What AI Can’t Do With Coaching

By Rod Heckelman

Can AI help with the most important step in coaching?

We all look forward to the advantages AI will bring to tennis instruction. It will help take us to the next step, way beyond the use of video analysis or data referencing, or any other technology we have added to our coaching in recent years. However, it remains a question as to whether it can help with what might be the most important step in coaching, and that is the recognition of the cause and the reason behind any student's shortcomings.  This may be especially true when recognizing the physical or mental attributes of each individual. As it stands right now, AI is not able to generate new ideas or solutions based on pre-existing data and rules. Also, AI lacks intuition, emotions, cultural sensitivity, pre-existing data, and rules. Most likely these issues will be eventually resolved. Still, until these important issues can be overcome, human experience will remain the most important quality any coach will have in determining the cause.

Understanding the coaching process

Let’s go back a step to understand the common coaching process better. When the tennis pro reviews a new student, in most instances they examine and analyze their approach and preparation, footwork, and stroke production. That might be a bit of a simplification of what takes place, but generally, that is the initial process. As time moves on, they will also assess their mental approach to their play and try to have a better understanding of how they strategically can best play the game with the tools they have. After a while, hopefully, they have evaluated the student’s needs, implemented their path in changing their game, and introduced a practice and playing program that will help them process these changes so that these changes become a natural part of their new game. Again, this is a simplified approach to coaching and thankfully, the abundance of teaching pros that have contributed their knowledge to the coaching industry has provided many alternative methods to achieving these goals.

As mentioned, the tools needed for this process have expanded through the years, and AI will help us take the next step. Imagine videoing a student hitting a ball and then having AI redo that video realistically depicting that student performing that stroke to their optimum level. The visual impact alone would be extremely helpful. The emotional and educational value would provide great motivation and an additional possible learning path for that student to follow. The coach and student would work together to find those paths and hopefully, productive changes would take place. But there is still something very important that needs to be addressed, and that is the cause.

Experience = recognizing the cause of problems

Finding the cause of a stroke issue is a step that separates the level of many teaching pros. It’s one thing to tell a student that they are doing something incorrectly and how they should perform correctly, but it takes a special experienced coach to be able to recognize the cause. Especially coaches who have enough experience that has provided them with a history of lessons where similar challenges with prior students have been resolved. Most experienced teachers know that every student brings to the lesson their own physical and mental characteristics. A coach will see those characteristics and engineer any changes needed with help from that library of experience. Let’s look at a few examples.

You have a student who prepares late for a groundstroke, but you have become aware of the fact that they have individual limitations and possible eye or visual issues, so they are unable to pick up the trajectory of the flight of a ball. The coach is going to have to find a new way for them to be as successful as possible taking into consideration this limitation. That is a good example of one type of physical hindrance, which can be brought on by injury, physical ability, or age. Another student has had a long-time issue with their toss when serving. You discover that when they abandon their wind-up and just serve with the racket starting the swing from the shoulder, they do much better. That may be an issue with cross-dominate behavior. They simply can’t orchestrate the combination of taking the racket back to power a swing and at the same time comfortably placing the ball gently into the correct location. This is an example of how they are neurologically wired. Again, everyone is different in how they process neurological information. As a result, the experienced coach reaches back into their library of experience with prior students and takes steps towards making the right adjustments.

Will AI be able to see that cause?

This brings us back to the one possible challenge with AI. Will it be able to see the cause? Will it be able to recognize individual physical or mental unique behavior? Are there just too many variables, essentially as many as there are people, for AI to recognize the cause?

These questions are important to ask because when a coach recognizes the limitations of a teaching tool, they are more quickly alert to find another solution. When a coach falls in love with any teaching tool, they may hesitate or even avoid, finding alternatives in their approach. Less we forget the lesson learned from the rackets that provided feedback via computer hardware in the handle. This new method of collecting data and factually understanding how the player hit the ball was at the time a great breakthrough. But it didn’t work out for many players, including the great Rafa Nadal. Let’s hope we learned a lesson from that and take that lesson into our relationship with AI. Let’s also take great pride in the underlying fact, that at the end of the day, it will be the coach's insight and experience that can never be replaced.