The Future of Sports: Dr Patrick Lucey of Stats Perform On The New Emerging Technologies That Are Disrupting The World Of Sports
You Don’t Frame Your First Sketch: AI projects are hard because you must worry about both the input data and the model. When starting an AI project that includes both live and spatial data, we aren’t sure of all the situations we’ll encounter. When I first started, I had the mindset that my solution had to address all possible paradigms, which results in building a complete monolithic system that’s difficult to maintain, update and improve. I quickly realized the key is to start small and build a small end-to-end system then refine, refine, refine until addressing all of the edge-cases.
New technologies have changed the way we engage in and watch sports. Sensors, Wearable Tech, Video Assistant Referees (VAR), and Instant Replay, are examples of new technologies that have changed the way we play and watch sports. In this interview series called, “The Future of Sports; New Emerging Technologies That Are Disrupting The World Of Sports,” we are talking to sports leaders, athletes, sports tech experts, and sports equipment companies who can talk about the new technologies that are reshaping the sports world.
As a part of this interview, we had the pleasure of interviewing Dr. Patrick Lucey.
Dr. Patrick Lucey is the Chief Scientist at Stats Perform leading the AI team with the goal of maximizing the value of the company’s deep treasure troves of sports data. Patrick has studied and worked in the AI field for the past 20 years, holding research positions at Disney Research and the Robotics Institute at Carnegie Mellon University as well as spending time at IBM’s T.J. Watson Research Center while pursuing his Ph.D. Patrick hails from Australia where he received his BEng(EE) from the University of Southern Queensland and his doctorate from Queensland University of Technology. He has authored more than 100 peer reviewed papers and has been a co-author on papers in the MIT Sloan Best Research Paper Track, winning best paper in 2016 and runner-up in 2017 and 2018.
Thank you so much for joining us in this interview series! Before we dive in, our readers would love to “get to know you” a bit better. Can you tell us a bit about your ‘backstory’ and how you got started?
I think I have the best job in the world as my role allows me to explore and utilize my two greatest passions: sports and AI. I grew up in Australia, and from childhood through my early adult years, played virtually every sport under the sun. Not only was I totally consumed with playing and watching sports, I developed an interest in sports data at a young age, as a child often collecting my own stats for soccer and cricket and memorizing players’ statistics from trading cards. I loved the way numbers could reconstruct the story of an athlete and allowed me to compare the strengths and weaknesses of my favorite players. Growing up, I always wanted to be a professional athlete, and although I didn’t quite “go pro,” I did play semi-professional soccer for eight years as a young adult. During this time, I conducted a lot of scouting and analysis on my team’s upcoming opponents. While I was able to break down the opposing teams strategically, the analysis was qualitative, which for someone who loves using data like I do, was very unsatisfying. Therefore, the question of how to improve measuring team performance was constantly top of mind.
During the same time that I was playing semi-professional soccer, I was also pursuing an undergraduate degree in Electrical Engineering. People often ask why I chose this degree, and my answer is the decision came down to a lack of available options at the time. My career ambition was always to play or be involved with professional sports in some capacity, but while I was trying to make that a reality, I needed to continue my education. My older brother had just completed his degree in Electrical Engineering and highly recommended it, so as most younger brothers do, I followed his lead. I earned my undergraduate degree then continued to pursue a doctorate in Audio-Visual Speech Recognition. While I’m proud to say I excelled at both my undergrad and graduate studies, the completion of my Ph.D. gave way to the realization that I wasn’t going to become a professional athlete, which was quite hard to come to terms with at the time. I wasn’t sure what to do next, so as Australians often do, I embarked on an overseas adventure.
I took off for Pittsburgh to accept a role at the Robotics Institute at Carnegie Mellon University where I worked on facial expression recognition. I was there for two years before receiving my next great opportunity at Disney Research Pittsburgh, where I spent the next five years working on their computer vision team to develop a fully automatic sports broadcast. While that team did groundbreaking work and generated a lot of great research papers, I wanted my work to have an impact and be utilized by people around the world. To do this, I knew I had to follow the data, and it led me to the place that has more sports data than anywhere else on earth — Stats Perform. So nearly six years ago now, I joined Stats Perform and have enjoyed every minute of it. I started the AI team from scratch and grew it from just myself to a group of over 50 scientists and engineers across the globe. We have the collective goal of maximizing the value of our data through products that revolutionize how sports is watched, consumed, and played. My work today utilizes many of the same methods I learned during my Electrical Engineering days, so it’s funny to see how things ultimately connect and goes to show that no hard work is wasted.
Can you share the most interesting story that happened to you since you began your career?
One of the most special moments as a scientist is the “Eureka!” or “Aha!” moment, and I’m fortunate to have had a few of those throughout my career. One I remember well came after giving a presentation where I proposed an idea related to using player tracking data in sport that was robustly debated, critiqued, and dismissed by the group. One particularly stinging comment I received was something like “video games have already solved AI for sports, so why are you doing this?”
While I knew video games and real-life sport were totally different, I couldn’t clearly articulate the difference, and that forced me to focus my efforts on this question: “What is the difference of AI for a sports video game like FIFA and real-world sport?” Later that night, I played FIFA on my PlayStation, and the penny dropped as I had one player dribble the ball from full-back to forward and saw that no-one had covered the defender. I thought “That’s the difference! In real-world sports, players are intelligent and automatically cover for their teammates where in computer games, the players don’t change positions and leave gaping holes.” This led to the realization that real-world sports require a dynamic representation of team play rather than a fixed one.
This key insight has led to a lot of great innovations using sports tracking data, such as being the first to automatically detect formations in soccer as well as ghosting and visual search, which leverages an “x’s and o’s” language and not just key words. Essentially, we’ve found that the hidden structure of team sport lies within our tracking data and have refined and improved that idea to now power nearly all of Stats Perform’s tracking-based solutions.
Can you please give us your favorite “Life Lesson Quote”? Can you share how that was relevant to you in your life?
I suppose my last story reinforces the importance of sharing ideas, being open to feedback and allowing yourself to “rethink” your position, but also it represents the necessity to give yourself time and space to rethink. Reflecting on this experience, I think a quote that captures the importance of being in such an environment came from reading Michael Lewis’ “Undoing Project,” which chronicles the relationship between Daniel Kahnemann and Amos Tversky, both pioneers in field of Behavioral Economics. One of Amos Tversky’s quotes that really resonated with me was “You waste years by not being able to waste hours.”
What I like most about this quote is it perfectly captures the need for a person to have the time to think and rethink a problem. Also, another key insight here is to not be lazy and waste time waiting for new ideas to be generated, but instead be relentless and refine the ideas and problems you care most about. In the long-term, if you don’t get the time to address the problem and define the scope, you’ll waste a lot of time and money working on a project or product that ultimately interests no one or prematurely scales before you get the right product-market fit.
None of us are able to achieve success without some help along the way. Is there a particular person who you are grateful towards who helped get you to where you are? Can you share a story about that?
In addition to my parents, wife and daughter, my eldest brother, who currently serves as Director at the University of Adelaide’s Australian Institute of Machine Learning, has been a huge influence in my life as he blazed a path that showed me the career I have today was possible. I’m also enormously grateful for the guidance and mentorship of our CEO at Stats Perform, who has shown me how to run a business as well as the importance of strategy and decisive decision-making skills.
Outside of my family and employer, the person who has played the biggest role in my career is Iain Matthews, my boss at Disney Research Pittsburgh who now works as a Principal Scientist for Epic Games. Iain first gave me the opportunity to marry my passions for sports and data through the position at Disney Research and allowed me the freedom to explore and pursue different ideas around tracking data. He taught me the importance of solving hard problems and never to shrink that responsibility. He also taught me to never settle for less than the highest quality possible. “Why have text when you could use an image, video or animation?” Whether you’re giving a presentation or preparing a demo, always strive to make it box-office quality. Iain would know, as he created the face tracking technology used in Avatar and other blockbuster films.
Is there a particular book, film, or podcast that made a significant impact on you? Can you share a story or explain why it resonated with you so much?
I read a lot of behavioral economics and science books as they deal with measuring and understanding human behavior from data and often look to domains outside of sport for inspiration because really what we do in sports is objectively measure human behavior using the data we have at our disposal. The only difference is the granularity of our data makes this complex than standard economic or behavioral studies. To that end, two books that have had the biggest impact on me to date are Atul Gawande’s “Checklist Manifesto” and Gary Kasparov’s “Deep Thinking.”
What I enjoyed most about “Checklist Manifesto” is the takeaway that simple lists and constant monitoring can be used to inform extraordinary improvements in complex industries like healthcare. The book emphasizes the importance of measurement and capturing KPI’s to show how lists can democratize the decision-making process to ensure the best decisions are made and based on merit rather than seniority. For “Deep Thinking,” I love how Kasparov expertly describes the future of AI as the symbiosis of human intelligence and computer technology and explains that the sweet spot is building technology that assists humans and improves our day-to-day lives. Both books have shaped how I approach my work at Stats Perform and the products our team is building.
You are a successful business leader. Which three character traits do you think were most instrumental to your success? Can you please share a story or example for each?
The three character traits I believe have been most instrumental to my success are mastering your field, following your passion and being self-aware.
1. Mastering your field is more than simply knowing the theoretical and technical aspects of your domain, it’s also knowing the business aspects and fully understanding what your customers value as well as the competitive landscape. You must know every inch of turf in your field to ensure you’re speaking the same language as your customers and creating what they want, not just what you think they need. To become a master in your field, Herb Simon and William Chase, which was later popularized by Malcolm Gladwell, suggested you need to have at least 10,000 hours of practice on a particular topic. This rule of thumb is well-established in the world of sports and, in my opinion, also applies to the world of sports technology. For example, at Stats Perform, our collective commitment to our craft is what sets us apart from competitors. My team has put in the hours needed to develop both the technical knowledge and the domain knowledge required to truly master our field, and it shows through our ability to know what can and can’t (or shouldn’t) be done.
2. My last point and this one go hand-in-hand. It’s much easier to master your field if you’re passionate about your work. The lens through which we see the world gives purpose to and drives how we operate. It also enables us to utilize the power of analogies and storytelling, which can simplify abstract concepts into a frame of reference that anyone can understand. It is also quite boring to talk to someone who is not passionate about what they are doing (even if it’s only a small aspect). However, it’s important to note that pursuing your passion often comes with sacrifice. For me, I had to leave Australia and my family and friends to pursue this passion and opportunity.
3. I’m fully aware that the ability to follow your passion is a luxury and requires having both the good fortune for an opportunity to present itself as well as the safety net to be able to seize that opportunity. Many people are never presented the opportunity to follow their passions due to situations outside of their control. Therefore, I believe it’s not only important to acknowledge the lucky situation people like myself have been afforded but also to explore ways to extend this special opportunity to others. For example, I was extremely fortunate to be born in Australia and have the resources to pursue my love of sport and numbers made available to me from a young age. I was also very fortunate that my time working in audio-visual and facial expression analysis gave me the experience needed to start working in sport. Lastly, for the work we do in sport, the real star of the show is the data. Without the deep treasure troves of data we have at Stats Perform, which are collected, stored and distributed by our world class operations and engineering team, none of the AI work we do would be possible.
How have you used your success to bring goodness to the world?
I strongly believe our work at Stats Perform matters. Sport is rare as it represents one of very few shared positive experiences. The reason sport is so exciting and why people continue tuning in to watch together from all over the world is because sporting events are ultimately played and coached by humans making decisions under stress and pressure. Those decisions are often irrational and wrong. Or so we think. As such, every single decision that a coach or player makes can be broken down, analyzed, and debated.
At Stats Perform, we are responsible for reconstructing the story of matches using our best-in-breed live data as well as the insights and analytics we generate. Whether people get this information via Google search, Amazon’s Alexa or Apple’s Siri, live video stream or a data feed from a sports book, the depth of data and level of insight we provide is crucial to how billions of people around the world consume and enjoy these events together.
Ok wonderful. Let’s now shift to the main focus of our interview. Can you tell us about the sports technologies that most excite you at the moment? Can you explain why you are passionate about it?
Absolutely. The “sports tech” trend or theme that’s currently most exciting to me is the use of AI to improve the consumption of sports content, whether you’re at-home watching on your TV, laptop, phone or tablet (or all these devices at once), a bettor, a coach or an analyst of a specific team. A few ways we can do this is by using computer vision technology to collect more granular data on every single action that occurs as well as using machine learning methods to objectively measure whether the decision made on that action was the right one. By using AI technology as the ultimate decision analysis tool, everyone associated with the sports world is empowered to make better decisions. This, in turn, affects consumers because better decision-making results in better quality play, higher engagement and more entertainment, all of which are extremely exciting to me.
In terms of specific technologies, the two I’m most excited about currently are broadcast tracking and analysis and live streaming predictions. Stats Perform’s AutoStats technology utilizes the latest computer vision techniques to capture player tracking data from a broadcast video feed, which is significant because this data historically didn’t exist or wasn’t available, and the organizations that have since started utilizing it are already seeing the value. For example, with AutoStats, we started by capturing tracking data for every college basketball game ever played. Through this data, we are able to detect every type of action (such as pick-and-rolls, drives, isolations, off-ball-screens, etc.), the defensive attributes of each player and how players performed against various opponents. This enables us to construct a much better understanding of a player’s specific strengths and weaknesses, which can help predict future NBA performance and inform more strategic draft picks. Stats Perform’s recent partnership with Orlando Magic leveraged AutoStats to help the organization do just that during the 2021 NBA Draft.
When it comes to live sports broadcasts, streaming is gaining popularity, and the experience of consuming sports through streaming is a different ball game, if you will. In soccer, we recently released a slew of new live AI metrics and predictions to accompany our live data feeds, including live possession value (calculates how dangerous each possession is as well as how valuable each action a player takes is), live season simulation (calculates the impact key events like goals will have on final team standings), and live team and player prop predictions (predicts what the end of match statistics will be at any point during a match). In tennis, we released our live interactive predictions, which tell you the likelihood of a player winning the next point, game, set or match at any point during the match. We’re excited about this product because, using our live predictive capabilities, broadcasters can predict the impact of future points before they happen. This “what-if” capability, which is based on our counter-factual work, is a current trend in making AI “explainable.” To make the consumption of such AI data easily accessible, Stats Perform released our PressBox solution, which enables broadcasters to generate any narrative they want using our best-in-class live data and AI content.
How do you think this might change the world of sports?
As mentioned, because it allows every decision and action to be objectively analyzed, I believe AI is the ultimate decision analysis and performance tool. As such, this technology is already having a massive impact on coaches and management in terms of decision-making support, but beyond that, I’m more excited about AI’s potential to improve fan engagement and consumption of sports in three areas: interactive experiences, personalization, and the long tail.
I believe AI has the potential to enhance the “shared” element of sport by making consumption a more interactive experience. In this “on-demand” world where most content is consumed asynchronously, sports are rare as fans still want to experience events at the same time. As an Australian now living in the United States, I still watch Australian sports. Likewise, my family and friends in Australia regularly watch US sports, and we often text during matches to share our thoughts and reactions. The idea of an at-home AI technology that makes these experiences more shareable and personalized to me as well as my friends and family is exciting to me. Additionally, whenever I watch a sporting event, there are so many decisions I’d like to analyze myself. Creating an interactive, AI-backed solution that allows me to ask the questions I want to ask and receive the information I find most interesting is another development I would love to see.
As far as personalization, generating personalized content for specific consumers is a massive benefit of AI as it enables us to improve the consumer’s experience and target them more strategically. But for a B2B business like Stats Perform, the “user” can be many things. It can be the direct consumer, where Stats Perform would use its Natural Language Generation technology to generate custom insights that would perhaps alert a user to an interesting game via our Smart Ratings. Or it could be a business with whom we share these insights, enabling their marketing and advertising teams to create more dynamic and strategic advertisements.
Lastly, it’s no secret that the best professional athletes are “the best” because they have access to the best analysis and coaching. My hope is incorporating AI technology into sport will democratize this level of analysis and feedback and make it available to everyone. A really good example of this is the partnership we have with the Women’s Tennis Association (WTA). As the association’s official data and AI partners, Stats Perform will bring AI solutions to the women’s game that do not exist in the men’s game. Additionally, we have recently focused our AI analytics in women’s soccer, which is something else I am really excited about.
Keeping “Black Mirror” in mind, can you see any potential drawbacks about this technology that people should think more deeply about?
While I haven’t seen “Black Mirror” yet, I do know what you’re referencing, and I think it highlights one of the most common fears about AI, which is that it’s used to monitor, control and curb individual freedoms of the public. In terms of sport, it could be viewed similarly as something that ultimately takes away the personal freedoms of individual players. To that end, I must re-emphasize what we do at Stats Perform as well as what we don’t do. We are the keeper of public record of sport, meaning for any game played in the public sphere, we transcribe what happened through our human and computer vision-based data collection methods. We don’t collect “private data,” which includes the physical measures of players like heart rate or any medical, contract and personal data. Focusing on match data alone protects Stats Perform against such privacy concerns and allows us to tell the best and deepest stories by using our data to give context to a specific event or individual performance as it relates to others throughout history.
What are the 3 things that concern you about the sports industry today? Can you explain? What can be done to address or correct those concerns?
While the positives that excite me certainly outweigh the negatives, there are still some things within the sports industry that concern me. The three things I’m most concerned by at the moment are:
1) Organization Data and AI Literacy: Given most sporting organizations are seeking an advantage of marginal gains to obtain that 1% advantage over their opponents, they are often seeking the “silver bullet,” meaning they are accepting technology or advice based on qualitative evidence and testimony rather than thorough scientific validation. As such, many new products and service providers in sport are using AI as a buzzword without validating the proper scientific evaluation. Therefore, one of my top concerns pertains to data and AI validation as well as literacy. As a decision maker within an organization, you need to know what data and AI can do, and more importantly, what it can’t do. To that end, Stats Perform has created a 4-part “AI in Sport” seminar series that is free and meant to educate on what can be done as well as limitations. Additionally, in order to maintain their competitive advantage or “secret sauce,” I’m finding many sporting organizations are reinventing the wheel by building and maintaining their own data and intelligence ecosystems, essentially becoming their own software or data company. Although this is laudable, it is not an effective use of resources when you consider the manpower, time and costs associated with building out such infrastructure in addition to maintaining and updating the system over time. A team’s value and intellectual property is centered upon its players and strategy or playbook. Organizations should aim to work with partners who specialize in this business, like Stats Perform, so they can focus their time and energy on preparing players to win.
2) Decision Transparency: My next concern is around transparency into the decisions determining key personnel in sports, specifically pertaining to the lack of black coaches. In basketball, soccer and American football, for example, where black players tend to dominate rankings, this discrepancy is glaring and, quite frankly, unacceptable. I believe AI can help solve this problem by identifying such biases and creating an objective method of comparison so every coach, regardless of race, is given the same opportunities.
3) Social Media Abuse: The online abuse athletes receive from people who consider themselves “fans” is both appalling and greatly concerning to me. From the horrific racist abuse directed toward UK soccer players after their 2020 Euro final loss to unwarranted online criticism of American gymnast Simone Biles following her decision to withdraw from the Beijing Olympics, the sports industry must come together to put a stop to this. At Stats Perform, our team is addressing social media abuse in sport by partnering with Signify Group to bring a best-in-class social media abuse monitoring service to the sports market. This service, called Threat Matrix, leverages AI and machine learning technologies to monitor millions of open-source social media posts across various platforms and flag abusive keywords, phrases, images and emojis in multiple languages in real-time. To my knowledge, Stats is the first sports tech leader to partner with a specialist firm on a project like this, and I’m incredibly proud to say I’m part of an organization that recognizes its responsibility to end this and protect both the athletes and the integrity of the games we love so much.
Fantastic. Here is the main question of our interview. What are your “5 Things I Wish Someone Told Me Before I Started” and why? (Please share a story or example for each.)
There was so much I didn’t know when I started, and I’m still learning every day. I’m amazed when I reflect at the end of the day on the amount of new insights and knowledge I’ve gained. Given that widespread productization of AI is relatively new in its emergence from academia to industry, I thought I’d focus my “5 things” on this technical element so people can pick up some useful tips if they’re exploring this topic for the first time:
1.) You Don’t Frame Your First Sketch: AI projects are hard because you must worry about both the input data and the model. When starting an AI project that includes both live and spatial data, we aren’t sure of all the situations we’ll encounter. When I first started, I had the mindset that my solution had to address all possible paradigms, which results in building a complete monolithic system that’s difficult to maintain, update and improve. I quickly realized the key is to start small and build a small end-to-end system then refine, refine, refine until addressing all of the edge-cases.
2) Focus on Edge Cases: Often business goals (i.e. what customers want) differ from how we optimize our models. In academia, our aim is to build the best model and achieve the best possible accuracy on the defined test set. In industry, however, the goals are slightly different. Whether a model is 89% or 90% accurate doesn’t make much difference to the customer, but customers will care when you can’t process a game due to some random edge-case. For instance, we dealt with a situation in soccer where the goalkeeper had been sent-off or injured and an on-field player had to step into be goalkeeper. Having domain expertise will immediately help with identifying and diagnosing such edge-cases, but it’s also key to include checks to gracefully deal with situations of that nature in the future. Within our AI team, we have developed best practices to operationalize and solidify our AI outputs following a similar strategy recently proposed by Andrew Ng.
3) Data >> Model: The most important aspect of developing any AI product is the data available to you, whether it’s the volume, granularity or how clean and reliable the data is. If you do not have any differentiation across these three dimensions, you should strongly rethink whether you should proceed because the model will only be as good as the data you give it. The shift towards “Data-Centric” AI (again, another strategy proposed by Andrew Ng) is a concept we have embraced over the last few years at Stats Perform as our differentiated data enables us to supply differentiated products.
4) Future of AI = Human + Computer: There’s a lot of hype around AI, especially the role of deep learning, which some believe is the “silver bullet” to solve all problems. Deep learning has proven itself to be extremely useful for learning the feature representations of vision, speech, and text data, where the deep neural networks are able to capture low-dimensional underlying signals using hundreds of thousands of hours (or more) of input data. We have used these methods for our computer vision systems to detect and track players, which in turn, has enabled us to achieve a level of performance we could only dream about years ago. However, at the end of the day sport is both played and officiated by humans, and you cannot expect the AI methodologies like deep learning to fully understand the rules and structure of the sport without any assistance. To build the most effective and robust technology in sport, I’ve found we must inject human domain-specific knowledge into the system to best capture the nuance and structure of the sport, which is why you need both technical and domain specific knowledge.
5) Managing Expectations Around AI: As I’ve mentioned a time or two now, there is truly so much AI can do, and while its potential is exciting, believe me, it’s equally important to know the limitations. In the healthcare industry for example, there are some situations where the utilization of AI has been over promised but underdelivered. It’s not that the AI technology is not capable, more so that the data needed to deliver on the promised outcomes is private and therefore inaccessible. This is true in sports with injury data, which is mostly private, fragmented and not available at the scale required to build out accurate and robust injury prediction systems.
You are a person of great influence. If you could inspire a movement that would bring the most amount of good to the most amount of people, what would that be? You never know what your idea can trigger. :-)
As I mentioned before, sports are rare as they represent one of life’s few shared positive experiences. Also, sports are a powerful vehicle for learning. As a field, AI can often be seen as abstract and complex. When taking an introductory course on AI, students are immediately initiated with theories and formulas. If students can overcome that initial barrier, this foundation sets them up for success in the field. However, some students are put off by this barrier and do not progress.
Similarly, to the untrained eye, sport can also seem abstract and complex. Take soccer and basketball, for example. Despite the complexity and nuances, billions of people all around the world have spent thousands of hours playing and watching these sports and, therefore, have profound understanding of these domains. At its core, AI and sport are very similar fields as both rely on learning behaviors directly from observations rather than having every action and rule specified. Not only are we creating technologies that enable people to better understand, interact and hopefully enjoy sports more, the work we’re doing makes AI more accessible to those who have a particular interest and understanding of sport. Instead of teaching theory and new concepts, learning about AI through sports can simply reveal what you already know. People would be amazed to see how the core concepts of AI are the same ones they already know from understanding sports. They just don’t know it yet.
Education is the most powerful tool to transform lives. I believe people often think fields such as data analytics and AI are too abstract therefore not viable career paths for them. My goal would be to use the power of sport and sports data as a vehicle for data literacy and AI literacy to empower more people to obtain this “abstract” knowledge and choose this as a possible career path if they so choose.
We are very blessed that very prominent leaders read this column. Is there a person in the world, or in the US with whom you would like to have a private breakfast or lunch, and why? He or she might just see this if we tag them :-)
Because my wife is by far the most interesting person I have ever met, I would always choose to have breakfast and lunch with her. However, if she was unavailable and I had the chance to have a sit down and coffee with anyone, I would like to pick the brain of fellow Australian Ange Postecoglou.
While I swear that I have interests other than sports and AI, the trajectory and progression of Ange’s coaching career is nothing short of amazing. For those who haven’t heard of Ange Postecoglu, I hope you will in the next five years or so. I won’t do his story justice, but at a very high-level, Ange was a solid football player who turned to coaching after having to retire early. He experienced immediate success in Australia with South Melbourne before becoming the Australian Youth coach, which was largely deemed unsuccessful. He turned this “perceived failure” into a positive by learning and reinventing himself into the coach he wanted to be. He took the Brisbane Roar from a team of no-names to the most successful team in Australian football history and achieving a 36-game winning streak. He then was equally successful as the Australian coach, qualifying for both the 2014 and 2018 World Cups and winning the 2015 Asian Cup. He had similar success with Yokohama FC and just recently moved to Celtic FC, where he is already transforming his team into a formidable outfit and starting to win over fans with “Angeball.” He’s done this not by having a star-studded roster, but by starting with the system and getting players to fit in within this system in order to execute at the highest possible level.
At the end of the day, what I do with sports and AI is build assistive tools and metrics to help decision makers better monitor performance and ultimately make better decisions. What I would love to pick Ange’s brain about is not how he uses these technologies and which metrics are most important, but how he ensures his players have the right character and mental build for success. I’d love to know what insights we could discover by marrying each organization’s private information with the Stats Perform’s public data.
How can our readers further follow your work online?
The best way to follow my work is on Stats Perform’s website at www.statsperform.com/artificial-intelligence. I also have a personal webpage — www.patricklucey.com — that links to all of the work we’ve done in this space as well.
Thank you so much for these excellent stories and insights. We wish you continued success on your great work!
