On the field: MONEYPUCK
Alexandra Mandrycky, IE 13, is helping the NHL’s Minnesota Wild put analytics to work on the ice.
Most professional sports leagues and teams have embraced data analytics as a way to evaluate player performance and influence playcalling, a movement most famously depicted in the book and movie Moneyball. But the National Hockey League has been relatively slow to embrace the approach.
Baseball, after all, has been a stats-driven sport for decades, and its fans are well versed in the value of BA, ERA, WHIP and even OPS and BABIP. And football not only employs an endless array of complex, drawn-up plays and schemes on both sides of the ball, but also collects and critically analyzes data on its players’ physical attributes and skills—such as height, speed, bench press repetitions and even hand size—like few other sports do.
Not long after NHL executives started looking beyond their scouts to bring in non-traditional voices into the decision-making processes that shape their multimillion dollar franchises, Alexandra Mandrycky, IE 13, was ready to be heard.
However, as a teenager growing up in Norcross, Ga., Mandrycky did not spend her evenings watching forwards and defensemen skate furiously on the rink. Hockey is not big in the South like college football is. But she eventually discovered the beauty of hockey when she was a senior in high school (the local Atlanta Thrashers hadn’t yet relocated to Winnipeg, Canada) and the sport broadened her horizons.
“I fell in love with the pace of the game and the unpredictable nature of it,” Mandrycky says. “When you are watching football or baseball, the action is more structured. Only one team is attempting to score for long stretches of actions, except when there are turnovers. Hockey is so fluid. You don’t know what is going to happen.”
That unpredictability has carried over to Mandrycky’s own experience as a hockey operations analyst for the Minnesota Wild. She started to learn about the specifics of hockey player development during her early stages as a fan. As she read more about the sport, she stumbled upon a growing movement that was using advanced statistical metrics –far beyond goals scored, assists, penalty minutes and the like—to measure player performance.
For example, there’s the Corsi rating, named for the pro goaltending coach who used shots-for vs. shots-against when given players are on the ice to measure his goalies’ workloads. This is about to get confusing, so see if you can follow: It’s a percentage calculated by adding shots on goal-for + blocked shots-for + missed shots-for divided by shots on goal-against + blocked shots-against + missed shots-against. This shot differential is a better indicator of team performance than goal differential.
Although watching and studying hockey was merely a growing hobby while she worked on her industrial engineering degree at Tech, Mandrycky realized that the data analytics she picked up from the sport also applied to her undergraduate classes. After she got out of Tech, Mandrycky decided to pursue data analytics in the hopes it could make following hockey something much more than a pastime.
Mandrycky reached out to the hockey world to see if anybody could use someone with her engineering background, passion for analytics and a growing love for the sport. She connected with statisticians Andrew Thomas and Sam Ventura, who started studying hockey analytics when they worked together at Carnegie Mellon University. Together the trio created War-On-Ice.com, one of several new websites that introduced and promoted a deeper analytical look into the sport. “The site started as more of an academic pursuit and I think their motivation was making this data available for anyone to analyze,” Mandrycky says. “You could download the stats and they encouraged people to do their own work.”
Mandrycky says she was brought onboard to help compile player compensation data for War-On-Ice.com. “But that soon led to work on the back-end infrastructure and database and data visualizations,” she says.
In the summer of 2014, a handful of NHL teams started hiring analytics gurus, stealing them away from the independent hockey stats websites. “A lot of people had to be taken offline because of their employment agreements,” Mandrycky says.
Mandrycky’s own shot at securing a job with an NHL squad first came through a consultant gig with the Minnesota Wild last September. In January, that turned into a full-time position as a hockey operations analyst. “It’s still surreal that I’m working for a pro hockey team,” Mandrycky says. “It says a lot about the Minnesota Wild that they are open to outside voices.”
Her specific charge for the Wild is to arm team personnel with information to make smart decisions. She finds and analyzes advanced data from a wide range of sources and then shares them with hockey operations staffers, the general manager and the coaches. With the NHL player draft held in June and the free agency period taking place immediately afterward, this offseason is a busy time for Mandrycky and her peers.
The communication among her friends in the industry is no longer an open flow as it was at War-On-Ice.com. After all, her analytic insights are now proprietary secrets. “I know people with other teams, but you are not supposed to talk about work with them,” she says. “I was at the annual Sloan Sports Analytics Conference and you see everyone and you talk, but you are not giving anything away.”
As Mandrycky absorbs information about a sport she was drawn to as a teen, she has learned that decisions with long-term implications are not made at a rapid-fire pace. Only so many players can be signed during a season or calendar year. “Not much actually happens immediately,” she says. “The satisfaction has to be in the process and the end results. It’s hard to know whether you did a good or bad job at the time. You just have to grade yourself on how you made the decision and how you followed the process.”
Mandrycky steeled herself for such an environment while studying at Georgia Tech. The industrial engineering curriculum made her comfortable working in many programming languages and environments. And working on long-range projects gave her the confidence to work on War-On-Ice.com.
However, she still enjoys watching the action on hockey rink that starts and ends in a few seconds. “Nothing can beat a good breakaway,” Mandrycky says. “Goals are still the most exciting part of the game.”