Defensive Zone Faceoffs: How Much Do They Matter?
by Jonathan Willis on Aug 9, 2009 5:30 AM PDT
in Analysis
15 comments
Gabriel Desjardins, Puck Prospectus:
It’s hard to downplay what happens here. After you lose a faceoff in the neutral zone, you have time to set up defensively and you don’t give up a particularly large number of good scoring opportunities. However, when you lose a faceoff in your own end, opponent shots on goal go up so quickly that it’s as though you gave the other team a 10-15 second power-play. For several seconds, the rate of shots allowed is as high as it is on a 5-on-3. The prospect of this level of defensive disadvantage, particularly late in a one-goal game, must give coaches nightmares.
David Staples, Cult of Hockey:
If you think about it, when a defensive zone faceoff is lost at even strength, it's just one small lost battle for the puck, one of many such small lost battles in a game, but in this case the defence is uniquely organized. Each player has a clear responsibility so it's not such a difficult thing to recover from this particular kind of lost battle.
I probably don't even need to say it, but I think Desjardins' view of what happens here is more accurate, and is even supported by David Staples' data. Let's quote Staples again:
Last season, the Oilers let in 156 goals at even strength. I reviewed each of those goals against to determine which Oilers were at fault on each goal. In the end, I counted up 22 times where a faceoff loss directly contributed to a goal against, with 17 of those faceoff losses coming in the defensive zone, five in the offensive or neutral zones.
So out of 156 goals scored against the Oilers, 17 of them came because the Oilers lost a defensive zone faceoff, 11 per cent of the goals against.
Only 11% of the Oilers goals against at 5-on-5 came about because of a lost defensive zone faceoff. That really doesn't sound like much, until context is considered. The Oilers as a team took 763 defensive zone draws at even-strength last year, and we know they averaged 47.9% in the circle. That means that the Oilers lost approximately 398 faceoffs in their own zone at even-strength.
Further, from Desjardins' data we know that the effects of losing a draw are negligible after about 25 to 30 seconds. Let's be conservative and multipy that 398 by 30 seconds - we get 199 minutes of affected ice-time. If we convert Staples' 17 GA in 199 minutes into a ratio, we get 5.14 GAON/60 - a goals against number 150% worse than the worst player on the team last year.
Desjardins states that the shots-against rate jumps to the level of a 5-on-3 powerplay, spiking nearly to the 120 shots against per 60 level roughly five seconds after a lost draw.
That's just the even-strength effects. The effects of a lost defensive zone draw on the penalty-kill are easy to imagine as well. This also doesn't allow for the flip effect - bringing in a competent third-liner would free up Shawn Horcoff to take draws in the offensive zone - and then Edmonton could receive the benefits of winning offensive zone draws, leading to more shots for, leading to more goals.
This seems like a decision that's fairly easy to make.
1 recs |
15 comments
|
Email
Print
Comments
MacT didn’t try all of the options he had.
Also. You are not asking the complete question. Do you want a short term fix, or a long term fix? A short term fix for tends to limit the long term potential upside of your team, because you are likely not getting a value contract and you are impacting player development. So the answer also depends critically on the price and duration the short term fix is coming at.
Role players and value contracts in a cap world should come from player development if one want to maximize the potential of your team.
MacT screwed up by not finding out whether Pouliot could do the job or not. MacT obviously decided for himself, but he obviously didn’t convince management, since the GM fired the coach and then hired the coach who developed Blair Betts.
Face Off %: Does it correlate to winning?
Bill James once said that the way to judge the worth of a stat is if it correlates to winning.
Lets look at the top 10 teams in FO% from last year vs the bottom 10 teams (according to NHL.com)
Top 10 Bottom 10
Pts per game – 1.2 – 1.04
5×5 outscoring – 1.05 – .94
PP % – 20.7 – 17.9
PK% – 81.4 – 80.8
It is pretty evident from last season at least that FO% correlates strongly with winning, outscoring and special teams superiority.
There are a few outliers in the top 10 such as Nashville, Ottawa and the LA Kings that are strong on the dot but finished out of the playoffs. What do those teams have in common? Face off specialists. Radek Bonk, Antoine Vermette and Jarret Stoll are all top 10 faceoff performers that skew the %’s in favour of their teams.
While its clear the winning faceoffs correlates with winning games, having a “specialist” that doesn’t contribute much in any other area (Bonk) is not a solution.
Side Note: Why the Oil don’t have Malhotra on the roster already is beyond me. Jesus, this guy is the answer to a lot of their questions.
One year’s worth of analysis is rather indeterminate.
If one uses statistics, one has to demonstrate that the data is statisticallly significant (i.e. show the error bar) before it is worth anything.
There are two parts to using statistics to make a point.
1) identify the deviation.
2) Demonstrate that the deviation isn’t random.
Doing (1) doesn’t prove anything without demonstrating (2).
Agreed on sample size
Too lazy on a Sunday morning to run excel spreadsheets for the past few years. I will take a crack at it on company time tomorrow maybe.
Actually, statistical significance requires calculating a p value and having it be below 0.05. Because I don’t know how to do that for myself, I rely on a stats program I use at work.
SNN Sports - A theoretical Oilers blog (i.e. theoretically, I write stuff there)
I’ll also take this opportunity to remind everyone that correlations only tell you so much, and even a statistically significant correlation doesn’t prove that Stat X is a contributor to winning.
SNN Sports - A theoretical Oilers blog (i.e. theoretically, I write stuff there)
OK, you win.
I am just a guy with a question and a spread sheet. You are much more statistically edumacated than I am. So, please to be answering the question. Does winning faceoffs cause you to win games? and if so, by how much?
In isolation, the question is rather pointless?
There may be factors more important than winning faceoffs, like a positive team Corsi, like PK effectiveness, like accumulated time of puck possession.
If you want to prove winning faceoffs is important to winning, you not only have to show that it itself is statistically significant, but that it is more statistically significant than other variables.
It makes no sense attempting to improve faceoffs, if in doing so you impact something relatively more important.
The moneypuck guys have a lot of hypotheses, and have accumulated some data, but they haven’t really conclusively shown the statistical significance, and the relative statistical significance of anything.
So all one really has are hypotheses, which are great, but which also could be wrong, or counterproductive, if one doesn’t do complete analyses.
I would think most moneypuck guys hypothosize that Corsi and EVP 5×5 are probably the most important..
So faceoffs would not only have to be shown to be important to winning, but more important than other things like Corsi, or one could potentially be making some very bad decisions.
…I wasn’t trying to win anything, but OK, go me. And anyway, I’m hardly more statistically edumacated than most — hell, I got a C in stats — but I know enough to know that you can’t really conclusively show much with the data at hand. You can show a relationship, but it’s confounded by a multitude of factors, some measurable some not. A relationship can suggest that a factor might be relevant, but it says nothing about causality, or its relative importance in the grand scheme of things. I have the data on hand, from back when I did the hits vs. point-percentage analysis a couple of weeks back, but didn’t get a chance to run it today (busy preparing a paper for work, on a three-week deadline). There might be a slight correlation, but just based on my sense of the game — which is the best thing any of us have — I would hypothesize that it’s a slight factor, particularly in terms of contribution to special teams, but that it’s not something I would prioritize over scoring ability or good hockey sense, which saw-him-good and moneypuck alike agree are the most important things in a skater. See also Brian’s comment, in that regard: there’s a lot of small-sample numbers that come out of this.
SNN Sports - A theoretical Oilers blog (i.e. theoretically, I write stuff there)
Shots must be elevated, yes, but aren’t the majority of them coming from defensemen?
In any case, it’s one thing to show that faceoffs are important, and another to show that a faceoff specialist would add much value. After all, the difference between a very good faceoff player and an average one is 5 or 6 wins out of a hundred draws. Even at the increased rate of shots against, we’re talking about, what, 2 or 3 shots? Meaning something like a dozen or so fewer shots over a season? I suppose it’s not irrelevent, but it’s pretty far down the list of important attributes.
I’m going to have to think about your comment, but I think you very well may be right. Thanks for contributing.
A posse ad esse.
Writer for The Copper & Blue and OilersNation.
by Jonathan Willis on Aug 10, 2009 10:20 AM PDT up reply actions
I think what he’s trying to say is that it’s not impossible to defend a defensive zone faceoff, because it would be ridiculously silly to suggest that losing a defensive zone faceoff does not deal a team a significant disadvantage. It’s like Corsi hell.
RT40 writes with An Oilers Refinery and is an avid hockey fan.
Gabe's article
Excellent work by Gabe Desjardins, as usual. One hint for Gabe if he’s reading this, I’d like to know though how much goal scoring goes up after a defensive zone faceoff loss, as opposed to just shots on net.
I know Tyler looked at the 2006-07 Oilers in this regard, over a 45 second time span, but that time span struck me as awfully long, and the 2006-07 team was uniquely bad at moving the puck (what with Jason Smith ringing it off the boards all the time and MAB’s giveaways).
Of course, I’m not saying there’s no problem with losing defensive zone faceoffs.
There is some issue, but I’m trying to figure out, like the rest of you, how big a problem that might be.
My intuition tells me it’s not such a massive issue, as players are in position and organized to stop a nasty shot on net. If you go by goals that come bang-bang off a faceoff loss, the Shea Weber scenario where a top shooter just blasts it in the net, I do believe it happened about five times last year . . . We lost far more goals with bad line changes. . . .
Now, a better faceoff man might have one a few of those faceoffs that led to goals against, and that would cut down on a goal here or there, but I don’t think this is where we’ll find the low-hanging fruit when it comes to cutting down on goals against at even strength, at least.
All that said, defensive zone faceoffs losses on the powerplay are a massive issue, I do believe. I’m studying all the powerplay goals against the Oilers on NHL.com to see if I can learn anything about this, but I’m hampered by the short time span of those video clips.
The Oil really do need a guy who can take faceoffs on the powerplay and help kill of the powerplay, which is why the loss of Brodziak is so unwelcome, IMO. He could take those d-zone faceoffs on the kill, and maybe he just might have improved as a penalty killer, stop being so over-aggressive and out of position out there.
by David Staples @ The Cult of Hockey on Aug 13, 2009 7:06 PM PDT reply actions
Jon's argument
I didn’t address your argument, Jonathan, about the goals against rate shooting up after a d-zone face-off loss. It has me thinking. Of course, I’m looking for holes in it ;), but it’s compelling on the face of it.
There’s certainly no doubt it’s best not to lose d-zone face-offs.
by David Staples @ The Cult of Hockey on Aug 13, 2009 7:19 PM PDT reply actions
Comments For This Post Are Closed
Secondary Sidebar
FanPosts
Community blog posts and discussion.
Recent FanPosts
Northwest Standings
Oilers Schedule
Next 5 Games
Calgary
|
Thu 10/07 | 7:00 PM PDT |
|
Florida
|
Sun 10/10 | 5:00 PM PDT |
|
@
Minnesota
|
Thu 10/14 | 5:30 PM PDT |
|
@
Calgary
|
Sat 10/16 | 7:00 PM PDT |
|
Minnesota
|
Thu 10/21 | 6:00 PM PDT |
Last 5 Games
|
@
Anaheim
|
Sun 04/11 | L 2 - 7 |
|
@
Los Angeles
|
Sat 04/10 | SO - W 4 - 3 |
|
Colorado
|
Wed 04/07 | OT - W 5 - 4 |
|
Minnesota
|
Mon 04/05 | W 4 - 1 |
|
@
Phoenix
|
Sat 04/03 | SO - L 2 - 3 |
Certain photos copyright © 2010 by Associated Press or Getty Images. Any commercial use or distribution without the express written consent of Associated Press and Getty Images is strictly prohibited.
