Baseball has embraced statistical analytics for quite some time now. It’s really gone mainstream within the last 10 years. There are hosts of websites and blogs dedicated to breaking down each granular iota of data. Virtually every MLB team now has multiple people dedicated to the angle of analytics.
Hockey is still a little bit behind baseball in terms of their analytical life cycle, but the sport is rapidly catching up to baseball. The issue is that baseball is more conducive to taking, processing, and applying statistical data, due to the stop-start nature of the game. Hockey is a free-flowing game with periodic interruptions, where players are changed on the fly for the most part. Most hockey advanced stats right now are centered on possession. The theory, correctly, is that if you have the puck more than your opponents and generate more quality opportunities than your opponent, you’ll win the game more often than not. Stats like Corsi and Fenwick are the first-generation of hockey stats; it seems like something is missing with the stats that have permeated into the general population, though.
Perhaps the next generation of advanced hockey stats will come from the extensive use of high-speed tracking cameras at NHL rinks. Many private companies have sprouted up in recent years, such as SportlogIQ with their player-tracking technology and machine logic algorithms and Stathletes with their video analysis. It’s no surprise that both of these companies are based in hockey-mad Canada.
But what can be extracted from these troves of data? Obviously, the possession-based advanced stats will improve. Perhaps it will help parse out a stat like Corsi to show more how each individual does in terms of possession, rather than the more team-based nature of Corsi right now. The real bleeding edge of hockey analytics, though, will come with advances that help with injury minimization of players.
The players are the true assets of the game, both in terms of the on-ice product for the fans and the real-world financial assets of the teams. These players cost a lot of money nowadays, so teams are going to want to do everything possible to keep them upright and playing as much as possible. A $6M winger does no one any good if he’s sitting in the press box for 20 games a year. With some non-intrusive player trackers, the in-game health of the players can be followed more effectively. It would be possible to download data after each game for each player directly to the team’s medical staff, where they could see energy expended, heart rate, caloric output, and oxygen levels.
The high-speed tracking cameras that are mounted in the rafters of some NHL arenas may also help training and medical staffs counsel coaches on hot spots to avoid on the ice. Cameras may be able to determine over time where injuries are more prone to happen that may not be as evident to the naked eye. Obviously the front of the net and the corners are where the action happens, but there may be a point on the ice where acceleration is maximized for both teams. Teams may game plan to avoid sending pucks into these areas without controlling them, so as to cut down on potential impacts and minimize injury risk.
It’s also natural to assume that networks will try to apply real-time analytics to their broadcasts in some aspect. Perhaps there would be a filter or toggle that a viewer could turn on to see stats like player speed/acceleration or we’ll be able to see impact strength in Newtons or foot-pounds on our screens.
It just seems like this era of Big Data that we’re in right now is going to lend itself to some sort of revolution both for teams and for the viewing public. Within the next five years, hockey as we consume it today may be totally different.