ELO ratings need to be used for junior rankings

[SHUtennis]

The tennisrecruiting star system needs to incorporate ELO ratings for a more accurate way of ranking juniors for college recruiting. The financially well off are at a much greater advantage- essentially a false one star increment to the players ranking. Recently from the Eastern sectional- at least 2 girls were over-rated while at least 4-5 where under-rated, making it harder for them to get the deserved scholarship. It's alarming when a girl has a higher national ranking but a worse sectional ranking & a worse head to head result. Unfortunately this is what the coaches have to base there decisions on.

 

[ClarkC]

These discussions are always the same. "Player A is ranked ahead of Player B at tennisrecruiting.net, but I think it should be the other way around."

When asked for details, either (a) other people look at the records and think that player A is ranked ahead of player B for good reasons, or (b) the original complainer cannot post names because of privacy, etc., which means the "discussion" is pointless.

 

[dallasoliver]

My name is Dallas Oliver, and I am in charge of overall content and direction at Tennis Recruiting.

I'll scattershoot and make a few comments on this thread.

- All ranking systems can be gamed. Period.

- It is much harder to game two ranking systems at the same time - particularly ranking systems that have very different properties. It is a good thing if people generally care about both USTA (points-based) rankings as well as Tennis Recruiting (head-to-head-based) rankings.

- The original poster mentions the "star system" when referring to Tennis Recruiting. Note that Tennis Recruiting displays both rankings (updated weekly) and ratings (i.e., the "stars" which are updated semi-annually). The ratings are extremely coarse rankings that are updated infrequently. We display them because they are somewhat sexy - and it allows us to use the language used in more popular football and basketball recruiting. However, those ratings do get out of date - especially this time of year when we are about to publish new ratings. College coaches understand the difference between rankings and ratings.

- We believe that our ranking system is very accurate. I invite people to send me examples of players who are ranked significantly different who are clearly ranked improperly. I had not yet seen an example where I thought our rankings got it wrong.

- Although we believe our rankings are very good, we are willing to look at other algorithms. We plan to experiment with other algorithms - including Elo - in the future.

Regards,
Dallas

 

[phinstennis]

tennis recruiting

Dallas is right. I have coached at all three major levels and found your service as a good way to judge talent. Most programs are grossly underfunded and coaches are part time. coaches cannot do the face to face research that gives you an idea on players.
One thing it gives you is contact info. When i coached DI in the mid 90'. This was the big issue. getting hold of the players to pitch the program. the accuracy of the top 100-150 does not come into play at my current position in Division II. i realistically have to go after 1,2,and three star kids. The top school in our conference only has enough money to purchase the service. That is their total budget. This certainly gives the lower starred American kids a chance at some tennis money. I specifically look at the kinds of wins over other starred players and look for trends.
Most programs are not big time and this creates more opportunity for other American kids that may not be able to follow the D I dream.
no matter what the players need to be very pro active when pursuing college tennis.

 

[JLyon]

many coaches now also use Universal Tennis Rating, but from my experience TRN is better judge as scores are correctly inputted versus UTR has many incorrect scores and the numbers do not always represent ability IMO from the players I have seen.

 

[Tennis player 43]

Universal rankings

many coaches now also use Universal Tennis Rating, but from my experience TRN is better judge as scores are correctly inputted versus UTR has many incorrect scores and the numbers do not always represent ability IMO from the players I have seen.

J Lyon, you may very well be right. However, I follow both ITF junior matches, as well as ITF womens futures a great deal. I can tell you from experience that the UTR rating hold up far more than the TRN rating. Now keep in mind, I only follow mostly the womens side, but I have found the UTR rating to be much more reliable. I enjoy the TRN rankings and they are great for giving one an idea where a player stands. They more than serve the purpose and of course only deal with US girls and boys. The UTR takes into consideration matches played around the world and has ratings for each player, not just US players.

 

[floridatennisdude]

The tennisrecruiting star system needs to incorporate ELO ratings for a more accurate way of ranking juniors for college recruiting. The financially well off are at a much greater advantage- essentially a false one star increment to the players ranking. Recently from the Eastern sectional- at least 2 girls were over-rated while at least 4-5 where under-rated, making it harder for them to get the deserved scholarship. It's alarming when a girl has a higher national ranking but a worse sectional ranking & a worse head to head result. Unfortunately this is what the coaches have to base there decisions on.

If you know for a fact that a coach uses names on one list as their only criteria for offering spots on a college team.......run.

And as far as well off kids having advantages, no $h1t. I know of no areas of life where that is not a fact. You can either look for excuses or look for opportunities. Excuses are far easier to find.

 

[JLyon]

J Lyon, you may very well be right. However, I follow both ITF junior matches, as well as ITF womens futures a great deal. I can tell you from experience that the UTR rating hold up far more than the TRN rating. Now keep in mind, I only follow mostly the womens side, but I have found the UTR rating to be much more reliable. I enjoy the TRN rankings and they are great for giving one an idea where a player stands. They more than serve the purpose and of course only deal with US girls and boys. The UTR takes into consideration matches played around the world and has ratings for each player, not just US players.

I agree UTR is great for the tracking foreign players, one issue we have seen is the inputting of College Scores is not always upto par.

I do think though the biggest judge is to look at a players record as the come into recruiting, is their ranking falling or rising? Were they one time Top 50 and now Top 150?
Motivation and love of the game are huge indicators IMO on the how a player will do in college, will they rest on their scholarship laurels and take it easy or keep pushing it to get better. Nothing worse than getting a 4* or better player who is burned out and is content to just go thru the motions once in school and on scholarship

 

[tennis5]

The tennisrecruiting star system needs to incorporate ELO ratings for a more accurate way of ranking juniors for college recruiting. The financially well off are at a much greater advantage- essentially a false one star increment to the players ranking. Recently from the Eastern sectional- at least 2 girls were over-rated while at least 4-5 where under-rated, making it harder for them to get the deserved scholarship. It's alarming when a girl has a higher national ranking but a worse sectional ranking & a worse head to head result. Unfortunately this is what the coaches have to base there decisions on.

Any good coach is not just looking at rankings.
It's a combination of many factors, rankings is just one part.

But, if you wanted to say which rankings, it's not your sectional ranking.
Most college coaches are looking at Universal Tennis rating and ITF rankings.
( no knocks against TRN, it is just going to be less effective without the cross play)

Just stay focus on your academics and tennis, academics first,
and the college will work itself out as long as you don't get stuck on ABC college and keep an open mind.

 

[ClarkC]

Most college coaches are looking at Universal Tennis rating and ITF rankings.
( no knocks against TRN, it is just going to be less effective without the cross play)

No, most coaches are not looking at rankings all that much. They are looking at head-to-head results. When a player has played 20 ITF tourneys and another has played 3 ITF tourneys, one has a lot more chance of accumulating ITF ranking points. A coach will look at the match results of both players to get an idea of who they typically beat and who they typically lose to.

TRN is a head-to-head system with head-to-head results available for easy inspection. That is the main thing that coaches care about. Parents, particularly whiny parents, are the ones who care the most about star ratings, etc.

 

[justinmadison]

TRN is a head-to-head system with head-to-head results available for easy inspection. That is the main thing that coaches care about. Parents, particularly whiny parents, are the ones who care the most about star ratings, etc.

The star ratings are a direct result of the head-to-head competition. I don't know why you support one while whining about the other.

 

[MarTennis]

If you use UTR & TRN you should have a very good idea where a player stands.

 

[ClarkC]

The star ratings are a direct result of the head-to-head competition. I don't know why you support one while whining about the other.

Because the star ratings are only updated twice a year, which Dallas has to explain on these boards about five times a year, while the head to head results are always available for inspection by anyone, including coaches.

 

[songdatennis]

Hi Everyone,

Here's the ranking before Roland Garros calculated by Elo guideline based on all ATP matches so far. What do you think? Does it look right?


1 Novak Djokovic 10,932 0 / 0
2 Rafael Nadal 10,826 0 / 0
3 Stanislas Wawrinka 10,698 0 / 0
4 Roger Federer 10,689 0 / 0
5 Andy Murray 10,565 0 / 0
6 Tomas Berdych 10,472 0 / 0
7 David Ferrer 10,456 0 / 0
8 Kei Nishikori 10,435 0 / 0
9 Milos Raonic 10,377 0 / 0
10 Grigor Dimitrov 10,372 0 / 0
11 Jo-Wilfried Tsonga 10,365 0 / 0
12 Tommy Robredo 10,325 0 / 0
13 Richard Gasquet 10,314 0 / 0
14 Kevin Anderson 10,280 0 / 0
15 Ernests Gulbis 10,261 0 / 0
16 Tommy Haas 10,260 0 / 0
17 Marin Cilic 10,240 0 / 0
18 Alexandr Dolgopolov 10,240 0 / 0
19 Julien Benneteau 10,239 0 / 0
20 Mikhail Youzhny 10,236 0 / 0

 

[songdatennis]

Hi Everyone,

Here's the ranking before Roland Garros calculated by Elo guideline based on all ATP matches so far. What do you think? Does it look right?

1 Serena Williams 9,541 0 / 0
2 Na Li 9,462 0 / 0
3 Flavia Pennetta 9,374 0 / 0
4 Maria Sharapova 9,373 0 / 0
5 Dominika Cibulkova 9,346 0 / 0
6 Agnieszka Radwanska 9,341 0 / 0
7 Ana Ivanovic 9,328 0 / 0
8 Simona Halep 9,287 0 / 0
9 Sloane Stephens 9,261 0 / 0
10 Jelena Jankovic 9,258 0 / 0
11 Ekaterina Makarova 9,248 0 / 0
12 Carla Suarez Navarro 9,218 0 / 0
13 Alize Cornet 9,192 0 / 0
14 Caroline Wozniacki 9,183 0 / 0
15 Sara Errani 9,178 0 / 0
16 Angelique Kerber 9,178 0 / 0
17 Eugenie Bouchard 9,172 0 / 0
18 Camila Giorgi 9,152 0 / 0
19 Petra Kvitova 9,138 0 / 0
20 Casey Dellacqua 9,136 0 / 0

The ranking of all 128 players participanted in Roland Garros can be found here (tab Seedings)
Roland Garros 2014 Men


Men's ranking:
Roland Garros 2014 Men

 

[ClarkC]

How do the Elo ratings compare to the ATP and WTA rankings? I don't care enough to spend the time to do a complete comparison, but maybe someone else does.

 

[sunskeeker]

Dallis Oliver, There is good evidence that many tournament directors think your ranking system is the best. I found that out when my kid was playing JR USTA circuit. Don't know if it has changed but then the draws were computer generated but they could be redone with a click. I am not sure if the directors could manipulate the individual slots. Regardless the draws at certain tournaments looked fine to the naive eye. But, if you those looking carefully would find that the home players were inevitably matched against the lower ranked players-lower ranked by Tennis recruiting rank but who had racked up USTA points by having played a ton of tournaments. Not sure if this was widespread habit. I'm taking Northeastern here. It obviously is a limited strategy since it usually only helped their players get to the 2nd round. But the reliability of it was amusing to us. And for clubs with jr players who don't really train but who take lessons to be the next generation of adult recreational players booking league time at the same club, it was the only way their players (except 1 or 2 across all ages) would make it to round 2 so it allowed the Director to save face.

 

[sunskeeker]

Also, your star system would be far sexier if you'd adorn them with fancy attire. Put some western hats on those stars! If you want them to be sexier you'll have to pretty them up.