The most well-known book about advanced statistics in baseball is Moneyball, which is odd, as it isn’t, strictly-speaking, about advanced statistics. Instead, Moneyball is a book about economics and finding market inefficiencies. It just so happens that in baseball, those market inefficiencies are generally found through the use of statistical analysis. In many ways, that analysis has its roots in the work of Bill James and his annual Baseball Abstract that was published from 1977 to 1988.
As advanced statistics in hockey grow in prominence, there appears to be a market inefficiency of sorts: there is no equivalent to the Baseball Abstract for hockey. While plenty of material has been published online developing statistics like Corsi and Fenwick and tracking things like zone exits and entries, no one has published a book covering these statistical developments in a way accessible to those unfamiliar with the work done online.
Rob Vollman has attempted to to fill the gap with his new book, Hockey Abstract, available in PDF format or in print from Amazon. While it falls short in some areas, it’s a fantastic resource for those new to advanced statistics in hockey and an engaging and enlightening read for those already familiar with them.
Vollman was kind enough to provide a PDF copy of his new book for me to review. Regrettably, there is no ebook version available, but I gamely loaded the PDF onto my Kindle and found it readable. Those with below average eyesight would be better served buying the print version or reading the PDF version on a computer.
Vollman has the appropriate credentials for a book of this type: he’s a founding writer at Hockey Prospectus and contributes to ESPN Insider. My favourite of his statistical work is the introduction of Player Usage Charts, which provide an easy to understand visual representation of a player’s role and general effectiveness on his team, but Vollman has also done excellent work with concepts like Quality Starts for assessing goaltenders and history-based projection formulas to help predict a player’s future performance.
By titling his book Hockey Abstract, Vollman invites comparisons to Bill James’s classic baseball sabermetrics annuals. Considering the influence that Bill James has had as a pioneer of sabermetrics and the influence that the Baseball Abstract books still have to this day, despite the last one being published 25 years ago, that’s a dangerous comparison and Vollman falls somewhat short. When taken on its own merits, however, it fares much better. Still, it’s worth drawing a comparison to James’s work.
Bill James, when outlining the difference between sabermetrics and sportswriting, said in his 1981 Baseball Abstract, “Sportswriters characteristically begin their analysis with a position on an issue; sabermetrics begins with the issue itself.” As a result, much of James’ writing starts with a question — “What’s the greatest pitching performance of the modern era?” or “Which of these two MVP candidates had the better season?” — then attempts to answer that question with objective, statistical analysis, frequently developing an entirely new methodology in the process.
Similarly, Vollman’s Hockey Abstract is structured to answer the basic questions that a hockey fan may have about the NHL, with chapter titles like “Who is the Best Player?”, “Who is the Best Playmaker?”, and “Who is the Most Undervalued Player?”, using those basic questions to introduce the statistical concepts that attempt to answer them. It’s a format that shows these statistics are not arbitrary, but have been created in an attempt to create an objective answer to questions that are generally answered subjectively by hockey fans and media.
It’s also an approach designed to be friendly to a reader unfamiliar with these statistics, as well as promote discussion. As Vollman says in his introduction, “I love hockey arguments — I want to refuel the conversations, not end them!”
This humble approach continues throughout the book, as Vollman is quick to point out the limitations of various statistics and the need for context. He frequently approaches the questions asked in each chapter by looking at subjective opinions, such as Hart, Vezina, and Selke voting, pointing out the importance of giving credence to the opinions of those who have spent a long time watching the game of hockey, before using statistical analysis to show what players might have been overrated, underrated, or missed outright.
Vollman doesn’t quite have the wit or breadth of insight of Bill James, whose opinionated and colourful writing style was a big part of the appeal of the Baseball Abstracts, but Vollman’s writing style is still eminently engaging. He breaks down the issues in each chapter clearly and concisely and introducing the various statistics slowly, rather than flooding the reader with charts, esoteric numbers, and acronyms all at once.
I would, however, have liked to see is a bit more in the way of precise definitions of statistics and the formulas behind them when they are first introduced. In general, when a new statistic is introduced, a note is included stating that the particular statistic will be delved into in more detail in a later chapter.
I understand Vollman’s desire to create a book that can be read and enjoyed by a casual reader with a passing interest in statistics, but showing some of the calculations involved in a statistic like GVT (Goals Versus Threshold) or even something as simple as Corsi would aid in understanding rather than being a roadblock. A reader who picks up a book on advanced statistics in hockey likely wouldn’t be turned off by the occasional simple formula or slightly more in-depth definition dropped into a chapter.
This is a minor criticism, however, and one that did not detract from my overall enjoyment.
When Bill James wrote the first Baseball Abstract, he was essentially speaking into a vacuum. There were very few people doing the kind of work he was doing at the time, with no internet to disseminate those ideas. As a result, James’s work was the first of its kind to be (relatively) widely read.
Vollman’s Hockey Abstract speaks into a different world. While it seems to be the first publication to place all of these statistical ideas in one place, there is a large community online that has been discussing and developing these ideas over the past 6+ years or so. Bloggers like Vic Ferrari, Gabriel Desjardins, David Johnson, Eric Tulsky, Vollman himself, and many others have helped develop many of the concepts in the book and have themselves been widely read online.
This both benefits and detracts from Hockey Abstract, as it gives Vollman a large foundation to work from but also limits the potential for innovation, given that this is an initial work that seeks to serve as an introduction. I was already familiar with most of the statistical concepts used throughout the book and those who regularly read myself or Cam Charron’s work here on Backhand Shelf will be familiar with many of them as well, as Vollman touches on Corsi, Fenwick, PDO, and Penalty +/-.
That said, there was definitely plenty of material that was new to me, such as the introduction of Vollman’s Passes statistic that he uses in his attempt to answer who the best playmaker is, and the statistics that I was familiar with were used in intelligent and innovative ways. The synthesis of various statistics into categorical indexes was particularly interesting, including the fun “Do-it-all Index,” which is used to help identify the league’s most undervalued player.
All-in-all, I thoroughly enjoyed Hockey Abstract and certainly recommend it to anyone curious about a more objective approach to hockey or who would like a one-stop guide to the burgeoning field of advanced statistics in hockey.