buy discount cialis

A Look At The Defense’s Quality Of Competition

When the Rangers lost Marc Staal and Mike Sauer, they lost more than two of their top four defensemen. They lost two of their most physical defensemen and two guys that have been relied upon for many games to neutralize the opposition’s top lines. They are out, and others are looked at to fill those gaping holes on defense. Although it is very tough to quantify how the Rangers will look when Staal and Sauer come back, we can look at who the workhorses are, and who is lining up against top quality –or bottom of the barrel– opponents.

Let’s state the obvious: Dan Girardi and Ryan McDonagh are the workhorses, and they are facing off against the opposition’s top lines night in and night out. They will be doing so for the foreseeable future too, as there does not appear to be any timetable for Staal and Sauer’s return. Using the eye test, it’s easy to say that Michael Del Zotto has been one of the Rangers better defenders as well, and to an extent he has been. But that’s why we have stats like QUALCOMP that measure the quality of competition that each player is facing.

A brief description of QUALCOMP, it is a positive or negative number, and the higher it is, the better competition a player is facing. Let’s look at the table below through the Rangers first three games:

Player GP QUALCOMP
Ryan McDonagh

3

0.635

Dan Girardi

3

0.475

Tim Erixon

3

0.144

Steve Eminger

3

-0.104

Jeff Woywitka

1

-0.494

Michael Del Zotto

3

-1.115

To sum up the table, the players facing the top quality competition are the ones you’d expect in McDonagh and Girardi. After that, it’s interesting to see that Tim Erixon is the player that coach John Tortorella seems to rely on the most when it comes to difficult opponents. The fact that they are the only three with positive QUALCOMP shows how much trust the coach has in them, and how his trust wavers with some of his other options.

Yes, Michael Del Zotto has been pretty good on defense this year, but he is way at the bottom of the QUALCOMP rankings. In fact, Del Zotto is at the very bottom in QUALCOMP for the entire league (excluding Sauer, who is injured). This is, of course, a very small sample size (just three games), and these numbers are likely to change dramatically as there is more of a regression to the mean (or progression to the mean for those slow out of the gate).

While those numbers themselves may change dramatically, the place where each defenseman resides on the list is likely to remain unchanged. It is widely known that with Staal and Sauer out, McDonagh and Girardi will be the top dogs on the blue line. What is troubling is that the Rangers have one defenseman that they are “breaking in” (Del Zotto), and another veteran –one that they relied on heavily last year– that are in the negatives with their QUALCOMP.

We do fall victim to small sample sizes here, as Jeff Woywitka has only played one game, and Steve Eminger took about 10-15 games last year to hit his stride and become that reliable defenseman that helped the Rangers through their injuries last season.  That said, trends will likely continue though, so expect to see Del Zotto facing off against lesser quality opponents, and expect the bulk of the heavy duty work to go to McDonagh and Girardi.

The positive to take out of this is that Tim Erixon has been facing some decent competition, better than both Eminger and Woywitka. He has been up to the task thus far. Although the number may change, he will still be considered to be a “second pairing” guy when it comes to the quality he is facing.

25 Responses to “A Look At The Defense’s Quality Of Competition”

  1. Matt J says:

    I don’t think Erixon will be sent down when staal and sauer return. He’s played too good. I think eminger will see some press box time to MDZ. I haven’t seen enough of woywitka to make an assessment on him. Our problems on defense will stop when staal and sauer return although giving up 4 ES goals in 3 games isn’t so bad.

  2. RangerSmurf says:

    I generally prefer Corsi Rel QoC to the regular qualcomp, which has some goaltending effects and sample size issues (goals are less frequent than shots), especially this early.

    By that, MDZ is 3rd, followed by Woywitka, Erixon, and Eminger.

    • Dave says:

      I use QUALCOMP to just see who’s getting top opposition, not necessarily who is most effective against that opposition.

      • RangerSmurf says:

        They both measure opposition, not effectiveness

        QualComp measures based on relative +/-

        Corsi Rel. Comp measures based on Relative Corsi.

        The one based on +/- has goaltending effects in it.

        • RangerSmurf says:

          Here’s the explanation:

          http://www.arcticicehockey.com/2010/1/15/1252520/slight-statistical-improvement

          It applies to just Corsi QoC, but Corsi Rel uses the same methodology, just with Rel. Corsi as opposed to just Corsi.

        • The Suit says:

          Maybe I’m just not grasping this, but when you say MDZ is third (or last) in icetime against quality competition, how exactly is that calculated using both methods QUALCOMP & Corsi Rel QoC? Spell it out for us. And dont use the word corsi in your answer haha.

          • Dave says:

            Basically one stat (QUALCOMP) uses goaltending, as it is derived using +/-, which can be affected based on the goaltender’s ability to stop a puck.

            The other stat (Corsi Rel Comp) is based off of the Corsi stat (shots directed at net that are either on net, blocked, or wide). Corsi itself is essentially a measure of puck possession. A higher Corsi Rel Comp means a player has the puck more often against higher competition.

          • RangerSmurf says:

            No problem.

            http://www.arcticicehockey.com/2009/10/8/1076586/frequently-asked-question-2

            If you don’t feel like reading that link, here’s the idea:

            Take a players +/- while he’s on the ice, compare it to when he’s off. That gives you the relative +/- for each player in the league.

            Next, identify the 5 players a player is on ice against, and find their rel. +/-. Average that total.

            Now, repeat for every instance a player is on the ice.

            For that other QualComp mentioned that I prefer, use the other stat instead of relative +/-

  3. The Suit says:

    is the league average zero for this stuff? what’s the benchmark?

    • Dave says:

      This early in the season it’s tough to really go with the exact numbers. But you can identify trends.

      • The Suit says:

        what about last season?

        • Dave says:

          QUALCOMP numbers don’t leverage prior year, but you can tell that MDZ hasn’t gone up against the top lines that often.

          That may change if he continues to be with Erixon, who has been getting 2nd line assignments.

  4. Chris in MA says:

    I would like to know where these whacky stat sites get their base stats. To my knowledge, the NHL does not have an API we can utilize to pull raw stats.

    Yahoo has something you can work with, though its limited in stat categories… but trying to read through/follow their documentation is damn near impossible.

    Do these guys just manually grab them from NHL.com, or do they scrape the site for the stats? or what?

    • Dave says:

      behindthenet.ca

      “In 1997-98, the NHL introduced a real-time tracking system for on-ice events and statistics. Behind the Net mines these statistics to find every player’s value when he’s on the ice. Our database now includes five seasons worth on data.”

      • Chris in MA says:

        Right… but how to the ‘mine’ it. If I were to try and do stuff on my own, I’d still have to do to them or NHL.com and sift through the data and pull out what I want manually.

        Im ultimately looking into trying to pull all the data into my own database where I can run my own queries, etc.

        • Chris in MA says:

          Wow… I should have proof-read that before hitting ‘submit’.

          Basically, I have an idea in mind, for my own personal info, and am looking into how to obtain all these stats from the NHL on an automated/ongoing basis.

          • RangerSmurf says:

            My understanding is they’ve all written scripts to scrape the stats out of the sites.

            If you ever figure out how to download all the play by play data, please let me know.