This is Part 10 in what I hope will be just a ten part series of answers to frequently asked questions about hockey analysis. Here are Part 1, Part 2, Part 3, Part 4, Part 5, Part 6, Part 7, Part 8 and Part 9.
The "Holy Grail" in any kind of sports analysis is to figure out how many wins a player is worth. One guy, one number - and relate it to his salary to figure out if his team is under- or over-paying for him. Baseball analysts have probably done the best job at this, particularly with hitting and fielding - if you look at Tom Tango and MGL's blog at "The Book," MLB players are evaluated using Wins Above Replacement, and you can pretty easily figure out if a player is worth his salary. Hockey (and most other sports) are not quite there yet.
What we do have in hockey is a sense of what players are worth in a general sense. Quite a while ago, I looked at what happened to team goals for and goals against when a player was in the lineup vs when he was out, an idea known as With or Without You (WOWY). And recently, I estimated the value of goaltender performance. Tom Awad has built a much more rigorous system, which he calls Goals Versus Threshold, that he describes in great detail here, here and here. The average value of the top 60 forwards, top 60 defensemen and top 30 goalies in our analyses are as follows:
|F||1.4 WAR/82||2.5 WAR/82|
|D||2.0 WAR/82||1.9 WAR/82|
|G||3.2 WAR/82||2.9 WAR/82|
In general, it's close, though I differ significantly from Tom on forwards because I just looked at forwards who had the most ice time - which may or may not be the same guys who are actually the best players. On defense, the guys who play the most are probably also the best D-men, and with 30 goalies, it's hard to set replacement level in the wrong place.
Do we know anything about individual players? Saku Koivu was a perfect candidate for the WOWY analysis, and he was worth about 2.4 WAR per season over ten years, which is a very high number. At his peak, he was worth something like five wins per season. Tom came up with valuations for every player in the NHL that make a lot of sense, though they underrate defensemen like Kurt Sauer or Karlis Skrastins, who take all the tough shifts for very bad teams. But that's the state of player valuation in the NHL - we don't yet know enough about how individual players impact team performance to put a very accurate number on performance. In a way, we're where baseball was 15 years ago - they could evaluate offense, didn't quite have pitching figured out, and didn't really have the data they needed to evaluate fielding. Technology and improved record-keeping will allow us to close the gap in a few years less time, but it will not happen immediately.