The Bruins were clinging to a 5-4 lead with just 17 seconds to play over a desperate Maple Leafs team. Chris Kelly had just been sent off for a delay of game penalty and with the Toronto net vacant down at the other end of the ice, the B’s found themselves on the short end of a 6-on-4.
Bergeron calmly won back the must-win draw, the puck was cleared from the Boston defensive zone, and seconds later, the Bruins had themselves another win over a hapless Toronto squad.
Winning face-offs is nothing new for Bergeron — one of the best two-way players in hockey right now. The statistics tell us that, as Jonathan Toews is the only player in the league who has taken 1,000 draws with a better percentage than Bergeron’s 57.9 percent success rate.
We know that Bergeron is a good faceoff man because the stats tell us that. We know that he’s a great faceoff man, however, because we’re able to see that he’s especially good when it matters most — like Tuesday in Toronto.
And therein lies the statistical debate that has heated up in hockey circles in recent years. Thanks to the use of advanced metrics in sports like baseball and football, there are those who want to infuse hockey and stats.
There’s certainly some momentum for the use of statistics in hockey, but at the same time, any sort of analytical revolution will be met with hesitation in a sport that is so reliant on incalculable intangibles.
“Stats are like a lamp post to a drunk,” offered Maple Leafs general manager Brian Burke this past weekend in Boston at the MIT Sloan Sports Analytics Conference. “[They’re] useful for support but not for illumination.”
His analogy is a bit extreme, but it’s a feeling likely felt by many NHL decision-makers who look at mental makeup as much or more as they look at a stat line.
“There has not been a statistical breakthrough that I’ve seen in hockey,” Burke added. “Everybody is looking for these Moneyball breakthroughs. … I have yet to see anything that has value in terms of an alternative way of evaluating players.
“This whole Moneyball thing aggravates me. It’s horse [expletive]. No one’s won a championship with Moneyball.”
Unsurprisingly, executives like Burke who grew up around and in the game, often prefer to look at intangibles that cannot be measured with stats when making personnel decisions. They want to know whether or not a player can bring his best when it matters most, which is, of course, something that is not easily quantified by numbers.
Still, there are some in the game who do embrace stats, even if it’s not to the extent that statisticians like Michael Schuckers, an associate professor of statistics at St. Lawrence University, who specializes in hockey and was also a member of the Sloan panel.
Some of the stats that do matter — to NHL execs that is — have more to do with the vitals of a player, rather than tabulating thousands upon thousands of scenarios trying to find a pattern.
Bruins general manager Peter Chiarelli said that some of the stats he looks closest at have to do with the size of a player. Chiarelli added that he tends to look at the weight of a player more over the height.
One look at the Bruins roster and this will be validated. Those bigger, stronger types of players are the same types who are strong on the puck and allow a team to possess the puck more frequently. If you win the possession battle, you’re likely going to win the game. Or put even more simply, you have to have the puck to score.
Chiarelli also pointed to the Detroit Red Wings as a model franchise when it comes to assembling a team that is built to win. The Red Wings possess the puck better than any team in the league, and the results usually follow. It should also come as no surprise that Pavel Datsyuk, who is considered by some as the best player in the world right now, was recently voted by his peers as by far the toughest player to take the puck from.
Still, statisticians like Schuckers are hoping there is further evolution in the game when it comes to statistics. He discussed the possibility of a wins over replacement stat — one that has swept through the baseball world — could soon be prevalent in hockey.
“As we get better data, we will have more advances,” Schuckers told NHL.com. “Part of it is getting out here and speaking. We [hockey analytics researchers] have to get ourselves heard.”
Stat-heads and old school talent evaluators can at least agree on one thing: Plus/minus is misleading when it comes to evaluating talent. Or, as Burke put it: “If you play on a horse [expletive] team, you’re going to have a bad plus/minus.” On the other hand, if you play on a good team, you’re probably going to have a strong plus/minus. Thus, Tyler Seguin (plus-25), Bergeron (plus-29) and Zdeno Chara (plus-26) are all in the league’s top 10 in that stat.
While there may be a future marriage between the new school and the old school, listening to someone like Burke and Shuckers debate the merit of faceoff wins shows you that there’s still a considerable gap in the two schools of thought. Schuckers stated that a team would have to win an extra 100 faceoffs to add one goal. Burke, of course, gruffly responded that he had seen plenty of instances where faceoffs turned into goals.
Yet, if we get back to Bergeron, there’s a reason why is widely considered one of the best two-way forwards in the NHL. He does a little bit of it all. His faceoff skills help garner puck control, as does his strong defensive play in all three zones of the ice. But Bergeron also possesses the intangible skills that analytics may never be able to quantify.
The real trick may be to be able to find the best combination of both evaluation metrics and making a decision off of that. The difference between being good and great, may be decided by which teams are able to use both. If that’s the case, Brian Burke may need to heed his own advice.
“If all you wanna be in sports is good, you’re gonna get your ass kicked,” he said.
As has been the case in other sports, whoever can find the right formula will be the most successful when it comes to the one stats that both sides of the room can agree matters most — wins and losses.
