In today’s analytics-driven game, statistics surround the world of hockey. Everything from shot quality, to score situation, to where netminders give up the most goals is tracked by statisticians and made available to organizations and, in some cases, the public.
With all of this data, it is sometimes difficult to sort through what is valuable and what isn’t. But irrelevant or misleading numbers have been around since before the analytics movement began.
Even mainstream markers have their issues, and here are five of the most flawed statistics commonly used in the game.
The end-game in hockey may be to win, but judging goaltenders based on wins and losses is something that does not make sense.
Hockey is a team sport, and while a netminder can occasionally steal a game for his club to secure a victory, he cannot do so consistently. A good goaltender can’t accumulate wins if there is no offense or defense in front of him.
Take Cory Schneider, for example.
Schneider has historically been one of the best netminders in the NHL. He owned a .924 save percentage and a 2.15 goals-against average last season with the New Jersey Devils. That ranked him fifth and fourth in the league, respectively, among goaltenders with at least 40 games played.
Despite those elite-level numbers, he had just 27 wins in that campaign; that was bad enough for 13th in the league.
The Devils were unable to support their netminder with sufficient goal scoring, and Schneider’s win totals paid for it.
Wins can also be inflated if a goaltender plays a high number of games. Jonathan Quick was second in the NHL in wins last season, but he also had the highest amount of starts. Win percentage would be a better statistic to use in this situation if you do buy in to the value of goaltender wins.
Hits are a misleading and misvalued statistic for two reasons:
First of all, they are difficult to track with consistent criteria. What one person counts as a hit may be different from what another person counts as a hit. Therefore, there is often a large discrepancy between totals from game to game.
Simply put, the count is unreliable.
The more important reason hits aren’t always great is because they have an inverted correlation with puck possession. If a player has to make a hit, it means said player did not have the puck to begin with.
In today’s game, puck possession is everything, and a high amount of hits shows that the player or team doesn’t have the puck as much as they should.
Physicality is a vital part of puck recovery, but having to use it too often should be alarming.
This statistic is on this list for almost the same reason as hits. If a player has to block a shot, it means he didn’t have the puck. Worse yet, it also means he gave up a shot attempt—something teams are striving to prevent.
Blocking shots is an important part of hockey, and sacrificing the body is solid team defense. The opponent can’t score if the puck never reaches the net.
But again, a high number of blocked shots means a player doesn’t have the puck as much as he should. Having to block occasionally is fine, but too much should be a cause for concern.
Let me begin by saying that giveaways are never a good thing. However, the issue with the statistic is how it is expressed.
The current giveaway leaders for the 2016-17 season are Brent Burns, Erik Karlsson, and John Carlson. These three players are highly-talented defensemen who are vital to their team’s success, yet they give the puck away more than anyone.
The reason their giveaway numbers are so high, though, is because they touch the puck more than most players. Of course Burns is going to lead the Sharks in this category; he has the puck on his stick every time he’s on the ice. He has far more opportunities to give the puck away than any of his teammates.
Other resources, like those which track zone exits and zone entries, more accurately display this kind of information. But expressing giveaways as a raw number rather than a ratio to puck touches is a flawed methodology.
Faceoffs are the first battle of every shift and—for the most part—dictate which team starts with the puck. There are many crucial moments centered around a single faceoff, particularly in the dying seconds of the game.
It’s not that faceoffs aren’t important, it’s that they are not important as some think they are.
Antoine Vermette of the Anaheim Ducks has won 65.9 percent of his faceoffs this season. However, he has negative possession and goals-for percentages. Meanwhile, Evgeni Malkin of the Pittsburgh Penguins has only won 43.4 percent of his draws and has positive numbers in terms of possession and goals.
There are many other instances where this relationship happens, meaning the value of faceoffs is a bit more ambiguous than once perceived. They are vital in certain situations, no doubt, but just how vital remains to be seen.
Honorable Mention: Plus/Minus
The analytics community is not a fan of plus/minus, and therefore it earns the honorable mention spot on this list. There are just too many factors that can work for or against a player in this marker that are outside of his control.
Typically, plus/minus would have earned a regular spot on the list, but the misleading nature of it is fairly well known. The other statistics are more heavily used without any noticed inaccuracy.