Frequently Asked Questions #4: Zone Starts (aka Offensive-to-Defensive Face-off Ratio)

by Hawerchuk on Oct 8, 2009 6:54 PM EDT in Statistics

This is Part 4 in an as-of-yet-unknown-number-of-parts series on advanced statistics in the NHL - here are Part 1, Part 2 and Part 3. So far, we've considered the impact of the quality of a player's competition on his performance, but we haven't considered the significance of territory in this analysis. We no longer have zone possession data by team, but at some point, Vic Ferrari popularized the notion of using faceoffs as a proxy for territorial control.

In other words, there are NHL players who go over the boards only for offensive zone draws or when it's clear that the other team is penned in in their end, and conversely, there are players who take all the tough defensive zone faceoffs late in the game. Coaches know who they want out on the ice in a given situation, and that determines the territorial advantage or disadvantage that a player starts out with before he racks up any statistics.

There's no question that this is important: a lost defensive zone faceoff costs a team approximately 0.25 shots on goal. For a guy like Kurt Sauer, who was out for 239 more defensive zone draws than offensive ones last year, that's incredibly costly - perhaps 3 to 5 goals versus an even split. Sauer, who was on the ice for 39 goals for at 5v5 and 43 goals against last year, could be a +1 instead of a -4. With six goals corresponding to approximately one win, that has a huge impact on a defenseman's apparent value. Indeed, the correlation between which zone a player starts in and the number of shots for and against his team while he's on the ice are very highly-correlated - Tyler Dellow again brings us the analysis of Corsi Number vs Zone Starts for defensemen.

Here are the Offensive Zone Start leaders for 2007-08 and 2008-09. You can see that Vincent Lecavalier gets to go out for a lot of offensive draws - 62% in 07-08 and 60% in 08-09 - while Jerred Smithson has to do a lot of the dirty work in Nashville (36.8% and 33.4% the last two seasons.) Again, keep in mind that these are only the roles that their respective coaches have chosen for them. It's possible that they could be used differently by other coaches and be more effective as a result.

Tweet Comment 8 comments | 0 recs |

Do you like this story?

Comments

Wait...

So the % on the last column is the % of draws that were taken either in the neutral or offensive zone, right?

Do you account for different teams getting different zon start splits? That is, Scott Gomez, for example, looked like he didn’t had that many defensive zone starts last year, but he actually was on the ice for over 30% of his team’s ES defensive zone Faceoffs…

This is a great, great series and is of tremendous help for those willing to have a bit of fun with your stat site…

by Olivier on Oct 8, 2009 7:35 PM EDT reply actions

The last column is O/(O+D). We don’t have a convention yet, so I just made one up. And I included all faceoffs, not just ones at the start of a player’s shift.

Glad you’re enjoying the series!

by Hawerchuk on Oct 8, 2009 8:12 PM EDT up reply actions

Why not include neutral zone draws, i.e. O/(O+N+D)? I wonder if the current format tends to overstate the extent to which players are shifted one way or the other.

More fun than a stick to the face!
On the Forecheck is SB Nation's blog covering the Nashville Predators.

by Dirk Hoag on Oct 9, 2009 4:49 PM EDT up reply actions

My choice was arbitrary – once everybody agrees how it should be calculated, I’ll follow suit!

by Hawerchuk on Oct 9, 2009 5:15 PM EDT up reply actions

I like the (OZ/(OZ+DZ)) metric better. No reason to think NZ faceoffs have any particular impact on shots or scoring chances. My hunch is that the % of NZ faceoffs is probably much more evenly distributed across players on the same team than the distribution of OZ and DZ faceoffs.

All things Thrashers + stats: www.birdwatchersanonymous.com

by The Falconer on Oct 13, 2009 11:31 AM EDT up reply actions

Shift End Statistics

I’ve made use at leafs.hockeyanalysis.com before of the data Vic posted for zone start, and shift end, and I wonder if that’s not a more valuable overall metric when comparing players. Obviously players who line up on offensive zone faceoffs are receiving some level of favourtism from their coach in that regard, as are players who are lined up predominantly in the defensive zone.

Indications of zone shift would thus draw out who is actually most successful at both transitioning from the defensive to the neutral and/or offensive zone, and who is giving way when starting in the offensive end. Players who end up in the opposition end more often than they start in their own would appear to be more successful from a puck possession and pressure perspective.

Just wondering what your thoughts on it are.

"Success is the ability to go from one failure to another with no loss of enthusiasm."

- Sir Winston Churchill

I'm pretty sure he's talking about the Leafs.

by Steve Burtch on Oct 26, 2009 12:19 AM EDT reply actions

Icing?

Does this take into account icing, whereby line changes aren’t allowed for the defending team?

If not, we’d have to be cautious about how we’re using the statistic. If we adjust a player’s rating to his time spent in the defensive zone, aren’t we creating excuses for players that get trapped in their own zones for faceoffs by icing calls? I don’t know that the %error would be terribly huge, but it might be something to think about before we canonize this.

"Without good hard work, it is impossible to reach the pinnacle of success." - Anatoli Tarasov

by PRC on Dec 3, 2009 1:43 PM EST reply actions