? The USTA believes their M ratings to be accurate enough. Of course I don't but hey this is the USTA's poor algorithm, not mine.
2)
@schmke can correct me, but I believe the USTA will generate a year-end rating after 3 matches (3 Spring matches, or 3 Mixed matches), therefore if those 3 matches will currently generate a year-end rating. By your own statement
If you do not include non-advancing matches then you will end up with so few matches that the ranking is going to be not accurate at all.
Well the USTA year-end rating generated by the USTA publishes ratings after just three (3) matches, so my suggestion is self-rated players play three matches before any local, state, or regional championship, to generate a USTA rating which every player must get verified locally before advancing to a championship match. All it takes is the local league coordinator running an update to every Self-Rated player's DNTRP. If below the threshold, the player can advance. If above the threshold, the player is bumped up and can't play at the intial self-rating level.
3) my point is this: The USTA's year-end rating (if they publish one at all, e.g, 2020 they did not) coupled with the Appeal Down option is broken and could be tweaked. They tweaked the 40+ leagues to four courts only, with only a few years of comments and complaints. Players, captains and teams have been complaining about the USTA's rating system for years, and the changes are small, and don't stop those who want to exploit their unlocked door. The USTA has a system that can and is exploited if players/captains choose to exploit it. Does it happen to a large percentage of players? I don't know but I think no. Do I think the USTA could generate spring ratings more quickly than they currently do that are more accurate than the M only ratings? Absolutely. Do I lose sleep over anything with USTA? Ha, no. Tennis is a chance to be with some of my friends and to have some fun. I like that I get to run and generate steps, sweat, and have more fun than running.
Read the first post from the thread below and the replies by some well-respected 4.5 players. Now insert this 18-year old (whose father claims he will double bagel 4.0 adults) into 18+ 8.0 mixed with 5.0 C lady, and 8.5 Combo with a 5.0C guy.
Do I want to play those teams? No, not with a 4.0 lady or a 4.5 partner.
If this 3.5S player who double bagels 4.0 "Adults", gets to play with 5.0 C rated partners in Mixed and Combo, of course, they are going to do well. So to answer your question, "Yes, I do not care to play players like that in Mixed or Combo". And by the way, Mixed has a National Champion. 18+, 40+ at least.
By the way I hit with a 5.0 lady weekly who plays for the USTA National team. No way I want to face her and an 18-year self-rated 3.5C who will double bagel 24-year-old 4.0s.
ok, I can see that. Now - it seems that for many folks hope of playing for championship with friends is what makes them join USTA league in the first place. It is going to be pretty hard to convince such a player 'well, you self rate this year, you play but can't join us if we advance, but _next year_ you will be all right. Unless of course people on the team change, get bumped, etc - in which case there's no _next year_'. So sure, that would make it perhaps 'more fair' at a cost of reducing the number of people that play USTA. Which is kind of contrary to what USTA tries to promote.
This narrative has been brought up so many times. My solution is to have the USTA used the same logic to generate a rating before any championship matches. The only way you get excluded is to not play 3 matches.
If the many players you write about join a team in hope of playing for a championship, surely, they can be dedicated enough to play 3 matches. If the USTA's algorithm works to generate an M Rating after 3 matches, so surely the USTA could change and generate a rating after 3 matches to prevent those players who will be most adsurely bumped at the end of the year, out of championship matches when they are clearly out of level. If the "S" rated player is below the level threshold, the player can play in the championship matches. If the "S" rating puts the player over the threshold, the player can't play at the "S" rating. It's really a rather simple solution that could be implemented easily. The USTA doesn't lose any revenue or players, and it puts a check-valve on the system. Captains and players would just stop with the excuses about exclusion and my player can't make 3 matches if the USTA made such a change.
Don't worry, this will never happen as the USTA doens't care. But you asked me "would you be so kind and explain how a rating system should be made better to make manipulation impossible?"
Anti-tanking code.
@schmke and I have discussed this before. It would be rather simple to include the code in the algorithm. For example, a high-level 4.5 player (player A) on a weak team that will not advance plays singles vs an average 4.0 player (Player B), a weak 4.5 (player C) and roughly same level 4.5 (Player D). The players loses all three matches 0&1.
The matches vs player B/C/D are compared to the expected results and the actual results are beyond the threshold for a given probability that player A should lose to player B/C/D.
Player A vs Player B: is not counted as either player's rating as 1) either Player A tanked or 2) player A was so injured he couldn't move, swing, etc. Playing injured at 1/5 or 1/10 capability results in the same net: it is tanking.
Player A vs Player C: if the actual match result is within the USTA threshold bounds of their "acceptable expected match result", it counts towards both player's ratings. If it isn't, the match is not counted.
Player A vs Player D: if the actual match result is within the USTA threshold bounds of their "acceptable expected match result", it counts towards both player's ratings. If it isn't, the match is not counted.
6) Do you think counting matches in a fall league that do not advance serve any valid purpose ? Do you think players, captains don't tank matches in these leagues? This is something that could be easily fixed, but it isn't fixed.
Why? The USTA doesn't care. If the USTA doesn't care about any of these issues, so don't lose sleep and enjoy your life. You asked and I answered.
No 4.5 team will ever put him on a team. Guess he can self rate again in three years.
