Recently I began reading the book "Range" which explores the concepts of learning and intelligence in the context of sports, reinforcing the idea that sports provide a valuable framework for understanding the world.
Learning Environments & Basketball IQ
In "Range," the book emphasizes the value of abstract thinking and problem-solving over specialization. It differentiates between two types of learning environments:
Kind Learning Environments:
In these settings, individuals receive immediate and accurate feedback, allowing them to improve through specialization. For instance, in sports like golf, a player can adjust their swing immediately after a sliced shot, thanks to clear feedback.
Wicked Learning Environments:
These environments provide confusing or incorrect feedback. Specializing in such environments can lead to the reinforcement of bad habits.
In the context of basketball, this concept relates to basketball IQ. An example is players who frequently jump passing lanes to steal the ball. When successful, they often score two points on a fast break dunk. However, unsuccessful attempts can result in various outcomes: missed open shots (rewarding the behavior), made shots (punishing the behavior), or defensive rotations that disrupt the team's defense. While there may be no immediate consequence, analytics reveal that over a season, the team loses points due to these risky plays. This is why some players often lead stat-lines in stats like steals despite being labelled by analytics as bad defenders. This illustrates the challenges of wicked learning environments.
Can AI Replaces Coaches?
The book also delves into the intersection of AI and chess, which got me thinking about how this relates to strategic thinking in basketball and the concept of automation.
Chess, with its vast number of unique outcomes, has a rich history intertwined with AI development. Initially, it was believed that computers could never surpass human players. However, in the 1990s, a computer employing rule-based calculations, primarily based on previously recorded moves, achieved the feat of defeating top human chess players. This marked a significant turning point in the relationship between AI and chess.
Subsequent breakthroughs occurred with the advent of Deep Learning, a subset of machine learning. While we won't delve into the technical details, Deep Learning in this case allowed computers to teach themselves how to excel at chess. This innovation led to the emergence of chess engines capable of generating tactics that were previously unknown to humans. It was a pivotal moment that sparked widespread discussion about AI's potential. It demonstrated that Deep Learning could power complex computations, comparable to human abilities, such as constructing sentences (ex. ChatGPT) and making strategic moves in chess.
In the context of AI's application in sports, it's evident that the integration of AI and sports analytics has advanced significantly. Currently, AI is employed to generate intricate metrics that go beyond traditional measures like Player Efficiency Rating (PER). These advanced metrics assess the impact of every micro-movement on expected points per action. It's worth noting that these highly valuable metrics are not publicly available, and their usage varies across sports organizations.
Can We See AI in Sports, the Same Way It Is In Chess?
However, the next frontier in this AI-driven evolution is the use of artificial intelligence to develop game plans and strategies. Achieving this milestone is a complex endeavor, and it might take considerable time to reach fruition. In sports like basketball, particularly the NBA, the incentives to build innovative, winning operations might not be as strong due to already high team values. However, in sports like soccer, there's immense potential. A struggling soccer team can undergo a substantial transformation by implementing cutting-edge technology. For instance, a French soccer team's value could increase from tens of millions to hundreds of millions through such innovations. In fact, there are private equity firms specializing in investing in struggling teams and creating their own advantages through data to boost commercial value, due to the importance of winning.
In other words, the opportunity cost for NBA teams may not justify the extensive investment required. However, in European soccer, this evolution holds significant promise. The difference between winning and not winning in European soccer leagues can translate into tens or even hundreds of millions of dollars in commercial value.
In my opinion the potential for AI-driven strategies to reshape sports extends beyond soccer and professional sports in the US. In the United States, College Football and College Basketball provide fertile ground for such innovations. College teams have diverse opportunities to capitalize on financial gains that comes from establishing winning programs. For example UTSA or Texas State can be in a place where they replace Texas Tech as the third biggest commercially valued school in Texas if they can establish winning cultures, because the fact of the matter is that winning brings students (money) growing the school's bottomline which is why programs invest heavily in their athletics programs. Currently, based on my exposure, the utilization of analytics and data in college sports may not be as advanced as one might expect but this is a place where I would not be surprised if new advancements in sport analytics and coaching came out of.
The research capabilities of academic institutions can be the driving force behind these initiatives. As the sports analytics industry matures, the emergence of a new generation entering the workforce with specialized expertise in sports analytics further fuels these developments. This synergy between academia and sports analytics may ultimately propel AI-driven game plans and strategies into the forefront of various sports, not just in Europe but also within collegiate sports in the United States.
If A Tactical AI Is Developed Will It Make Coaches Redundant?
The emergence of AI in sports analytics raises the question of the role of coaches in this evolving landscape. As AI systems become capable of generating game plans, coaches will have the opportunity to focus on their core strengths. Drawing a parallel with chess, AI engines excel at chess-specific intelligence but fall short in broader, general intelligence. This is why human-AI pairs can outperform chess AI. Humans possess the ability to consider the bigger picture and make nuanced decisions that AI cannot generalize.
Similarly, in basketball, the human element remains crucial. Beyond the tactical aspects, psychology plays a significant role in both chess and basketball. AI systems that focus on game plans may not easily encompass the psychological dimension of the game. This suggests that, even with AI assistance, human coaches will continue to contribute by recognizing and responding to patterns in the moment, enhancing team performance and strategy and this is one point the book heavily argues for.
Conclusion
Range has been a great read 3 chapters in and I highly recommend it to all people, I am sure there will be at least one positive take away from it. Of course when reading books, it is ok to challenge some ideas in it and I am not saying what it is saying is a fact but it introduces a lot of interesting and applicable perspectives in many different industries such as sports but also in life, as in how we approach our lives. Recently, I have picked up playing a new sport and I can see how this applies.