Talk Tennis

Talk Tennis (http://tt.tennis-warehouse.com/index.php)
-   Junior League & Tournament Talk (http://tt.tennis-warehouse.com/forumdisplay.php?f=36)
-   -   What is a "Good Win" on TRN (http://tt.tennis-warehouse.com/showthread.php?t=447649)

Tennischick 12-08-2012 04:19 PM

What is a "Good Win" on TRN
 
So what makes a "Good Win" on tennis recruiting. Does graduation year make a difference? For example, if I beat a blue chip 2 years older than me is that as good as beating a blue chip 2 years younger than me?

Alohajrtennis 12-08-2012 05:09 PM

Quote:

Originally Posted by Tennischick (Post 7053446)
So what makes a "Good Win" on tennis recruiting. Does graduation year make a difference? For example, if I beat a blue chip 2 years older than me is that as good as beating a blue chip 2 years younger than me?

Graduation year matters. Better to beat a 5 star or maybe even 4 star 2 years older than a blue chip 2 years younger than you..

dallasoliver 12-09-2012 08:38 AM

Aloha got it right, but let me comment...

Good wins are wins against players with "higher rankings" than you overall. Note that when we calculate our Tennis Recruiting rankings, we first create a master rank list that includes all players of the same gender - that master rank list includes players of all ages, graduation years, etc. Note also that only results are used in the rankings - age, etc. does not factor into the rankings.

Once we have the master ranking list, we then filter out the various graduation years to create the 7 lists for 6th-12th grade.

So... wins against players significantly above you in the master rank list are "good wins". Losses far below you are "bad losses". Almost every player has a handful of wins above and a handful of losses below - but most wins are over players ranked below them.

Addressing the original question directly...

Quote:

Originally Posted by Tennischick (Post 7053446)
If I beat a blue chip 2 years older than me is that as good as beating a blue chip 2 years younger than me?

The rule of thumb is that "older kids are better than younger kids". The No. 1 senior is usually higher ranked than the No. 1 8th-grader in our master rank list, and a Blue Chip senior is usually ranked higher than a Blue Chip 8th-grader as well. However, there is nothing to say that this is always the case... a freshman phenom that wins, say, Kalamazoo, might be ranked higher than most Blue Chip seniors. In this case, beating that freshman would be a stronger win.

So, to be precise, age/graduation year does not factor in directly - except insomuch as older kids are typically ranked higher than younger kids in our master rank list. But yes, beating a Blue Chip 2 years older is almost always worth more than beating one two years younger.


I hope this helps.

Best,
Dallas

Tennischick 12-10-2012 04:30 AM

Thank you for the clarification.

strike1 12-10-2012 01:59 PM

Thanks Dallas, that's really helpful. Are those master lists published publically somewhere? And is there an algorithm for determining exactly how much a match will help or hurt a player's ranking?

andfor 12-10-2012 05:02 PM

Quote:

Originally Posted by strike1 (Post 7056476)
Thanks Dallas, that's really helpful. Are those master lists published publically somewhere? And is there an algorithm for determining exactly how much a match will help or hurt a player's ranking?

I would think that would be proprietary. Hopefully Dallas will answer.

Why would anyone really want to bother with the trouble of figuring out what the ranking value would be for a win or loss for player x against player y? Just asking.

jmnk 12-10-2012 06:16 PM

Quote:

Originally Posted by andfor (Post 7056741)
Why would anyone really want to bother with the trouble of figuring out what the ranking value would be for a win or loss for player x against player y? Just asking.

what do you mean why? To know who to avoid.

andfor 12-10-2012 06:27 PM

Quote:

Originally Posted by jmnk (Post 7056874)
what do you mean why? To know who to avoid.

And that can be accomplished by knowing the algorithm how?

I'm just curious on this train of thought. Some may think I'm combative here, but I'm really not. I'm puzzled why a player or coach would obsess over going to that kind length to judge a good win versus a bad one, pre or post match. Isn't that counter productive to maxing the athlete's performace potential short and long-term?

jmnk 12-10-2012 08:55 PM

Quote:

Originally Posted by jmnk (Post 7056874)
what do you mean why? To know who to avoid.

Quote:

Originally Posted by andfor (Post 7056894)
And that can be accomplished by knowing the algorithm how?

I'm just curious on this train of thought. Some may think I'm combative here, but I'm really not. I'm puzzled why a player or coach would obsess over going to that kind length to judge a good win versus a bad one, pre or post match. Isn't that counter productive to maxing the athlete's performace potential short and long-term?

I'm just joking. I'm with you - I have no idea why one would try to figure out up front what a 'good win' is.

But yes, knowing the algorithm would allow you to do just that. While you can't change the draw, you could potentially do 'strategic pre-match retirement' if you do not feel up to par and you know that a potential loss would be a 'bad loss'. Similarly, if you know that upcoming match is at worst (when you lose) a not-very-ranking-changing one, and a very good win (if you prevail) - you can play with no pressure.

Those rankings are most likely based on a variation of ELO ranking where the strength of the opponent and the result are taken into consideration to come up with some kind of match value. Read up on Canada Rogers ranking - they are likely similar.

The exact algorithm is not published for the same reasons why USTA algorithm is not published. To avoid having people try to game the system.

Disclaimer: these are all just my thoughts, I have zero inside knowledge.

strike1 12-11-2012 01:28 AM

I asked the question about the algorithm because I was at a college recruiting seminar recently -- sponsored by the USTA -- and one of the speakers from a private recruiting firm said TRN's rankings are determined completely subjectively by people looking at the matches each week and simply deciding how match the win or loss should be worth. Intuitively I knew that wasn't right, but thought it would be good to hear from Dallas to put away those misperceptions.

andfor 12-11-2012 05:28 AM

Quote:

Originally Posted by strike1 (Post 7057201)
I asked the question about the algorithm because I was at a college recruiting seminar recently -- sponsored by the USTA -- and one of the speakers from a private recruiting firm said TRN's rankings are determined completely subjectively by people looking at the matches each week and simply deciding how match the win or loss should be worth. Intuitively I knew that wasn't right, but thought it would be good to hear from Dallas to put away those misperceptions.

Who ever told you that is highly uniformed. I would stay far away for anyone like that representing themself as recruiting consultant.

We were approached by a college recruiting firm and it was clear their college tennis recruiter was not a tennis player, never played college tennis, but worse knew little about the game. Lucky for my player I knew more than they did.

sbw2z 12-11-2012 08:10 AM

Not exactly sure about the details, but there is a lot of information here:

http://tennisrecruiting.net/faq/rankings.asp

... and here:

http://tennisrecruiting.net/about/RecruitingLists.asp

What does seem clear is that whoever gave you that advice doesn't know what he's talking about. There are tens of thousands of players, and they come out with rankings once a week. There is no way they sit around a table, discuss individual matches, and decide who is ranked what.

strike1 12-11-2012 12:30 PM

That's exactly what many of us sitting in the forum were thinking. It was surprising though that someone there would present such wrong information . . . And Andfor, you are right -- this group also seemed to have no particular expertise in tennis, but rather did college recruiting for all sports.

dallasoliver 12-11-2012 12:51 PM

Quote:

Originally Posted by strike1 (Post 7056476)
Thanks Dallas, that's really helpful. Are those master lists published publically somewhere? And is there an algorithm for determining exactly how much a match will help or hurt a player's ranking?

No, we do not publish the master lists anywhere today for general consumption. We do have some future features on our roadmap that will provide additional insight into our rankings, but we do not plan on simply publishing the master lists.

Best,
Dallas

dallasoliver 12-11-2012 12:53 PM

Quote:

Originally Posted by jmnk (Post 7057081)
Those rankings are most likely based on a variation of ELO ranking where the strength of the opponent and the result are taken into consideration to come up with some kind of match value. Read up on Canada Rogers ranking - they are likely similar.

The algorithm we use is actually an iterative algorithm based on the old STAR system that the USTA used back when they calculated head-to-head rankings.

STAR is very complicated - I'm not going to go into it more than to point out the FAQ entry above that another poster references.

Best,
Dallas

dallasoliver 12-11-2012 12:55 PM

Quote:

Originally Posted by strike1 (Post 7057201)
I asked the question about the algorithm because I was at a college recruiting seminar recently -- sponsored by the USTA -- and one of the speakers from a private recruiting firm said TRN's rankings are determined completely subjectively by people looking at the matches each week and simply deciding how match the win or loss should be worth.

Let's just say that the other folks on this thread are correct, and the so-called experts from the private recruiting firm are - to put it delicately - misinformed. Our rankings are completely objective. We get phone calls all the time from people asking us to give them consideration for injury or sickness or whatever... our response is always that the data is the data. We feed data into our ranking system, and the rankings come out the other side.

Best,
Dallas

Tennischick 12-12-2012 04:12 AM

One more question...how do loses effect my rating? I am a 5 star and had an off day had rwo recent loses to 3 stars. Does this effect my rating? Those are my only loses to 3 stars.

sbw2z 12-12-2012 05:38 AM

Tennischick, it could impact your weekly ranking, but your rating won't change. They only rate players twice a year. The next time will be in the spring:

http://tennisrecruiting.net/faq/ratings.asp

kme5150 12-12-2012 06:49 AM

Quote:

Originally Posted by jmnk (Post 7056874)
what do you mean why? To know who to avoid.

That was hilarious :)

ClarkC 12-12-2012 04:23 PM

Quote:

Originally Posted by strike1 (Post 7057201)
I asked the question about the algorithm because I was at a college recruiting seminar recently -- sponsored by the USTA -- and one of the speakers from a private recruiting firm said TRN's rankings are determined completely subjectively by people looking at the matches each week and simply deciding how match the win or loss should be worth.

Wow, I thought they did it all by computer. I did not realize that the people at TRN have that much time on their hands every week. :)


All times are GMT -8. The time now is 04:11 AM.

Powered by vBulletin® Version 3.6.9
Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
© 2006 - Tennis Warehouse