Recently, I expressed my frustration with the lack of advanced stats being discussed in a meaningful way by many analysts on TV broadcasts of NHL games. My main point in expressing that frustration was to point out that many hockey fans are not on social media and often do not have access to or knowledge of the resources available to explain stats they may not be familiar with. I was advocating for the inclusion of these stats in broadcasts so that people who are unfamiliar with them could feel more comfortable with them and get some exposure to them in an accessible setting. This level of comfort would decrease the pro-stats versus anti-stats strife and be better for everyone in my opinion. Unfortunately, some read this as me trying to put down people who are not on social media or who aren’t into stats, which was not at all my point. Perhaps I could have done a better job making that clearer but apparently I fell short in that regard.
I received many and varied responses to that post but one of the best basically said that when more NHL teams admit to using them, fans and media would begin to feel more comfortable accepting advanced stats as a useful tool. With that said, I was intrigued this past week by new information regarding the use of advanced stats by some NHL teams. While many in the “advanced stats crowd” have known for a while that there are several teams in the NHL that use their analytics budget for more than just tracking hits, information came out this week that expanded upon that and more importantly, the information was publicized by NHL.com thereby reaching a far larger audience.
While doing interviews with the coaches and players of the San Jose Sharks and L.A. Kings, NHL.com writer Corey Masisak (on Twitter @cmasisak22) asked questions about puck possession and the teams’ use of stats. He was met with a fairly frank discussion that led to confirmation from both teams that they do in fact track puck possession data and use it in their strategies, player usage, etc.
During media availability for the Chicago Blackhawks and St. Louis Blues series, Blues head coach Ken Hitchcock admitted that both he and Blackhawks head coach Joel Quenneville use advanced stats for their teams. On a much smaller level, I was able to get information from a woman who knows one of the Blackhawks analytics people regarding the team’s use of advanced stats. She informed me that the Blackhawks internally track Corsi, Fenwick, zone entries and exits, among other data. This information is then presented to the coaching staff for their use. Obviously, my Twitter feed is a drop of water compared to the ocean that is the reach of NHL.com and the many other media sites and publications covering the NHL, so I do not expect that to have anywhere near the impact of the aforementioned information.
Regardless, this advanced stats usage by NHL teams such as the Sharks, Kings, Blues and Blackhawks being mentioned and explored by major media sources is a big deal. It is very intriguing. I write about and tweet about advanced stats regularly and sometimes meet with resistance to the information I disseminate. I wondered if perhaps seeing that these very successful teams use advanced stats would sway the opinions of those who had been skeptical as to their usefulness. I took to Twitter and asked for thoughts on the matter. I also made sure to explain that I was not trying to insult or make fun of anyone in order to encourage more honest responses. Obviously my audience, for the most part, are pretty accepting of advanced stats but some of them are not.
The majority of the responses I received from the skeptics boil down to this:
Numbers and stats are fine but they are not the “end all be all” of hockey analysis. You cannot just rely upon numbers to decide who the good players are. The eye test has to agree for the stats to mean anything.
Obviously, I am paraphrasing and summarizing here because several people really took the time to lay out their thoughts and, I would like to add, the majority who responded did it in a very respectful and congenial manner.
The “End All Be All” was the most widely mentioned criticism of advanced stats in the tweets and emails that I received. There seems to be a wide spread notion that folks in the advanced stats crowd begin and end all analysis with the numbers. Perhaps it comes across this way because there just are not that many people in each team’s fan base, at least for the American teams, that focus on this information. Being one of the very few who do stats analysis for a team may lead people to think that stats are the only thing that particular person looks at when in fact it is simply the lack of that type of information from other sources making it stand out.
For the past year, I have followed along on Twitter and frequented the websites and blogs of as many advanced stats people as possible. That is a very short time and happens to coincide with the amount of time I have been on Twitter. In that short time, I have learned an amazing amount from people who have revolutionized stats in hockey and people who use stats as part of their analysis of the teams and players. In that year of paying close attention to and having frequent discussions with these people, not once have I ever heard even one of them say that advanced stats are the end all be all in analyzing hockey. Never.
I think the “end all be all” argument against stats is more a function of perception than reality. Often the advanced stats pack provides a counterargument to more “traditional” analysis by citing ways that the numbers show a different reality than what is being alleged. If the majority of what you have heard in the way of hockey analysis is not numbers related and then it is attacked based upon numbers, it makes sense that you may develop an aversion to the numbers. Again though, I have never heard any in the advanced stats crowd say that numbers are the only way.
I know of several hockey sites that do a wonderful job of breaking down plays and strategies. They are not all that plentiful so people have to search them out. Hockey is so often presented as hit, shoot, score that the actual strategy to the game is left by the wayside fairly often. Further, there are people who are doing tracking projects for zone entries, zone exits and passing. These manual tracking projects require a lot of time to carry out and frankly, I do not think it will shock anyone to learn that not every NHL team has a member of the public that does this.
The long and short of it is this: Hockey is a tough game to analyze. It is more fluid than many other sports and lacks the technological applications that some other sports have to help track data. Even the data provided to the NHL is tracked by people, so as fans we have to make do with what we have. No, Corsi and Fenwick are not perfect, but they are the best tools we have at this point to analyze puck possession and puck possession is the name of the game. All told, I think that the more people are hearing about teams using advanced stats, the more they are coming to accept that they are useful, but they still have some reservations.