Over the past couple of years the development of advanced metrics has exploded.
With more people talking with regularity about corsi, fenwick, zone starts, and PDO it has become a little more accepted. It obviously still has its critics, and it still is in a stage of infancy with general public acceptance.
This season the NHL implemented a huge overhaul in its stat gathering and stat posting page on NHL.com, with reference to most, if not all, of these advanced metrics. While they were under different names, the meaning was clear. Advanced stats have gained a foothold in a very traditional game which is somewhat reluctant to open itself to new age thinking.
However, it has not really stopped for those who crunch the numbers, ponder the possibilities and try to dream up bigger and better evaluative properties for our great game of hockey. If Baseball is the holy grail, with all of its sabremetrics, hockey wants to be equal or close to that.
One of the major stats in Baseball that is currently under development with regards to the NHL is WAR, or Wins Above Replacement. WAR in baseball is essentially the stat that tries to hone in on just how important a player is to his team. While it is imperative to use more than one statistic when evaluating a player, WAR has quickly become one of the more telling statistics when it comes to showing the true value of a player and what he contributes to his team.
In recent years there have been developments of multiple parties in regards to hockey WAR.
SB Nation's
Artic Ice Hockey and
BehindtheNet founder(?), Hawerchuk, has opted to evaluate with WOWY charts, or With or Without You charts. These give us a good idea of what players contribute on their own, or with certain partners, in corsi, goals for, fenwick, etc. etc. We have used them on this blog before and they are a fascinating tool to play around with.
You also have Tom Awad, a writer over at
Hockey Prospectus who developed GVT, or Goals Versus Threshold. This attempts to demonstrate how useful a player is by combining a number of traditional stats ala shots against, goals for, goals against, etc. etc.
It is explained in much better detail by Awad himself right here
Each of these have limitations in one form or another, but both are pushing for the same thing. A simple, straight forward way of analyzing the overall impact of a player on his team in terms of wins or goals. It is exceptionally hard to do this with hockey, considering not everyone plays the same position or has the same role. Baseball players all get to hit, and hit in the same way. Hockey is less fortunate, and ergo it is more difficult to evaluate properly.
A third metric has popped up on,
War-On-Ice.com another site that our blog here loves to use. The good folks over there which includes Stephen Burtch, whose dCorsi Impact numbers have been talked about, have set up a Career WAR stat for NHL players. Not only that, they have set up a GAR (goals above replacement) by season.
After playing around with it,
and you can too by going right here, I've noticed it seems to be fairly accurate in evaluating players. I, of course, applied it to the Kings team. What the outcome was made complete and total sense. Take a look at the Kings roster and lineup based on WAR and GAR.
The numbers next to these players is their Wins Above Replacement. Anze Kopitar, for example, is 28 wins above a replacement level player, being replaced. His contributions to the LA Kings are worth 28 wins. Basically if the Kings did not have him, well, it would be ugly. This is the entirety of the database dating back to 2005 mind you.
Marian Gaborik (18.61) - Anze Kopitar (28.95) - Justin Williams (15.27)
Dwight King (3.26) - Jeff Carter (17.58) - Tyler Toffoli (4.26)
Dustin Brown (16.54) - Jarret Stoll (1.95) - Trevor Lewis (-0.02)
Kyle Clifford (-0.3) - Nick Shore (0.08) - Jordan Nolan (0.05)
Andreoff (-0.23)
Pearson (1.68)
Mike Richards (14.31)
Just on the forwards alone you can see that the players playing in elevated minutes generally are the players who have been more impactful over their careers. When things look a little fuzzy, like Trevor Lewis being at a -0.02 or Tyler Toffoli only being a 4.26, it is time to bust into the goals above replacement year by year.
Trevor Lewis was indeed a 6.45 GAR this year. Tyler Toffoli was a whopping 17.10 GAR. On the flip side, Mike Richards was a -0.72, and Dustin Brown, who had an off season, was a meager 3.67.
These are interesting metrics to play around with and they seem to give a good sense of the reality of it. If the Kings had removed Mike Richards from the lineup they were a better team. If they played Tyler Toffoli more than Dustin Brown they usually scored more goals. That information is somewhat reflected in these numbers. Someone like Toffoli, who has a low WAR on his career, will only see that number rise as he continues to perform at an elevated level.
Now for the defense.
Drew Doughty (9.17) - Jake Muzzin (4.89)
Alec Martinez (4.81) - Robyn Regehr (-3.41)
Brayden McNabb (0.37) - Matt Greene (1.51)
McBain (1.56)
Sekera (4.49)
Again, it makes sense. Developing defenseman like McNabb are rated a bit lower. Fringe players ala Nolan, Andreoff, Clifford and Jamie McBain are rated slightly above replacement if not right at it. Third pairing defenseman are slightly above, bottom line are for the most part slightly above replacement. The Kings top pairing has the best WAR, the top line has the best WAR. So on and so forth.
Most of this, coinciding with possession stats like corsi and fenwick and the beloved eye test, seem to confirm what most people see. For the more year to year stuff, the GAR is a good indicator for isolated evaluation. Toffoli had an amazing year, and that is indicated in his GAR. His career is still young, and he has only meant a few wins here or there for the Kings. Again, that will change as he progresses.
As always, advanced stats are to be taken with lots of other stats to support them. Just simply quoting WAR or GAR in an effort to evaluate a player is probably not the wisest thing (Although if Lombardi had referenced Regehr's WAR he may not have extended him...). Be that as it may, we are getting close. Between GVT, WOWY, and WAR/GAR, we are seeing progress made in a "Holy grail" style evaluative stat.
Play around with the WAR and GAR stuff. Some of it is pretty surprising and may or may not confirm your long standing feelings about a player or two.
Follow me on twitter for news and notes about the Kings and the NHL
Also be sure to like HockeyBuzz on facebook!
++++I AM CURRENTLY LOOKING FOR ADVERTISERS! If you, or anyone you know would be interested in placing an ad on the blog here at HockeyBuzz then send me a PM!+++++