How the right kind of sports analytics can transform business thinking

Famed sports announcer Howard Cosell once remarked, “Sports is the toy department of human life.” This succinct observation highlights how sports, much like toys in childhood, can serve as a space for understanding the world. Cosell also observed, “Sport is human life in microcosm.” The highs and lows, the strategies, and the split-second decision-making in sports mirror the complexities of our world.

Just as sports reveal patterns in human behavior, decision-making, and performance, analytics offers a powerful framework to distill these patterns into actionable insights.

In business, as in sports, success often hinges on informed decision-making, adaptability, and the ability to interpret data effectively. The rise of sports analytics demonstrates how data can uncover inefficiencies, optimize strategies, and predict outcomes. Through the lens of sports analytics, we can gain a deeper understanding of competition on the field and in the boardroom.

Simulating the unlikely

For several years, my phone would ring with calls from Burbank, California. On the other end was a producer from an ESPN show: I knew I was about to be posed a problem that had stumped their analytics crew. On the show, exact probabilities were rarely needed. A statement like 1 in 400 was often as helpful as, say, 1 in 389.42, so I often required only rough estimates of an underlying phenomenon. 

To illustrate the point, we can apply this estimation approach to Stephen Curry’s remarkable shooting performance against the Knicks at Madison Square Garden in February 2013. Curry racked up 54 points, making 11 of his 13 three-point shots — a clearly impressive feat, but how improbable was it? During the 2012–13 season, Curry had a three-point shooting percentage of 45.3%. This is close to 50% so we can simulate Stephen’s shooting with the flip of a coin. If you flip heads, you make a shot, tails you miss. Flip the coin 13 times and see if you get 11 or more heads. In a room of 100 people, on average one person will accomplish this.

So, the feat was improbable — but not historically so for someone of Curry’s skill. While we lost accuracy, we used a quick way to gain a sense of the level of improbability. Simulation offers a way to uncover statistical concepts without requiring advanced statistical training.

Navigating uncertainty

With our analysis of Curry’s performance, we accepted imperfection to prioritize speed and simplicity of computation. Sometimes, imperfection is inherent in a process.

Every March, national attention turns to NCAA Division I men’s basketball and the frenzy of “March Madness.” The 68-team single-elimination tournament begins with Selection Sunday and culminates in a national champion. By the Thursday after Selection Sunday, fans complete their brackets, predicting outcomes from the first round to the championship. These brackets fuel friendly competitions and high-stakes pools, including a $1 billion prize Warren Buffett insured in 2014 for a perfect bracket — a feat no one achieved.

Try Big Think+ for your business

Engaging content on the skills that matter, taught by world-class experts.

Request a Demo

Using analytics to create successful brackets is central to my research. Analyzing data from roughly 5,000 games each year reveals patterns that improve predictions. The odds of a perfect bracket through random guessing are 1 in 9 quintillion — equivalent to picking a single second out of 285 billion years. Many fans rely on some form of analysis, whether it’s statistical models or a preference for team mascots. The average accuracy at predicting March Madness games is 70%, resulting in odds of 1 in 5.7 billion. A mere 1% increase in accuracy further reduces the odds to 1 in 2.3 billion.

Focusing solely on perfect prediction, however, can limit innovation. Randomness in sports makes perfection inherently improbable. Analytics, instead, identifies trends and enhances decision-making — on the court or in the boardroom. Embracing imperfection leads to more practical and impactful insights, moving us closer to understanding the unpredictable.

Turning data into actionable insights

The office is stuffy after hours of analysis. Scraps of paper litter the desk, whiteboards are smudged with ideas, and questions hang in the air: How can outcomes improve? Can weaknesses be minimized? What adjustments are needed? This is a typical day for many coaches. It’s also the setting for my analytics group, Cats Stats, to collaborate and provide actionable insights.

In 2013, three math majors walked into my office and asked if we could create an analytics group to support Davidson College’s men’s basketball team. We partnered with the team and quickly realized that not all analytics were equally useful. While some insights intrigued us, coaches would respond, “This isn’t coachable.” Inherent in that response is a fundamental principle of analytics: align the work with the stakeholders’ goals.

Analytics are often produced in an iterative process — analyze, present, and refine. Coaches rely on clear, practical takeaways. By maintaining open communication, we ensure our analytics enhance their decision-making.

In 2015, Cats Stats gained national recognition through features in The New York Times, NPR’s All Things Considered, and ESPN’s FiveThirtyEight. The students emphasized a recurring theme in their interviews: Actionable insights are our focus. Our analytics are ready on the coaches’ desks as soon as one game ends, preparing them for the next. Like coaches, business leaders benefit most from actionable data tailored to their needs. To deliver this, analysts must ask: Who is the audience? What do they find actionable? If these questions cannot be answered, time should be set aside for discussion. Such conversations can uncover hidden opportunities and refine insights, ensuring maximum impact.

Analytics isn’t just about numbers — it’s about creating clarity and driving action. Whether for a basketball team or a business, actionable insights are the key to turning data into meaningful results.

Double-take on data

Modern NBA games can be analyzed with unprecedented precision using NBA tracking data. For over a decade, the league has employed cameras mounted in stadium rafters to record the ball’s location and every player’s position on the court 25 times per second. Additionally, three-dimensional coordinate data is provided for the ball. This level of detail allows analysts to address questions that were once nearly impossible to answer.

Let’s examine LeBron James’ shooting patterns while highlighting the importance of analyzing data from multiple perspectives. In 2017, the Crumpled Jumper blog examined LeBron’s shooting during the 2016–17 regular season, noting that after making one shot and missing the next, he moved an average of 5 feet closer to the basket on his following attempt. A similar analysis of LeBron’s 2023–24 regular-season performance revealed that, under the same conditions, he moved an average of 4.7 feet closer to the rim.

Such a result falls into the intriguing but not “coachable” bin. Coaching staff typically seek analytics that informs strategy or preparation. If LeBron missed a shot, how should a team play him differently? Such a result, as currently stated, does not offer such insight.

At first glance, the statistic suggesting LeBron shoots closer to the rim after making and then missing a shot seems to offer defensive value. However, further analysis clarifies this pattern. In such sequences, only 27.5% of his missed shots were layups or dunks (i.e., close-range misses). Yet, after the miss, he attempted layups or dunks 51.1% of the time — close to his overall game frequency of 46.5%. LeBron’s change in the distance has more to do with the distance of shots missed at a higher frequency.

Digging a bit deeper, after making a shot followed by three or more consecutive misses, LeBron’s frequency of layups or dunks rises to 60%, which is higher than his average frequency. This change in “shot-type frequency” could provide actionable insights for defenders. Players might be coached to anticipate an increased likelihood of layups or dunks during streaks of misses. Further analysis, including a video review, could refine these insights, exploring whether LeBron’s behavior is unique or shared by other forwards.

By examining the percentages and context of LeBron’s shot types, the seemingly surprising “5-foot adjustment” statistic becomes clearer. This approach highlights the importance of digging deeper into the data to uncover trends that offer strategic value. Insights like these, when well-analyzed and contextualized, can bridge the gap between interesting patterns and actionable results.

Sports offer more than just entertainment — they provide a unique lens for understanding analytics in action. By observing how data informs strategies on the court, you can draw parallels to your business challenges. Whether it’s uncovering patterns, making informed decisions, or adjusting tactics in real time, the principles of sports analytics can inspire innovative approaches to problem-solving. 

As you enjoy the game, consider how these lessons could apply to your organization. Asking the right questions and tailoring insights to your needs can turn analytics into a powerful tool for business success.