Never has a headline struck me as so patently false as, "Nashville Predators evaluate players with advanced stats."Josh Cooper's story tells how David Poile really connected with the book Moneyball, about the startling success of the small-market Oakland A's in the early 2000's. The book, of course, details how Billy Beane and his lieutenants used advanced statistics years before other front offices, and leveraged that knowledge to sign undervalued players to cheap contracts.
The appeal of such a story to the General Manager of the historically cash-strapped Nashville Predators should be obvious. David Poile might even rightly even see himself in the book's protagonist, Billy Beane. Turning Long Island ball hockey superstar Joel Ward into the 3.5 million dollar man has to be a feat as impressive as anything involving Scott Hatteberg. Though I'm not sure Brad Pitt would play Poile in the movie version (maybe Kevin Bacon?).
I even believe that Poile knows of hockey's equivalent to sabermetrics. And I know there are people within the Preds front office that understand such stats.
Still, to read a story about how David Poile uses advanced statistics to evaluate players is a little odd, since every player they've brought into the organization in the last 2 years has had terrible advanced stats. Here's 2 years worth of transactions and their Corsi number for the Preds last season:
- Paul Gaustad (trade): -11
- Andrei Kostitsyn (trade): -4
- Hal Gill (trade): -9
- Brandon Yip (waivers): -15
- Niclas Bergfors (signing): -14
- Jack Hillen (signing): -11
- Marek Svatos (waivers): -4
- Mike Fisher (trade): -13
- Sergei Kostitsyn (trade): -9
- Matt Halischuk (trade): -16
- Francis Bouillon (signing): -9
Now look how the players they dumped in that same span fared this year:
- Jerred Smithson (trade): -16
- Blake Geoffrion (trade): -13
- Cal O'Reilly (waivers): -12
- Cody Franson (trade): +2
- Matthew Lombardi (trade): -9
- Shane O'Brien: +8
- Joel Ward: -2
- Steve Sullivan: +14
- Marcel Goc: +5
- Alexander Sulzer: +5
- Mike Santorelli: even (0)
- Dan Hamhuis: +10
- Jason Arnott: +9
Listen, I'm not saying that all those acquisitions were bad, or that they should have paid top dollar to keep Ward or Hamhuis. Obviously, there are mitigating circumstances and other dimensions of analysis to consider with every move listed here. And Corsi only tells about part of a player's value.
Still...Corsi number measures to what an extent a player's team does or does not get outshot, when that player is on the ice. And when, over the course of two years, you dump nearly all your players with good Corsi ratings and only bring in players with bad Corsi ratings -- no matter the other considerations -- you might find your team getting outshot, often and by a lot, like the Predators were this year.
To have success while getting outshot every game, as Gabe Desjardins talks about in Cooper's article, you basically need the world's best goalie and special teams. And thank the hockey gods the Predators did! But that kind of success is hanging by a very thin thread. Nothing varies year-to-year as much as goalie performance and special teams play. Just look at the SV% or PP% leaders from one year to next. Always a different list.
And besides this brand of success being very fleeting, it's also very obvious. Moneyball was about finding players who did things that were subtle, but no less important to winning games. Finding a player who can rip a wrist shot on the power play or block a shot in the defensive end? Not subtle. Bringing in a Joel Ward type, who controls the end boards, making sure zone-exits are smooth? Subtle.
I wrote about the Predators and this oft-made comparison to Moneyball in October, when the movie was coming out. In my piece, I said the Predators more closely resembled the villains of the book (the indignant scouts and media), than Beane and Paul DePodesta. If you'll allow me to quote myself from that article:
We can even make a simple application of Walker's theory to hockey: if hockey is equally about taking shots and preventing shots--and a team taking a shot is necessarily, at that moment, not allowing a shot--forward offense is hugely more important than forward defense.
If there's a correspondant Oakland team to today's Predators, it's the mediocre A's of the past five years, not the hugely successful Moneyball teams of the five years prior. Since the Red Sox made traditional Moneyball tactics work for a big market team, Oakland (among other teams) tried to stake out a new niche in building defense-first teams.
And while teams like Oakland and Seattle no doubt reached their goal of evaluating defense better than the rest, they betrayed their ultimate goal: winning. Oakland's slick-fielding, low-OBP speedesters dominate that 5% of the game. But alas, it's still just 5%.
Today's A's teams are hard to score against, but beatable. Sound familiar? Unfortunately for Billy Beane, he's running out of options--On Base Percentage is now the opposite of a secret. Fortunately for the Predators, hockey is decades behind baseball, in this regard.
I'm not calling David Poile a liar here, he probably does use Corsi numbers -- or whatever -- to compare players on his own team. But to really accept the message of the advanced stats would entail critically rethinking fundamental tenants of his player personnel philosophy. And seeing as the Predators have doubled-down on the extremely conservative "Predator Way" in recent years, I don't think that's what's going on.
On the other hand, David Poile's obviously not actively working to go against advanced statistics with his transactions. So with hopefully a better offseason -- and a step forward from young Corsi monsters like Ryan Ellis, Craig Smith, and Gabby Bourque -- the Predators could be a more well-rounded team next year.