August 07 2013 04:31AM
The Ottawa Senators will be the best team in the league, the Toronto Maple Leafs were the luckiest in the league last season, Alexander Mogilny is Steven Stamkos' closest historical match and Toronto is expected to be 28th in the league next season.
This is all buried within Rob Vollman's first self-published Hockey Abstract, available in print at Amazon or as a .pdf. No e-book version is available, but if you have a tablet or an e-reader that's comfortable loading .pdf files, you'll be fine. Vollman says the novella isn't a textbook, and it makes for a much easier read than anticipated, simply attacking the same problems in a different way rather than attempting to write a treatise between stat-oriented and viewing-oriented hockey fans.
Among the questions answered: "Who is the best defensive player?" "Who is the best goalie?" It's not anything groundbreaking in that regard, and fans who have been reading Vollman's work (he's a major player in the hockey analytics blogosphere) won't be blown away by too much new data, but it's a good reference point with how to attack these problems with the data we have in 2013. Hockey Prospectus' proprietary GVT "Goals Versus Threshold" ranking is leaned on heavily here, as is data from BehindTheNet.ca and HockeyReference.
Consider his section "Who is the best defensive player?" Vollman doesn't start with a conclusion ("Datsyuk") and find data to weave around, but looks at a fair number of different methods, tallying up Selke Trophy votes since 2008-2009, counting defensive GVT, and determining the percentage of minutes a player has spent killing penalties for his team. His conclusions in each section are not deterministic, but there's a lot of evidence to lean towards Patrice Bergeron being the best all-around defensive player, or Henrik Lundqvist the best goaltender or Sidney Crosby as the league's best playmaker. Much of the data presented shows many ways that the above players are effective, and the tables presented are deep enough that you can find some interesting comparables when looking at players a certain way.
Samuel Pahlsson and Jordan Staal for instance, are noted "defensive specialists" that frequently take on heavy defensive zone and quality of competition usage, but next on the list is Tampa Bay utility-man Nate Thompson and new Leafs acquisition David Bolland. It's hard to think of Shawn Horcoff as a defensive specialist, but he's been handling tough minutes in Edmonton for years now with all the rookies and his scoring has suffered.
One thing that this book has that you can't find many other places just yet is the "passes" statistic that Vollman developed this year, looking at a player's percentage of overall assists and comparing it with the number of on-ice shots for in an attempt to find the number of times a player set up a shot. At the top of that list sit Sidney Crosby and Joe Thornton. I'm a fan of that statistic and like where it's going.
There's no colourful prose to heighten poor arguments and the statistics used are worth more in the "well, if you looked at it this way, you'd find..." which is just as valuable. My own opinion isn't that analytics are the be-all and end-all of player evaluation, but it's important to find data that contradicts an assertion you've made, since the data at least makes you think about the viability of your own argument.
Consider the unique "quality starts" table found towards the end of the book. Even though both Jonathan Bernier and James Reimer are listed as having the same save percentage (.9150) since 2007-2008, Bernier has quality starts in 64.8% of his appearances while Reimer just 48.0%. I like quality starts as a metric for goaltenders better than wins and losses because it better tracks consistency, takes out some of the team influence of the statistic and also tests a goalie's longevity where simple save percentage wouldn't.
It would be a more complete resource if there were additional evidence or arguments or an attempt explanation for quirks like that. Vollman did himself no favours by titling his work "Abstract" like Bill James did for his groundbreaking work in the 1970s and 1980s since that invites the comparisons to the books that changed the way people look at sports. James' original Abstracts had a brief essay about each team, but Vollman has stayed away from that, focusing on general problems.
There's a lot of meat and Vollman attempts to attack problems in different ways and offers a lot of tables and illustrations not to prove points, but to introduce another element into the discussion. As a result there were some tables that focused too much on certain statistics (such as GVT, a catch-all stat I'm very iffy on) and glosses over others I'd like to see in more depth, such as penalty differential. At 232 pages, it's quick to get through, ideal for a flight or a road trip or a freak thunderstorm.
This is a good record of what we have right now in the way of data to judge hockey players, one that blends the objective statistics found online with the subjective opinions of writers and coaches close to the game. If you're at all interested in hockey data or looking for a resource to summarize the first ten or so years of dedicated analysis, pick it up.