Analytics and Business Intelligence Have Changed Sports Forever

Here’s how the story begins: A four-year-old boy growing up in Ontario in the mid-1960s grabs a pencil and a couple of pieces of notebook paper and sits down in front of the television to watch a hockey game.

He does this all the time. He draws the rink on his paper and as he watches the players skate and pass and shoot and check, he traces the movement of the puck over and over again from his living room floor. All these hundreds or thousands of individual lines record the seemingly random trajectory of the puck throughout the game

Over the course of a game, a season or many seasons, some discernible patterns emerge from the chaos. Clusters of goals are scored repeatedly from very specific locations on the ice. But even long before the puck goes into the net, this little boy looks at his drawings and notices how and where the puck travels before a goal is scored. It informs him not only of where the puck has gone, but where it will be before a scoring opportunity develops.

He internalizes this crude data and analysis and takes it with him to the ice when he plays hockey. He’s playing with kids several years older than himself. He’s smaller than all his opponents and his teammates. He’s not the fastest skater. He doesn’t have the hardest or most accurate shot. And yet he’s absolutely dominant in every game he plays.

Ten years later, he’s playing professional hockey. He’s still not the biggest or the fastest or the most physically gifted player, but he’s always in the right spot. His timing and positioning and play-making skills are surreal. He’s a bona fide prodigy.

Wayne Gretzky will go on to play professional hockey for 20 seasons and shatter every individual scoring record in the history of the sport. These records (total goals, assists and points) still stand more than two decades removed from his last shift and likely will never be threatened. He will win four Stanley Cup titles and be named the league’s most valuable player an unprecedented nine times, also a record that figures to stand forever. 

The “Great One” was a stone-cold data analytics pioneer.

This seemingly random and unstructured data has always been there. But knowing what to actively look for rather than passively responding to it after the fact is where the value can be found.

In the multibillion-dollar industry of professional sports today, whether it’s called data analytics, business intelligence, sabermetrics or advanced statistics, the goal is the same: gather and analyze as much information as possible to give your team and your organization the absolute best chance to win.

These same principles play out every day in the business sector. Thanks to the ease and volume of data collection now possible across all industries, data-harvesting and parsing tools will continue to redefine the optimal way to use people and technology to maximize profitability.

In the world of big-time professional sports, this so-called analytics revolution has fundamentally changed the way coaches coach, players play and fans consume their favorite sports teams and games. Many of these innovations – based on oceans of data that always existed but could only be effectively captured and contextualized thanks to recent technology – challenge conventional methods, strategies and evaluation practices that have prevailed for generations.

“The frontier of analytics is just beginning and there is no end in sight to the potential,” Lynn Lashbrook, founder and president of Sports Management Worldwide, a Portland, Ore.-based sports agency and management training institution, told WorkInSports.com, a blog dedicated to guiding sports business professionals. “Technology has opened the door for infinite analytic advancement.”

Put simply, the combination of data-capturing tools – digital cameras covering the hockey rink or the basketball court, GPS tracking sensors on individual players, applications that track the trajectory, location and exit speed of thrown and batted baseballs during hundreds of thousands of games, etc. – and the willingness of teams to embrace and curate this information to improve on-field performance has fundamentally changed the way games are played.

This reality has also given birth to an entirely new industry of its own. For decades, general managers, the people responsible for building rosters within the confines of a budget or a league-mandated salary cap, were almost exclusively former players. Today, those high-profile, high-paying positions are just as likely to be filled by Ivy League graduates with degrees in finance or analytics – or both.

Theo Epstein, currently the general manager of the Chicago Cubs and former GM of the Boston Red Sox, is a Yale alumnus. During his tenures atop two of the most storied and tormented MLB franchises, the implementation of advanced metrics and data analytics helped bring championships to two teams that hadn’t won a World Series for a combined 194 years. This resonated with MLB owners to such a degree that today roughly half of all professional baseball teams have Ivy League graduates serving as either their general manager or president of baseball operations.

Imitation, of course, remains the greatest form of flattery. This philosophy – to turn over the day-to-day operations of major sports franchises to the nerds and the technophiles –has reverberated throughout all the world’s major sports leagues. Today, interns and assistant GMs and even coaches are making their way up the ranks in football, baseball, basketball, soccer and hockey without ever having played the games they’re now reconstructing.

Big data and cloud-based applications are common and integral to just about every business segment today. Retail, banking, transportation and health care, to name just a few, all rely on a deep and wide portfolio of technology tools to improve efficiency, maximize profitability and, theoretically, provide an improved experience for their customers and employees.

Comparatively, professional sports leagues and teams were a little late to the party. Part of this reticence was based on nostalgia and tradition. There was – and still is to some degree – a sentiment that only those who played the game really understood the best strategy for success and were uniquely qualified to evaluate the skills and potential that gave the team the best chance to win.

“I’ve always believed analytics was crap,” Charles Barkley, a Hall of Fame NBA player and prominent basketball analyst for TNT and TBS, famously said of the data-driven mindset of the modern sports personnel executive. “All these guys who run these organizations talk about analytics, they have one thing in common – they’re a bunch of guys who have never played the game and they never got the girls in high school, and they just want to get in the game.”

“The NBA is about talent,” he said.

Barkley’s simplistic dismissal and on-brand contrarianism aside, there is a middle ground to be found somewhere between what the data is suggesting and what really happens between the lines. Investment in business intelligence and data analytics in sports isn’t going away. Now it’s about refining the process to blend the “hard” data with the so-called “eye test” to extract the best performance from individual players and their teams.

Spending on sports analytics applications and the data-capturing equipment feeding this insatiable hunger for data is projected to grow at a compounded annual growth rate of 40.1% a year through 2022, eventually topping out at nearly $4 billion by 2022.

This investment will go far beyond the team management and roster-building analysis tools to include everything from fan engagement to marketing endeavors and injury and health assessments. The data analytics revolution has already taken root in every nook and cranny of the sports industry.

It’s happened fast, too. In fact, if a sports fan were to awake from a years-long coma and tuned into an NBA basketball game or MLB baseball game today, it’s likely he or she wouldn’t recognize much of what they were seeing.

In the recently completed NBA regular season, teams attempted an average of 32 three-point field goals per game – the most in the league’s history. During the 2012-13 regular season, teams averaged a total of 20 three-point shots per game. That’s an astonishing 60% increase in three-point shots in just six seasons. In the 1979-80 season, the first year the three-point shot was adopted, teams averaged 2.8 attempts per game.

Why?

Data and math. Whether you’re a nerd or a jock, everyone understands the only way to win more games is to score more points than your opponent. This season, teams made an average of 11.4 three-point shots on 32 attempts, a 35.5% success rate. Meanwhile, the standard two-point field goal – taken closer (sometimes much closer) to the basket – was converted 46.1% of the time.

If a team takes 100 2-point shots in a game and converts them 46.1% of the time, those made shots are worth 92.2 points per 100 shots. If a team takes 100 3-point shots and is successful 35.5% of the time, it results in 106.5 points per 100 shots. This past season, teams averaged 89.2 total shots per game (32 3-pointers and 57.2 two-pointers).

So every team should just shoot 3-point shots exclusively, right?

This is where the business intelligence component of optimal game strategy takes flight. 

Los Angeles-based Second Spectrum is the official video tracking technology provider for the NBA. It, like its predecessor, SportVU, uses digital cameras installed at every NBA arena to track and record 3D spatial data of the ball, the players, the referees and other “spatiotemporal patterns” to give teams, fans and media an entirely new way to watch the game and process what is happening on the court.

It augments data extracted from the video to identify the type of patterns and in-game sequences that Wayne Gretzky was doing with a pencil and paper in his living room 50 years ago. 

But instead of only gathering data for one game at a time maybe a couple times a week, this database is constantly refreshed with every second of every game played. Moreover, this data can be reviewed in real time, making it possible for coaches and players to make in-game adjustments on the fly.

New statistics such as Quantified Shot Quality and Quantified Shooter Impact are now part of the lexicon, giving general managers, coaches and players even more nuanced information to take to the court to improve their chances of scoring and winning games. 

Points, rebounds, assists and shooting percentage are still informative and worth noting, but advanced statistics such as PER (a per-minute rating that boils down every individual player’s positive and negative accomplishments into a tidy per-minute rating) and Effective Field Goal percentage (which places more weight on 3-point attempts and conversions) are the advanced statistics that general managers, players and their agents are paying closer attention to these days.

By having these tools at their disposal, coaches, general managers and players have plenty to mull but, as Barkley said, at some point it does come down to talent. Basic analytics suggest that NBA teams should be taking more 3-point shots per game because they have more value. 

But the type of data generated by Second Spectrum tells everyone something even more valuable: who should be taking those 3-point shots and where they should be taking them from. The data found that most players shoot a higher percentage of three-point shots from the corners of the court or from the top of the arch. Thus, teams practice and design their offenses to get their best shooters isolated in those targeted areas – even at the expense of taking closer and much easier 2-point shots. Because, math.

Not surprisingly, this data-driven revolution is changing the way the game is taught and played at lower levels. College and high school basketball teams are shooting more three-point shots than ever before. Young kids starting to play basketball are practicing shots from beyond the three-point line, ensuring the next generation of players are more efficient and effective as the game evolves into the future.

In Major League Baseball, an individual sport masquerading as a team sport, statistics are as much a part of its culture and history as the hot dogs and Cracker Jack. Advanced statistics based on data harvested from tens of thousands of games made it abundantly clear that teams needed to change the way they positioned their defensive fielders.

The shift — the repositioning of defensive players to one side or the other of the infield — has radically changed the way professional baseball is played. The data proved that most hitters, particularly left-handed hitters, hit a much higher percentage of balls to the right side of the infield between first and second base.

None of this is particularly new. Ted Williams, a Hall of Fame outfielder for the Red Sox who is widely considered the greatest pure hitter in baseball history, was so feared that teams started using a variation of the shift back in the 1940s and 50s. But this adjustment was an anomaly. Today, coaches and players will set their defensive alignment for each individual player during each and every at bat based on the statistical data collected throughout his entire career.

It’s a cocktail consisting of big data, customization, advanced metrics, business intelligence, artificial intelligence and data mining all in one.

Batting average, home runs and runs batted in have long been the key batting statistics held in most regard (and nostalgia) by baseball fans and personnel executives. Today, on-base percentage, slugging percentage and on-base-plus-slugging percentage are the metrics general managers and coaches rely on when building a roster and filling out a lineup.

For pitchers, it’s no longer just about wins, earned run average and strikeouts. Now FIP (a measurement of the pitcher’s effectiveness and earned run average independent of fielding errors and mistakes) and WHIP (average number of walks and hits allowed per inning) are in vogue.

Advanced metrics have found their way into hockey, too. For decades, the prevailing strategy for teams on the power play was to either have the puck-possessing player enter the offensive zone onside on his own, pass to a player closer to the offensive zone or fire the puck into the offensive zone and then go chase it down to create scoring opportunities.

With the help of mountains of data and video files, forward-thinking hockey minds realized that it made more sense to have an offensive player bring the puck up to or near the offensive zone and then pass the puck to a trailing teammate behind them – a technique that was largely bemoaned and coached against for years – in order to better organize the offensive players as they entered their opponents defensive zone.

As a general rule, more players on the field – football and soccer, for example – results in more unpredictable patterns and variance. But both of these sports are incorporating wearable technology and more nuanced video-capture technologies to identify patterns, opportunities and inefficiencies during games.

For industries outside of professional sports, the message is clear: while you may not be gathering the same degree of detailed information about your employees and business processes on a daily basis, these teams and leagues have provided a blueprint for how you can begin to create your own analytic framework to evaluate employees, sales programs and new product and service rollouts without slavishly holding to tradition and antiquated benchmarks.