Seven stats: Non-traditional statistics for evaluating player performance

Goals, assists, time-on-ice, even plus-minus. There are a lot of stats hockey fans pay attention to on a daily basis. These stats give us some idea of what kind of contributions our boys are making to the team's fortunes. We decide whether they are under- or over-performing based largely on these factors.

But, let's face it. They don't tell us everything. Not even close. It's a team effort to score goals, and assists and plus-minus only partially help us understand what goes into creating a win or a loss.

Luckily (?!) there are a number of other stats available, if you look a little harder. These can clarify some questions about who is doing what on the ice. I thought that this week I'd take a look at a few of the less quoted stats to get a somewhat fuller picture of what kinds of contributions different players are making. [Note all of these are players with 10 or more games played and are current through Saturday night (Feb. 11).]

  1. CorsiRel: stands for Relative Corsi.

    Quick tutorial: A player's Corsi is all the shots directed towards the net (whether on goal, missed, or blocked) while he's on the ice minus all the shots opponents direct at the player's net while he's on the ice. BUT, raw Corsi doesn't take into account the fact that some teams/lines are just better than others. So if you want to know how a player compares to the rest of his team, you have to adjust. A player's CorsiRel subtracts his team's Corsi while he's not playing from his Corsi (while he is playing) to understand whether he's doing better or worse than his teammates. (It's CorsiOn-CorsiOff.)

    Top 5

    Bottom 5

    Marc-Andre Bergeron

    14.3

    Nate Thompson

    -14.4

    Bruno Gervais

    11.6

    Adam Hall

    -11.4

    Teddy Purcell

    10.6

    Brett Clark

    -10.5

    Steve Downie

    10.2

    Dana Tyrell

    -9.5

    Brendan Mikkelson

    8.5

    Eric Brewer

    -9.3


    Defensive players are expected to have lower Corsi ratings than offensive players. MAB isn't 28 points more talented or effective than Nate Thompson. He's just in a different role, facing different opponents in different situations. You have to take CorsiRel in context with the following 2 stats to be able to make sense out of it.

  2. CorsiRelQoC: stands for Relative Corsi (Quality of Competition)
    This is the CorsiRel rating for all the opponents a player faces. Higher QoC means you face guys who generate more shots against you. Because this takes a lot more players into account than CorsiRel, the spread is going to be a lot smaller. For the most part, CorsiRelQoC tells who a player's coach matches him up against. It's common to see offensive players with low or negative QoC numbers and defensive players with higher QoC numbers.

    Top 5

    Bottom 5

    Nate Thompson

    1.27

    Brendan Mikkelson

    -0.53

    Eric Brewer

    1.13

    Bruno Gervais

    -0.41

    Tom Pyatt

    0.99

    Marc-Andre Bergeron

    0.05

    Dominic Moore

    0.93

    Martin St. Louis

    0.12

    Victor Hedman

    0.9

    Vincent Lecavalier

    0.24


    I expected Steven Stamkos to be in the bottom 5 there. His CorsiRelQoC rating is 0.32, which is good for the 18th highest on the team. Mostly, I assume, because opposing coaches put their best defensive guys against him. But the fact that VL4 and MSL have such low ratings suggests that Stamkos is taking on some pretty tough assignments at home, as well.

  3. Offensive Zone Starts:
    This tells what percentage of a player's shifts start in the offensive zone. It's based on faceoffs rather than on-the-fly line changes, which may be too unwieldy to track. It essentially tells what situations a guy plays in. It's easier to score when you start in the offensive zone; if you start in the defensive zone, your job is to prevent goals and get the puck out. Coaches approach these things differently. Every coach uses particular players in particular situations, but some are more stringent about it than others.

    Top 5

    Bottom 5

    Marc-Andre Bergeron

    71.4

    Eric Brewer

    31.5

    Brendan Mikkelson

    69.6

    Tom Pyatt

    34

    Bruno Gervais

    63.5

    Adam Hall

    34.2

    Vincent Lecavalier

    56.1

    Dana Tyrell

    35.8

    Ryan Malone

    54.9

    Nate Thompson

    36.3


    Players who are high on this list will tend to be lower on the QoC list and higher on the CorsiRel list because they are being used in ways that are designed to get them into scoring position. Players who are low on this list will tend to be higher on the QoC list and lower on the CorsiRel list because they are being used defensively.

  4. Penalties Drawn Differential: Penalties drawn/60 - Penalties taken/60
    This stat gives an idea of which players tend to put the team on the power play (higher numbers) and which ones tend to put the team on the penalty kill (lower numbers). Of course, this season both the PP and the PK are pretty iffy, so either one is a concern.

    Top 5

    Bottom 5

    Martin St.Louis

    0.7

    Victor Hedman

    -0.7

    Ryan Shannon

    0.6

    Pavel Kubina

    -0.6

    Dana Tyrell

    0.6

    Dominic Moore

    -0.6

    Edward Purcell

    0.4

    Eric Brewer

    -0.5

    Nate Thompson

    0.4

    Brendan Mikkelson

    -0.4



  5. Hits:
    Hits are not just any collision on the ice. A hit is registered when the collision forces a change in possession. [Note: yes, the counting of hits varies greatly from rink to rink. However, since all of these guys are playing in the same arenas for roughly the same amount of time, teammates' Hit numbers are subject to far less distortion than hit numbers across teams.] I have adjusted these for ice time, so they are Hits/60.

    Top 5

    Bottom 5

    Eric Brewer

    6.86

    Martin St Louis

    0.69

    Nate Thompson

    6.74

    Brendan Mikkelson

    0.83

    Ryan Malone

    6.25

    Ryan Shannon

    1.08

    Adam Hall

    6.05

    Edward Purcell

    1.31

    Brett Clark

    5.72

    J.T. Wyman

    1.65


    No surprises here. Big guys hit more. Little guys skate more. Not sure what's up with Mikkelson, though. He is most certainly being used in offensive situations and is being called upon to be fast rather than physical.

  6. Giveaways:
    Self-explanatory. Same note about rink-to-rink distortion applies. Also adjusted for ice time (GvA/60)

    Top 5

    Bottom 5

    Vincent Lecavalier

    1.97

    J. T. Wyman

    0

    Martin St Louis

    1.84

    Dana Tyrell

    0

    Steve Downie

    1.81

    Brendan Mikkelson

    0.17

    Edward Purcell

    1.73

    Ryan Shannon

    0.45

    Brett Clark

    1.72

    Adam Hall

    0.66


    Yep. Dana Tyrell will be missed.

  7. Takeaways:
    If you want to be on the bottom of the above list, you want to be on the top of this one. TkA/60.

Top 5

Bottom 5

Steve Downie

2.56

J. T. Wyman

0.14

Martin St Louis

2.07

Brendan Mikkelson

0.17

Nate Thompson

1.95

Brett Connolly

0.41

Edward Purcell

1.89

Matt Gilroy

0.72

Ryan Malone

1.62

Eric Brewer

1.03

In each of these categories, players are making contributions (some good, some bad, some neither) that ought to affect how we look at their overall performance. It's easy, for instance, to say that Steve Downie takes too many penalties, but he draws enough to offset that somewhat, while Victor Hedman doesn't--or at least not to the same degree. At the same time, Downie leads the team in Takeaways per 60 minutes of ice time. Ryan Shannon may not be scoring a lot but he is very good at maintaining possession of the puck. Bugsy hits a lot, but he's being outmuscled by Nate Thompson right now.

So, what stands out to you?