OnTheLine
Hall of Fame
Although I understand your frustration and your point, obviously the computer disagreed with you. The players he beat were not as low as you thought.
And kudos to you for not encouraging him to manage his scores
USTA DQ
- Thread starter ACTG
- Start date
Although I understand your frustration and your point, obviously the computer disagreed with you. The players he beat were not as low as you thought.
And kudos to you for not encouraging him to manage his scores
Was the doubles player the first match? This is one of the unfortunate side effects of a single DNTRP rating. The same thing happened in Middle States last year. A 4.0S rated guy signed up for a team and in the first match, he played the captain of the other team, who is a 60 year old vet of decades of USTA league play who is still a solid double player but who struggles at singles now. I guess he didn't have singles players for this match and decided to take the bullet himself and lost 6-0 6-0. If his "true" doubles level is 3.8 and singles 3.4 or something and he plays 90% doubles, then his overall rating will be around 3.75 and a 6-0 6-0 match is clearly a strike against a 3.75 rating, but may not be against a 3.4 rating. This was the guy's first match and he never recovered from that first big rating and was eventually DQ'd. In the Middle States case, then guy in question really was borderline 4.0/4.5 but probably wouldn't have been DQ'd if he hadn't caught a doubles specialist in singles or even if the matches had just occurred in a different order (i.e. playing the doubles guy after he had an established rating instead of first).
schmke
Legend
Perhaps caught a bit playing a good doubles player in singles, but I wouldn't say any of his opponents were lower end 4.0s. The lowest rated player was 3-2 on the year and squarely in the middle of the 4.0 range. And the "doubles" player was 1-0 in singles last year with a 3.89 match rating so certainly not that bad a singles player, albeit a single data point.
Whether the strikes were "right" or "wrong" I have no idea, but my ratings show the last one was very close, but the other two pretty clear strikes. Remember there are "true" 4.5s that are in the 4.0x range, and I'd be interested to see how your player would do against them. I hope he doesn't quit USTA and gives it a go and we can see.
Vox Rationis
Professional
The computer can disagree but that doesn't always make it "correct". I think this situation happens more than people realize. Tennis has such wide range of results and match ups matter too much. It's too easy for a likely good 4.0 to play some weaker ones, win convincingly and get DQ'd despite being worse than the other good 4.0s in the area. It's hard to believe the system succeeds in it's goal of properly rating people in these situations. But it's not going to get everything perfect.
I admit, I am a bit frustrated. He's a nice guy and I just don't think he's a 4.5 level player. I think he got some bad luck. The guy that mainly plays doubles is a much better doubles player but the ratings have no way to distinguish that. Also the first guy he beat must have taken a couple of years off. He has prior history in 2017 at the 4.5 low level and nothing since. I watched that match and the guy he played looked like at one time he had the strokes but couldn't move that well or put things together. So it's more likely that guy was out of tennis the past two years and just started to get back into it but his rating was still above 3.75. Again, the computer has no way to factor that in either. If he had played quality 4.0's at the top of their game then maybe the matches would have been close and win or lose he would not have gotten DQ'ed. I hope he'll find a team to keep playing USTA but I don't think he will.
Last edited:
schmke
Legend
Fair point on the first opponent, he'd taken 2018 off so hard to say if his rating on the books was representative or not. But this can go both ways, sometimes a player improves significantly during the non-league season and your player would have benefited from playing an underrated player.
But there are 3 strikes so a single "bad luck" match isn't the death knell for a player. It sounds like your player may have had back to back "bad luck" scenario matches and perhaps just the third was a legitimate/accurate result. Unfortunately, sometimes that happens.
I agree. I'd rather see them trying to do something with self rates than nothing at all. Who knows, maybe against a solid player he would have picked his game up even more and would have won easily. It's just impossible to know only watching the guys he played against.
Last edited:
Vox Rationis
Professional
For his sake I hope he's able to hold his own at 4.5 and have fun. Maybe he's better than everyone realizes.
Matthew ATX
Semi-Pro
What is the match rating threshold for a strike?
kylebarendrick
Professional
It varies by level, but generally believed to be about half way into the next level.
Sent from my Pixel 2 using Tapatalk
For 4.0, it's around 4.20, which is not just barely above level but rather well into the "solid 4.5" range. The tolerance is even greater at lower levels where large in-season improvement is more likely.
schmke
Legend
Yes, particularly at lower levels a player is really "clearly above level" before they get a strike. For example, a 2.5 has to nearly be playing at a 3.5 level before they get a 2.5 strike. Similarly a 3.0 has to be playing closer to the 4.0 level than 3.0 to get a strike. The USTA really does give ample room for players to naturally improve, arguably too much room to be fair to computer rated at-level players that have to play against these improving players.
Matthew ATX
Semi-Pro
Interesting. I guess (not that didn't already know) TR isn't super accurate then.
I have four matches over 4.2 listed on there.
I have four matches over 4.2 listed on there.
1. Correct. TR is not super accurate a lot of the time.
2. Are you looking at the match rating or the rating? It's the rating (the USTA DNTRP that includes averaging with prior ratings) that is used to determine strikes.
Cindysphinx
G.O.A.T.
My 3.5 ladies 40+ team had 12 matches (over a 3 month season)
My 3.5 ladies 18+ team had 14 matches (over a 2.5 month season)
My 7.0 mixed 18+ team had 15 matches (over a 2.5 month season)
My 7.0 mixed 40+ team had 12 matches (over a 2.5 month season)
Matthew ATX
Semi-Pro
I was looking at match rating. How would individual strikes work if they're going by your overall dynamic rating? Say one match puts my overall rating above the threshold at 4.2; then my next two matches keep it static. Is that three strikes?
Regardless, I'm not in any danger if it's going by overall rating.
Last edited:
schmke
Legend
This is well above average. When I looked at the average number of matches per team a few years ago, 18+ teams average 8.72 matches per season and 40+ 8.34.
Cindysphinx
G.O.A.T.
Yup. I like that we have so many matches. We can have big teams, more depth, and more protection against injury. It's also easier to get your weak players into the rotation without risking a loss. You can play them just 3-4 times against the weakest time.
OnTheLine
Hall of Fame
It is how condensed they are that is impressive to me .... 15 matches in 75 days means you are playing more than 1 time a week
And the number of matches is nice
In contrast our mixed 40+ is run as a weekend challenger .... 7 matches condensed into Friday night - Sunday ... winner of that weekend goes straight to sectionals ... fun weekend but doesn't feel like "league"
Yes. The use of the DNTRP instead of match rating makes the strike probability "sticky" in both directions. That is, if your DTNRP is far enough over the threshold, then a match where the match rating is not over the threshold may generate a strike and if your DNTRP is below the threshold, a match with a match rating in the strike range may not generate a strike. Of course, the latter is a much more likely case (i.e. that the person's DNTRP creeps up as they get match ratings in the strike range but doesn't become a DNTRP strike for a long time), but it also emphasizes the importance of the first two matches for a self-rate when there is no existing DNTRP to average with.
Consider these two cases. First a 4.0 S-rate plays his first two matches at a 4.3 (strike level) and then plays at 4.0 for every subsequent match:
No Match DNTRP
1 4.3 4.3
2 4.3 4.3
3 4 4.2
4 4 4.2
5 4 4.175
6 4 4.14375
7 4 4.129688
8 4 4.112109
9 4 4.096387
10 4 4.084546
The first two matches are obvious strike, but on match 3, he strikes out on a match that is not even above level, and, in fact, it would take until match 5 before his DNTRP falls back below strike range.
Second, a player who plays his first two matches at 4.0 (below strike level) and then 4.3 at every subsequent one:
No Match DNTRP
1 4 4
2 4 4
3 4.3 4.1
4 4.3 4.1
5 4.3 4.125
6 4.3 4.15625
7 4.3 4.170313
8 4.3 4.187891
9 4.3 4.203613
10 4.3 4.215454
11 4.3 4.22674
For this player, the first two matches are fine, but then he plays 5 matches in the strike range before the DNTRP finally catches up and gives him a strike at match 9, his SIXTH with a strike range match rating, and finally DQs him at match 11, his NINTH straight match with a strike match rating.
So, the first player strikes out after 2 matches with match ratings in the strike range and one at level, whereas, the second player strikes out after 2 matches at level and NINE in the strike range. This is a flaw in the system.
Cindysphinx
G.O.A.T.
Yeah. It's one reason why we can't have matches on a set day of the weak.
I remember one year we had 18 matches in the 2.5 week season, which was too much. We sometimes had matches on back-to-back nights. The current league coordinator is trying very hard not to let that happen nowadays.
Three strikes DQ, not dynamic DQ - totally different
Vox Rationis
Professional
Dynamic DQ is the three strikes DQ. They are one and the same thing. The other type of DQ is a self-rating grievance, which A-rated players are NOT subject to (except under very extreme circumstances).
Last edited:
Similar threads
