Posted by Neil Paine on November 3, 2010
A somewhat non-basketball-related dispatch from Tango and The Book Blog: Brian MacDonald has written a paper applying Dan Rosenbaum et al's adjusted plus/minus theory to hockey players.
To which I say... Finally! I've been waiting for someone to use that methodology on the sport that popularized the plus/minus stat in the first place. Even though the standard errors are huge (especially for goalies, whose impacts are difficult to disentangle from teammates because they rarely leave the ice), it's still encouraging to see the effort be made.
For now, I'm still partial to the great Tom Awad's Goals Versus Threhold (GVT) as my go-to NHL metric (download the all-time spreadsheet here), but MacDonald's work could have implications for hockey statheads for years down the road. Nice work!