Zone Defence, Corsi and Mark Fayne

This last year I’ve been tracking zone transitions for the Oilers defencemen. I finished 31 games and will be planning on getting all 82 done next year (with help). One of the things I wanted to do this summer is analyze the data to see how these stats impact game-flow. Through this, we might be able to get a sense of what tactics and skillsets are particularly important for Oiler defencemen. We start this series with Zone Defence: preventing Controlled Zone Entries.

We know from Eric Tulsky’s work (and many, many people since him) that entering the zone with control generally leads to more shots than dumping and chasing it. There are a small number of exceptions, but generally this is the case.

By extension, then, if a defenceman prevents Controlled Zone Entries (and forces more dump-ins), they should have fewer shot attempts against. Since I have 30 games of Zone Defence data, we can take a stab at testing this theory.

The Oilers

I’ve taken every defenceman I’ve tracked and plotted his Controlled Entries Against vs his Corsi. Remember, Controlled Entries Against were only recorded when a defenceman was specifically targeted on an entry and he allowed the other team to get the puck in with control.

Well look at that! It’s a reasonably tight correlation. Defencemen who prevented Controlled Zone Entries had better Corsi. By the way, this correlation is pretty much equally strong whether it’s Corsi-for or Corsi-against.

Of course, it’s also possible that this is just correlation, not causation. Maybe the Oilers defencemen who are better at Zone Defense also have other skills that drive possession.

We can also look at this data at a team level. That is, in games where the Oilers defencemen (as a whole) allowed fewer Controlled Zone Entries, did the team have fewer shot attempts against?

Turns out… yes! There’s more noise in this graph… obviously a lot of factors can impact an individual game’s Corsi. However, there still remains a correlation between allowing fewer Controlled Zone Entries and fewer shot attempts against. In this case, the correlation was not all that strong with Corsi-for. This means that in games where the Oilers prevented Controlled Zone Entries they allowed fewer shot attempts against their own net but did not necessarily have more shots on the other guys.

So how does Mark Fayne fit into all this?

There’s an immense divide in opinion on Mark Fayne. Some feel he’s a legitimate shut-down defenceman. As evidence they point out he played that role not that long ago with Andy Greene in New Jersey. Statistically, he was near 50% Corsi and played tough competition, which is pretty solid for the Oilers.

On the other side of the coin, there’s people who believe he’s sub-replacement level and doesn’t belong in the NHL. They counter with the fact that he passed through waivers without a single team picking him up. Backing up their argument statistically is Fayne’s 47.8% xGF%, which is behind every defenceman on the Oilers except for Darnell Nurse, Andrew Ference and Brad Hunt. (xGF is an ‘expected goal’ calculation done by Manny Elk and is available from Corsica.hockey. It takes into account shot distance, type, rebounds and a multitude of other factors to weight shot attempts by how dangerous they are.)

Explain Yourself!

What do those stats tell you when you put them together?

When Mark Fayne is on the ice, the Oilers are pretty close to evening out in possession despite playing reasonably tough competition. However, when you account for shot quality, he’s struggling relative to the other Oilers D.

How can this make sense? What is Fayne doing that is helping his Corsi?

Let’s go back to where I started this post… with Zone Defence. This table is sorted by Controlled Zone Entries Allowed with the best being at the top.

There’s Mark Fayne, right in the middle of the pack in preventing Controlled Zone Entries. Okay, that’s not bad, though not world-beating. So what’s my point?

Consider that Mark Fayne spent almost all of his 5×5 TOI last season with either Andrej Sekera (49.6 CF% in 656 min) or Oscar Klefbom (50 CF% in 172 min). Either of those pairings, together, would be pretty good as a pair at preventing Controlled Zone Entries… which should lead to fewer shot attempts against and a better Corsi. The only other partner Fayne played more than 50 minutes with was Darnell Nurse (44% Corsi in 88 min). That combination would be quite poor in preventing Controlled Zone Entries… and it reflects in their Corsi.

Now, I’m not trying to argue that Zone Defense is the only reason for Fayne’s Corsi. However, there is some evidence it’s a contributing factor and it makes sense when you consider who he played with.

That said, despite what he gains (along with his partners) in shot attempts, he’s giving it back elsewhere on the ice. The Expected Goal numbers are concerning, especially since it’s now two years in a row that he’s been near the bottom for the Oilers (last year he was ahead of only Brad Hunt and David Musil).

What Does This All Mean?

Firstly, in general it means that defencemen who prevent Controlled Zone Entries also tend to have better Corsi and if the Oilers improve that aspect of play as a team, they should allow fewer shot attempts against… which is obviously hugely important.

Secondly, Mark Fayne, specifically, isn’t bad at preventing Controlled Zone Entries and he was pretty consistent with it throughout the year. That, and the fact that he almost solely played with other defencemen who were good at that same skill, probably contributed to his break-even shot attempt differential. This is important. That’s a real NHL defenceman skill. He gives it back in some other area, though. One hopes that his Expected Goals improve next year when he’s taking on softer competition. (The recent work by Woodguy and Oilers Nerd Alert on the WoodMoney metric hints this might actually happen.)

What’s Next?

In the next post in my series, I’ll take a closer look at Zone Exits how that relates to game flow.