Growing up in Toronto in the early 2000’s, I was raised during a boom in Canadian basketball. Vince Carter, Chris Bosh, and (unfortunately) Andrea Bargnani were the names that defined my sports-childhood. As I stared wide-eyed at awe-inspiring displays of athleticism, and experienced heart-stopping buzzer-beaters at the end of games, I, and many young Torontonians like myself, decided that we would grow up to play in the National Basketball Association.
Unfortunately, most of us were not blessed with the physical attributes or skills necessary to realise our dreams. As we got older, we came to terms with the fact that we would never be able to compete professionally in the game we loved. Nevertheless, the desire to be a part of the world of sports still burned within our unathletic hearts. So we used all the strength in our puny, undefined arms to pick up as many books as we could. Maybe, if we studied hard enough, we could enter the realm of athletics from some other angle. NBA-doctor, NBA-journalist, NBA-agent: these were the fields we pursued, and these were the new dreams that occupied our minds.
Well, good news to all the uncoordinated, flat-footed, butterfingered, and otherwise athletically-inept basketball fans: the age of the NBA analytics guru is here, and these all-knowing nerds are turning the world of basketball on its head.
Well, good news to all the uncoordinated, flat-footed, butterfingered, and otherwise athletically-inept basketball fans: the age of the NBA analytics guru is here, and these all-knowing nerds are turning the world of basketball on its head.
The transformation of the NBA has in part been driven by advances in computing software. In every NBA arena, there is a system of Sports VU cameras hanging from the rafters. These cutting-edge cameras track every player on the court, and gather information about their performance at a rate of 25 times per second. Among other things, the cameras measure each player’s movement and speed. They track who shoots the ball, where the shot is taken from, and most importantly, whether or not the shot goes in. This data alone, however, is rather useless. It is only when it has been analyzed by math-savvy savants that the data is transformed into the beneficial statistics now ubiquitous in the NBA universe.
Today, every team has a highly valued analytics department that evaluates players based on criteria few people had even heard of 10 years ago. For example, a popular advanced stat is Real Plus Minus (RPM), which evaluates a player’s impact on the difference between how many points their team scores, versus how many points opponents score against it. Unlike traditional plus minus, which is heavily influenced by a player’s teammates and opponents, RPM uses statistical models to adjust for environmental factors, and isolate the true impact of each specific player. This is just one example of advanced statistics among many — Player Efficiency Rating (PER), win-shares, true shooting percentage, to name a few. Consequently, if you were to walk into the average NBA executive’s office today, or even a sports bar, you would hear people casually talking about advanced stats.
Journalists have embraced the data revolution more than anyone. Bleacher Report columnist Adam Fromal has written a series of articles that objectively ranks players based on various metrics. Because those parameters used different scales, Fromal standardized them by using adding their respective z-scores together. This type of math is relatively standard practice in today’s NBA, but if you asked Michael Jordan or Larry Bird how high their freaking z-scores were during their playing careers, they’d probably knock you out.
Overall, these statistics are starting to replace traditional “eye-test” player evaluations, and are altering NBA team’s — and the fan’s — perceptions about which players will lead them to success. Not only has this made opposing team scouting and internal player development more efficient, it is has helped teams make decisions in the most uncertain environment of all – the NBA draft.
Deciding which players will excel in the NBA has been an inexact science for as long as the entry draft has existed. Steph Curry, the two-time NBA champion and two-time most valuable player award winner, was picked 7th in 2009. Do you know who was picked 6th? Johnny Flynn. If you’ve never heard of Johnny Flynn, that’s because no one has. The point is, drafting is like trying to hit a moving target while blindfolded, on a horse that’s also blindfolded.
Steph Curry, the two-time NBA champion and two-time most valuable player award winner, was picked 7th in 2009. Do you know who was picked 6th? Johnny Flynn. If you’ve never heard of Johnny Flynn, that’s because no one has.
Nevertheless, Houston Rockets General Manager and notable analytics pioneer Daryl Morey is using statistical models to try and better predict which college youngsters will have future NBA success. In Moneyball author Michael Lewis’s most recent book, The Undoing Project, Morey discusses how he uses projection models as a part of his drafting process. He and his staff use historical player data to complement traditional scouting, in an effort to avoid the systematic errors in judgment made by NBA executives in the past. The Rockets have the NBA’s second-best record this season, and executives around the league are noticing that Morey’s math is working.
As integral as player evaluations are, advanced algorithms have altered the NBA in an even more fundamental way, revolutionizing basketball strategy over the past decade. The three-pointer is now viewed as one of the most desirable and efficient shots to take in the modern NBA. Once reserved primarily for oddball specialists whose lack of athleticism prevented them from scoring closer to the basket, the three-point shot has evolved into an integral part of every team’s offence. The reason for this about-face is the simple mathematical reality that three is more than two. For example, if a player averages 42% shooting from two-point range, then the expected value of that shot is 0.84 points. If a player averages 35% shooting from three-point range, the expected value of that shot is higher, at 1.05 points, even though the field goal percentage is lower.
The three-point line was introduced in 1979. It may have taken the NBA 30 years to listen to its nerds, but today every team embraces the efficiency of the three-point shot. The league average of three’s taken per game more than doubled in the past 20 years, from approximately 13 attempted per game in 1998, to 27 in 2017. This year, Morey’s Rockets have attempted a league leading 43 threes per game, representing over 50% of their total offensive effort. To put that into perspective, the entire league only took around 220 threes in total in the 1980 season. The Rockets eclipsed that number six games into this year!
It’s not just the volume of threes that are changing, but the types of players that are shooting them have shifted as well. Three-point shooting is valued so highly in the NBA today that even centres — big men who traditionally play close to the net — are chucking up three’s at an unprecedented rate. This season, the top five three-point shooting centers average upwards of 4 attempts per game. In 2012, the top five shooters took less than one three-pointer per contest. Walk into a college gym today, and you’ll see seven-foot giants heaving up practice attempts from beyond the arc. Five years ago, those same players would have been benched by their coach for standing more than 5 feet from the basket. The nerd revolution is fundamentally changing the way basketball is being taught to the next generation of players.
So, at the end of the day, all of you basketball fans who are 5’6 on a good day, who can’t jump, who can barely run, and who sometimes trip while standing still, remember this: hope is not lost. Pick up your stats textbook, lock yourself in your room, and put your glasses on. Because if you study hard enough, you may just end up being one of the math nerds who really run the NBA.
