This article isn’t about statistics.
This article isn’t about PDO, regression, Corsi, Fenwick,
save percentages, shooting percentages, or any of the other advanced analytical
categories that are bandied about so often.
This article is about hockey.
Specifically, this article is about a book about hockey
written by a guy who loves to use statistics. So yeah, maybe there’s some stats in the article, but what I’m
doing today is telling you about Rob Vollman’s Hockey Abstract Update 2015-2016, and why it is as much a hockey
book as it is a primer for advanced analytics.
Some of you may know of Vollman, he of the ever-popular
Vollman Player Usage Charts, a precursor for simplified visual rankings such as
the more recent HERO charts. He’s also a former contributor to FlamesNation from
a few years’ back with his weekly Black Box articles, which provided a treasure
trove of raw possession data on Flames players.
Vollman has been writing on Hockey Prospectus since its
beginning, the two most recent McKeen’s Hockey issues and written his own Hockey
Abstract going on two years now.
Fair to say, this guy loves the game.
This year Vollman will be moving from self-publishing to a
professional publishing arrangement with ECW Press, a small Canadian press
devoted to Canadian literature, poetry and non-fiction. As such, future
iterations of Hockey Abstract will be available in major book stores coming fall 2016. PDF updates will be available for download for those who have purchased the print copy to ensure that the data and analysis within is completely up-to-date.
Vollman’s Hockey Abstract Update 2015 is a book about hockey
through the lens of statistics. Certainly it is something those of you who are
comfortable discussing PDO regressions and NHLE predictions would feel at home
with. But for those of you who want a hockey book to be, well, about hockey,
you can find it here as well.
Who’s the best goalie in the league?
Datsyuk or Bergeron?
Weber or Lidstrom?
Why isn’t Lindros in the Hockey Hall of Fame?
Who’s going to win the Stanley Cup this year?
How will the Flames do this season?
These are basic water-cooler type questions from hockey
fans. And every one can be approached (though not necessarily answered) by Vollman’s Hockey
The information provided in the 163 pages of this update is,
to put it modestly, eye-opening.
Of course, many of us all familiar with the basics of Corsi,
Fenwick, and the terms that circulate around the analytics field about
possession metrics such as sample size, regression and so on. What Vollman
provides in his update is not only a greater level of detail in these
statistical categories, but also a very readable and relatable interpretation
and explanation of many of these terms.
There is the common complaint against the analytics movement
that it is all number crunching removed from the passion of the game, wrestling
the thrill and excitement of the experience away from the fans and putting it
into a cold, clinical environment, as it were.
Nothing could be further from the truth and Vollman
explicitly says as much on several occasions. He argues that the best metrics
usually help to tell you things that you probably already believed in your gut,
that is they resonate with the instincts of an informed fan.
Patrice Bergeron is a terrific 2-way center and deserved
Olympian, right? Well the numbers back that up and Vollman shows you why. Kessel is a top-scoring threat on the wing who tends to be more of a one-dimensional
player? Again, the numbers would support that assertion.
Then why do we need them? Because sometimes our eyes lie to
us and we need an objective check against our own biased impressions.
The other aspect of Vollman’s work that has value is in
identifying the really potent players who often fly under the radar. The poster
boy for that around here is, of course, Mikael Backlund. Kent Wilson began
waving that flag years ago based on his understanding of Backlund’s possession
metrics which were outlined by the same work Vollman details in his Hockey
Now, you’re probably thinking that anytime a book spends
160+ pages talking about statistics, even hockey statistics, it must be drier
than a Baptist Barbeque in Lynchburg. Not true. He makes this a
surprisingly easy read for the layperson, occasionally layering in a personal
touch. Case in point, when discussing his Quality Starts goalie
metric being taken from baseball’s equivalent, Vollman frames it thus: “I
almost immediately applied the same concept to hockey, despite being just 10
years old and having to count them by hand using box scores in the newspaper.
And yes, it was a number of years after that before I kissed my first girl.”
My point is this isn’t a boring, pedantic snorefest. Vollman
writes in an intelligent, yet expressly comprehensible fashion to best communicate
a given concept.
Enough about the writer. He’s good, does excellent work and
for anyone interested in hockey analytics is worth paying the price of his
But what does his HA 2015 Update actually cover? I’m not
going to reveal all the data or the interesting tables that lay out which
players are (surprisingly or not) the best in various categories or which teams
have untapped potential based on their underlying numbers.
Instead, I’m going to throw out some teasers by describing
some of the topics included.
Hall of Fame
Inductinator by Iain Fyffe – a predictive model designed to determine which
players are most likely to make the Hockey Hall of Fame based on historical
Goal stats, including
Quality Starts and Home Plate sv% – some new metrics that might just lay
the groundwork for cracking open the enigma that is goaltending. I’ve said the
first guy to figure out goalies in a meaningful, predictive analytical model
will have a 40ft statue made in his honour. It has been said that goalies are
No, we can understand voodoo.
Goalies are when your wife/girlfriend stopped speaking to
you sometime between dinner and dessert and on the drive home from the
restaurant you have no idea why but only know that you are way too afraid to
(We wish we had Fenwick, PDO and Quality Starts % stats on our relationships with women)
Coaching data –
often seen as unquantifiable outside of systems and deployment, the Update includes some interesting
information about coaching records.
Standings – the initial steps at hammering out a better predictive model
for league standings that takes “luck” out of the equation. Better than
throwing darts at a board, which is about where the industry standard is today.
Setup Passes –
Corsi, Fenwick and all of its derivatives measure shooting but there is very
little room in the analytics world right now for pure passing data. One of the
most important skills in the game sits in the shadows and here Vollman and
company begin the process of deciphering that data and creating a usable
Do-It-All Index –
A measure for the most effective players who display the widest range of skill
sets in the league. Coaches love them and teams often win on the backs of these
He also introduces some
new statistical categories which are being explored by he and a number of
others working in the field, including Tom Awad, Michael Parkatti, Stephen
Burtch, Micah Blake McCurdy, Tom Tango and a host of others.
Some of these stats may fall by the wayside as they prove to
be either too cryptic or redundant, while others could well grow up to dominate
the landscape the way Corsi and Fenwick have. These include Adjusted Corsi,
Delta Corsi, Score Effects, Bayes-Adjusted Fenwick Close (a predictive
possession metric), SWaSH (Score Adjusted Weighted Shots which may prove to
have greater predictability with forwards than Corsi), Expected Goals, Adjusted
Save% or xSV%, WAR (Wins Above Replacement which hopes to form a catch-all
statistic), and Prospect/Draft analysis that hopes to expand on the traditional
NHL Equivalency method.
Rob Vollman’s Hockey Abstract Update 2015 is an
exceptionally informative and surprisingly readable resource for any fan,
whether you’re involved in fantasy leagues, blog about hockey, or are just a
Vollman expands on the most basic concepts of hockey
analytics by providing easy to understand descriptions of the concept that is
being measured, how it relates to hockey, and what impact that can have on a
I’m not given to recommending specific book purchases, so
when I say that Rob Vollman’s book is worth the investment for even the casual
hockey fan with an interest in analytics, please take that to heart. You can
purchase a copy of Vollman’s Hockey Abstract here.