How much does the Avs PP miss JM-Liles?

Photo from The Denver Post

Ever since JM-Liles went down with a shoulder injury I think people are unclear just how much the Avs actually miss him. Since the Avs are have still been winning games without him, I'm not sure people understand the impact that the injury has actually hurt the Avs. In fact because of the improved play of Kyle Cumiskey some people have even mentioned that Liles and his injury might even be expendable at the trade deadline.

So I wanted to examine just how much impact the injury to JM-Liles has had on the Avs. Before JM-Liles injury the Avs PP was cruising along at a phenomenal 38.5% (10/22), and since JM-Liles went down with a shoulder injury it has limped along at 10.5% (2/19). That's a huge drop, and it's easy to point at that and say "It's been huge", but it's impossible to tell how much of that drop has to do with normal regression to the mean, and how much of it is due to the loss of one of the Avs top PP players.

It's been my contention that JML is the best passer on the Avs blueline, and that since power play scoring is strongly dependent on either passing or shooting from the blue line, that his loss has been the biggest reason for the sudden drop in the Avs PP efficiency.

First I looked at the Avs Power Play Points per 60 minutes of PP ice time, to try and get an indication of how each Av is scoring while on the Power Play:

7 29.27
16 9.73
5 8.23
23 8.05
4 7.17
8 6.32
40 5.84
9 5.45
26 5.25
27 4.91
10 0.00
22 0.00
24 0.00
25 0.00
37 0.00
39 0.00
44 0.00
52 0.00
54 0.00
55 0.00

Now you can throw out TJ Hensick (#7)'s numbers because he has 1 Power play point in his 2 minutes 3 seconds of PP ice time. That's too small a sample size. In fact with most players getting 4 or less points, this doesn't really do anything to determine whether Liles injury is affecting the powerplay, or whether it's normal regression to the mean. It's interesting to note that Tucker, Hejduk, Clark, and Liles have been the Avs most efficient scorers on the powerplay, however let's keep in mind that these are small sample sizes, so it's tough to read into these numbers too critically.

However there are enough scoring chances that you can at least get a preliminary idea of how important someone is to a power play, by taking a look at on ice scoring chances per 60 minutes of PP ice time. Just to properly set up the data here: Every time the Avs have had a scoring chance while on the ice The Gospel of Hockey blog marks it down, and whether it's on the PP, PK, or ES. You can see the Avs through the first 10 games here . I took it a step further and extrapolated the scoring chances for every 60 minutes a player is on the ice. (why 60? because that's what Behind the Net uses in all his stats, and so it's pretty much standardized, even if I think 20 would be a better metric, whatever it's a factor of 3).

Player # PPSC60
24 89.26
7 87.80
4 64.50
9 59.49
39 54.14
26 53.94
16 48.65
27 47.84
23 47.56
40 44.71
5 43.65
22 41.86
10 39.69
8 23.86
25 15.06
37 12.86
44 0.00
52 0.00
54 0.00
55 0.00

This last data set should give a pretty good feel for who is creating scoring chances for the Avs on the power play. First throw out Salei (24), Hensick (7), and Galiardi (39)'s data because they have been on the powerplay for too little time (just over 2:00 total over the first 10 games each). This shows that while JM Liles is on the ice for Power Plays, there's 64-65 scoring chances for every 60 minutes he plays. Or, there's more than 1 per minute of PP ice time, better than anyone else on the team. (Interestingly enough, Duchene's 2nd).

In order for the Avs to maintain success this season they are going to need to continue to have a strong power play, as they are being severely out shot at even strength. This shows me that this season it is critical to have JM Liles on the blueline for the powerplay, as the drop-off between him and his PP replacements (Clark-5, and Cumiskey-10) has been extraordinary To me, this shows that, at the very least, JM Liles is extremly valuable on the power play.

Yes the Avalanche's recent PP woes are the partially a normal regression to the mean, but there can be no doubt that they really miss Liles' contributions from the blue line with the man advantage.


  1. PP %s are super trickey. How do treat a PP at the very end of a game with the lead? What about a PP only lasts 20 seconds? Or a 15 second penalty at the end if a game?

    I went through the games after Lules went down and here is what I counted up.
    Montreal 0/3 + 23 seconds on a 4th pp.
    Detroit 0/3
    Wild 0/2
    Carolina 2/4 + 40 seconds for Staal getting kicked at the end of the game.
    Detroit 0/1 + 25 secs for Holstrom at the very end of the game.

    So, from what I can tell they went 3/13. But honestly, you can call it 3/15 because of one 5 minute major and a double minor. There was one Acs goal against Carolina only 20 seconds after a penalty expired where the PK unit couldn't change.

    As a whole I'm not saying the team is better without Liles. By just watching you can see the PP looks better with him on it. I just don't think there's a large enough sample size to make a point statistically. And those PP stats aren't always what they seem.

  2. Jibbs, I hope your not turning into a numbers nerd....

  3. Dario,

    Yeah, I was a bit weary of that. It's really hard to work that into it, which is kind of why I blamed part of it on statistical normalization. I mean those sort of things just happen.

    But for the meat of the analysis i used PP time on ice instead of PP opportunities, so that it would be a decent one.


    I've always been a numbers nerd.