Photo by “Herkie” via Wikimedia Commons
That #8 Blue Jacket up there is Jan Hejda, one of a number of defensemen coined “defensive defensemen.” I’ve been a defensive defenseman skeptic for quite awhile now, primarily because I’ve just not been convinced that they make much difference in the possession game. My rationale runs like this: when it comes to defensemen compared to forwards, the actual areas of the rink wherein one versus the other skates and possesses the puck are undeniably different. In diagram:
With the defensive zone being in the left half of the rink, the green area is the rough regular coverage area of forwards. These are the areas they do everything, including disrupt passes, make passes, shoot, score, assist. Most top 4 defensemen have a different coverage area:
They cover lower in their own zone than forwards, though both forwards and defensemen play active roles in the scoring chance areas on down to the offensive zone. It’s in the offensive zone that there’s major divergence, as defensemen cannot afford to be caught much lower than the tops of the circles (and if they do, they will return to the top of the zone soon thereafter). Theoretically, defensive defensemen won’t even go that far; their responsibility is even more devoted to containing the zone or protecting against their partners being caught deep. Their coverage, then, should look more like this:
Notice the similarities between this chart and the one above it. It’s important to point this out…in the defensive zone, the responsibilities of defensemen (offensive, balanced, defensive) aren’t much different. So, from the bat, if your idea is that there are defensemen that are specifically shutting down the opposition in their own zone, and they are “defensive defensemen,” there is limited evidence that one or the other defensemen gives up higher-percentage shots in their own zone. The actual, more important difference is in the offensive (and, to a lesser degree, neutral) zone, where their responsibilities make a greater divergence with forwards and lesser among defensive vs. offensive/balanced defensemen. How could this manifest itself in something a bit more tangible?
Well, thanks to Greg Sinclair’s awesome Super Shot Search, I can use shot locations to give us a sense of how this might demonstrate itself. Reaching back to 2008-09 through 2012-13, I created four groups of defensemen (whom had all played 35+ GP in their respective seasons):
- “Offensive defensemen” – defensemen with 3+ min PPTOI/G and less than 1 min SHTOI/G
- “Balanced defensemen” – defensemen with 2.5+ min PPTOI/G and 2.5+ min SHTOI/G
- “Defensive defensemen” – defensemen with 3+ min SHTOI/G and less than 1 min PPTOI/G
- “Replacement defensemen” – defensemen with less than 1 min PPTOI/G and less than 1 min SHTOI/G
Basically, I’m placing faith in coach and scout assessment, which is not entirely unfounded procedure if you’re looking league-wide. Each group gave me 85-100 player seasons, with a lot of micro-data to tease out the values/detriments of these defense groups (which will be expanded upon in the second and third parts). Their shot locations:
The sheer shot volume difference is fairly obvious; the offensive and balanced defense groups simply shot more. But you can definitely see the difference between defensive defensemen and their balanced and offensive counterparts.
From this charting, we can see the theoretical “coverage zones” play out in shot location. Defensive defensemen are less likely to commit to playing deeper in the zone and shooting (or passing) from closer in. Overall, this should make sense; coaches probably don’t want them jumping in too far, nor are they likely so good offensively that it would be worth the risk. Having watched Steve Burtch’s work on measuring quality defensive play over at Pension Plan Puppets, I’ve become more and more convinced that defensive defensemen are less consequential than their balanced or even offensive counterparts. I mean, if offensive D are clearly (above) jumping further into the offensive zone, aren’t they then engaging more fully a greater area of the ice and thus playing a greater stake in the possession game? Also, if defensive defensemen shot locations are so closely mirroring replacement defensemen (who aren’t trusted with prime defensive minutes), and they share similar assignments to balanced and offensive defensemen in their own zone, are they even a thing? So, that was ultimately where my hypothesis ended up, being skeptical of both the value and existence of defensive defensemen based on the above work; I carried this to Part 2 (viewable here), which looks closer at these shooting and offense contributions, and drew some conclusion in Part 3 (viewable here).