In my previous article, I discussed the problem that underrated players present in USTA league play. It is difficult to be critical of a system without being able to propose an alternative system. In this article, I hope to put forth what I believe can improve the accuracy of rating new players coming into USTA.
The new system is called Placement Matches because it consists of a number of placement matches the new player would have to submit to playing in order to receive a rating. After the initial questionnaire is finished by the new player, they would be required to play three matches (two against players of the minimum rating allowed, and one against a player one level up from the minimum rating allowed). The results of three matches against players would give the current computer algorithm a very good idea of what the new player should be rated. Part of what the current system tries to do is eliminate the human element.
Under this new system the human element is completely removed. All that goes into the rating is your tennis and the computer algorithm that USTA already has in effect. There is no subjective observation of the player, nor questions that can be maneuvered around. A new player that is given a minimum rating of 4.0 will more than likely play competitively against the two 4.0’s they play, and get beat by the 4.5. How well they do in the two 4.0 matches and how well they respond to the better 4.5 player is what will really show where they should be. If a player is given a minimum rating of 4.0 and they beat the two 4.0’s they play 6-0, 6-1 and 6-2, 6-2 and then go on to lose to the 4.5 player 6-4, 6-4, it can be suggested that the new player (although given a minimum rating of 4.0) should probably be playing at the 4.5 level. This may not always be the case, but still it gives our current computer algorithm system more data with which to work with. Going forward, if placement matches counted towards a person hidden rating (the rating we don’t see, where if it exceeds a certain threshold we get bumped up or down) the new player would be placed into a more accurate rating faster than if they were to enter it without any previous match history. The system of placement matches is simple and intuitive in that it is exactly how our league works right now at this very moment.
There are several issues that can be brought up with this system, so I’ll respond to the obvious ones. A player that really wants to be rated 3.0 will dump their 3.0 matches, and dump their 3.5 match as well. If someone really wants to get around it, they will get around the new system the same way they got around the old system. This is completely true. If someone is determined to attain a certain rating, there is no system perfect enough to stop them. This new system though provides much more resistance than the current one.
From the comfort of a chair questions can be answered with dishonesty, but this is a far cry from deception in person. Deception in person requires a far more determined unscrupulous individual. Already in this scenario we eliminate the captain who signs up a player who doesn’t know his true level. We also eliminate the honest players who answer the questions truthfully, but are better than the minimum rating given.
For the individual who is going to answer the questions dishonestly (and also dump all the matches necessary to get the rating they want) they have a much tougher road ahead. In tennis as in many things in life, like recognizes like, and so a player who has played 3.0 for years has a pretty good idea of what a 3.0 player looks like. Similarly a player who plays 4.5 knows what a 4.5 player looks like and what 4.5 strokes look like. A player can try to dump as many matches as they want, but it is an incredible challenge to appear to be something you are not in tennis. The ability to hit a ball out without making it seem obvious or clip the net without making it apparent that you’ve done it on purpose is very very difficult. Most times when a player tries to play down it shows because they’ve played the way they play for so long that’s all they know. I doubt very seriously though that someone would go to such great lengths for something so small as a half level underrate.
Another issue that could arise from the new system is the problem of getting every new player three matches. It seems tedious for every single new player to play three matches before they even start playing in a regular league. Although this is true, and I would agree it probably would be a pain, it’s a small price to pay to be part of USTA. Under the old system new players would have to go to take time out of their day to go and get rated and had a separate cost associated with it. USTA is not a free club that anyone can join. It is a club with membership costs and responsibilities. That being the case Placement Matches would be part of cost of joining our association. If we want to improve the association this would be a strong step forward.
The new system also begs the question, who will the new players be playing against. Ideally speaking we would like to get a player who has been rated at their level for a long period of time who has a win/loss ratio of around 50%. Incentives could be offered to players willing to play in placement matches, although I would hope that in the interest of better play, that players would be willing to volunteer their time to help new players get an appropriate rating. The problem of underrated players is significant enough that I feel most people would be open to the idea of helping.
Some might ask if three matches are really necessary. Chances are that for most players three matches would be a bit of overkill, but I’d rather have one system that applied to all rather than one that changed for each new player. There will be a small number of cases where three matches will be necessary. A player might have a good day or a bad day in one or two matches. There are many variables, but three matches worth of data for the computer’s algorithm to take into account should help balance out all the variables that push cases into the outliers.
I hope that this article has given you something to think about. The new system eliminates human error more than our current system, puts the new player into a real playing environment, allows the USTA algorithm to do its job, and offers more resistance to those that would try to subvert our current system thus allowing for better ratings for new players and improving the quality of our matches. While the proposed system will have challenges and issues this could be said of any system. No system is flawless but Placement Matches seems like a better alternative. All questions, comments, or suggestions are welcomed. You can send them to me via email (UnderstandingTennis@gmail.com).
Miguel Cervantes III
<p>Miguel Cervantes III teaches at Carefree Racquet Club and privately outdoors. Miguel specializes in teaching beginners, training juniors and coaching doubles. He may be reached by e-mail at <a href="mailto:UnderstandingTennis@gmail.com">UnderstandingTennis@gmail.com</a>.</p>