The Metrics We UseSeptember 16, 2011, by
In an effort to gain more of an understanding of a player’s value to his team, specifically players we can’t watch daily, we are making more of an effort to use metrics in our analyses. It’s not to say we are shunning goals, assists, powerplay points, and stats of that ilk; but we will generally try to stay away from stats that are more team based, like plus/minus. It’s simply because there are now better options to use.
All but two of the metrics listed here were created by Behind The Net. The other two metrics (WAT, PVT) are based off of metrics they have created.
Goals Versus Threshold (GVT): GVT was the first of the metrics to gain popularity. To first define the “threshold”, it simply means “replacement player”. The “replacement player” here is the top AHL/minor leaguer (note: NOT the top prospect, per se), or the best available free agent in the middle of the season. Glen Miller of SNY Rangers Blog put this into Ranger terms and noted that a replacement player is someone like Chad Kolarik, Kris Newbury, or Andre Deveaux. Simply put, GVT is the number of goals a player is worth in relation to a replacement player over the course of a season.
In the 2010-2011 season, Marian Gaborik finished with a GVT of 10.2, which means his presence on the Rangers meant a little more than 10 additional goals that season. Comparatively, Henrik Lundqvist finished with a GVT of 29.6, which means that his presence on the Rangers saved them a little less than 30 goals throughout the season.
GVT is a combination of a player’s offensive, defensive, and shootout contributions for skaters, and the goaltending and shootout contributions for the goalies. The formula is very complex, and I did not come up with it, so I cannot comment on it. But, the formula does a fair job at weighing these attributes, and adjusting them for position. For example, a defenseman’s defensive contributions will count more than his offensive contributions. Ice time and quality of ice time is also measured in this formula.
Points Versus Threshold (PVT): PVT is a metric that I created after reading Miller’s post on WAT (explained after the jump). Simply put, I felt that points were an easier way to judge a player’s worth in the standings than wins, as points earned in the shootout play a large role in hockey. Using Tom Awad’s note that 6 goals (GVT) = 1 win (WAT), and 3 goals (GVT) = 1 point (PVT), I simply took the GVT and divided by 3 to come up with PVT.
Wins Above Threshold (WAT): WAT is a metric created by Glen Miller in an attempt to turn GVT into a number of wins in the standings. The reasoning is that while goals are great, wins is a better mark of a player’s worth. Basically Miller takes GVT and divides it by 6, where 6 is the number of goals that equals a win (per discussions with metrics guru Tom Awad). This helps you realize how many extra wins a player’s presence on the team can bring.
On/Off Ice +/-: On/Off Ice +/- is a metric that is designed to show how a player matches up with his team, as opposed to the rest of the league. It may seem a little misleading if just reading the name. Let’s first start with +/- per 60 minutes, which is a rate that that calculates a player’s +/- per 60 minutes on the ice. Next, there is +/- per 60 Off Ice, which calculates a player’s +/- per 60 minutes not on the ice.
On/Off Ice +/- simply subtracts the +/- per 60 off the ice from the +/- per 60 on the ice to arrive at its value. The reason why this is important is because it takes into account ice time, which traditional +/- does not. Simply put, a player with more ice time has a better chance at finishing with a higher traditional +/-. The rate stats here account for the ice time per player, and averages them out to a stat that is rounded per 60 minutes.
Quality of Competition Faced (QUALCOMP): This is fairly simply to understand. The more time seen against opponents top players, the higher the QUALCOMP. This stat is based off of BTN’s On/Off Ice +/- (explained above). Basically QUALCOMP takes the On/Off Ice +/- of the opponents faced for one shift, and turns them into a stat to allow people to see what the level of competition faced is. Naturally someone like Marc Staal, who regularly sees top scorers, has a higher QUALCOMP than a bottom pairing defenseman.
Using traditional +/-, one could make an argument that Mike Sauer had a better year than Marc Staal. However, QUALCOMP shows that Staal faced far greater competition. With QUALCOMP, the greater the number, the harder the competition faced.
This post will be a permanent page on the blog. The link is above on the right.