FRB Book Club: Rob Vollman's Hockey Abstract






As each season passes, the field of statistical analysis in Hockey gets more prominent and essential. Rob Vollman, long regarded as one of the pioneers in this field, set forth in an ambitious endeavor to create Hockey's companion to what Bill James did with his innovative and revelatory Baseball Abstract; to establish a guidebook to the fundamentals of metric-based opinion. Shot differentials, quality of competition, quality starts for goalies, and the new passing metric are all covered within. As a hockey fan with a somewhat decent grasp on these concepts, I decided to drop some scrimp at Amazon and check it out.
By now we are fairly familiar with things like Corsi and Fenwick numbers, zone starts, usage charts, etc., and Vollman uses these to answer 10 questions he puts forth; who is the best *insert position or player type*, coach, etc. By applying these measures to build and support arguments, the reader discovers things about certain players, whether it is good or bad- whether they are overvalued, undervalued, or getting paid just about right for what they bring to a team. The questions are as much about finding the answers, but also illustrating how Vollman got there and with what metrics.


As someone who scouts various leagues in the Upper Midwest region, what I find most intriguing about this book is how I can apply some of the philosophies in the book (which Vollman definitively states are a starting point, not the conclusion) to how I view and value players- at some point after diagnosing the mechanics of the player, you start to look for results. The newly debuted "pass" metric- counting how many pass distributions from a player lead directly to a shot attempt (and surprise surprise, Sidney Crosby lead with roughly five "passes" a game)-  is particularly intriguing. It isn't necessarily about replacing the live viewing option, but about tracking elements to find out if a player, regarded as this or that, is actually expressing that trait during the course of the game. Since its been proven that shot attempts = goals = wins, you look for players that do the things that lead to that; and some players may be overrated in that regard, while others go unrecognized for their efforts, which can lead to diamonds in the rough.

I don't have too many qualms about Hockey Abstract; at times, for me of the short attention span, the barrage of charts and numbers can be overwhelming and the numbers becoming monotonous. I found myself keeping a thumb in the glossary at times, reading a chart and then flipping back to try and remember what exactly the metric measures. There really hasn't been a definitive metric established for defensemen quite yet; so instead of finding how how well a D can, well, actually defend (blocks, takeaways, deflected passes, making plays that create possession interruptions) a combination of things like Relative Corsi and Zone Starts are used. That said, there has been some headway in how well a player exits the zone and turnovers/touches, so maybe a more blueliner-oriented stat is on the horizon. And Wild fans won't be pleased that the team is projected for 82 points, which would be fifth in the new Division.

Maybe the most important part of Hockey Abstract is in the Introduction, with Vollman saying this:

"It is also important to remember that statistical analysis was never meant to replace anything and no one in this field has ever suggested otherwise- probably because that would be mistakenly ridiculous. This type of objective analysis makes an excellent companion to the more traditional ways of studying the game and never has, and never will, act as a replacement."
For what seems like an unending misunderstanding between some of the proponents of each side, this is important and should be recognized as such; both elements act in unison in the development of opinions on players- "Player X is a good player, and here are the numbers to show why." Just like a scout can watch a player and find that the raw numbers agree or disagree with his assessment, someone can look at the numbers and then see the player do the things that the data would suggest. There is no such thing as too much information, in my opinion, in regards to how we view the world of hockey; and Hockey Abstract is a fantastic way to begin that union between raw data and how we view the game with our naked eye.

No comments:

Post a Comment