I just can’t Fayne surprise

My fellow fancystaters are frustrating me with respect to Mark Fayne. There are a lot of strong opinions out there, and they go something like this:

“Mark Fayne’s Corsi or CorsiRel is good, therefore he must be good. And anyone who disagrees is an idiot.”

Presumably lumped into that ‘idiot’ category are not just the usual crew of “seen ‘im bad” types, but also two pretty experienced hockey guys – Chia and TMc – who came to the same conclusion.

Now, one thing you have to keep clearly in mind with the fancystats is that there is *no* perfect stat. Context is everything.  A handful of stats that paint a complete picture are pretty much always better than a single stat.

But that’s not happening here.

Yes, Fayne’s CF/Rel is good … but by eye he has not been anywhere near as good.  Shouldn’t the first response be “why are they different?” … rather than “your eyes are wrong and you’re an idiot”?

I track a lot of numbers on a lot of players, including my own adjusted version of Fenwick, which I weight based on distance and shot type. By this Dangerous Fenwick measure, Fayne consistently looks worse than he does by raw metrics.

That’s a tell that you cannot ignore.

For example, here are the ‘game flow’ charts of Fenwick and Dangerous Fenwick for Mark Fayne from his last two games:

Notice how the purple “Dangerous” line is well below the raw Fenwick line?  That means that while the raw numbers were modestly negative for Fayne, when you account for the danger involved in the shots, he got shelled both times.  Especially Dallas. Wow.

These are not atypical results for Fayne.  This pattern – where his danger adjusted numbers are bad and often the worst of all the defenders – happens with regularity.

This is the answer in my view. Fayne doesn’t give up as many shots as others (hence good Corsi and CorsiRel) … but when he does lose coverage, it’s often deadly.

A few folks have looked at Fayne’s poor on ice save % and asserted that, in light of Fayne’s shot metrics, it must just be bad luck.

I’m here to tell you, I don’t think it’s bad luck.

Since I don’t yet track my stats cumulatively, I will give you a war-on-ice chart that will make that situation very clear:

The chart is showing clearly that WITH Fayne on the ice, the Oilers give up almost 20% more shots from close in, and 35% (!) more shots than the league average from the high slot and circles. But 3% fewer shots from the blueline, meaning the other team is NOT keeping things to the perimeter.

WITHOUT Fayne, the Oilers actually look like a decent defensive team!

From the dangerous areas, the Oilers give up sightly fewer shots than the league average, while giving up more from the blueline. This is a good thing.  The CorsiRel tells us that there are more shots on the right hand side.  But the shot rate hextally chart, like the Dangerous Fenwicks, are telling us that those shots aren’t as dangerous.

@DTMAboutHeart’s xG (eXpected Goals metric) shows much the same thing:

I don’t know if this was true last year but it is definitely the case this year – Fayne has struggled.  Pull up Eric Gryba’s hextally chart and you’ll see – even he’s been markedly better at reducing dangerous shots.

This could also explain why even the “seen ‘im” opinions on Fayne range from good to bad. Those who are paying attention to the volume of shots probably like Fayne. Those paying attention to breakdowns probably do not.

Todd McLellan and Peter Chiarelli are not morons.  There is a reason Fayne got waived, and it is NOT incompatible with the fancies. But you have to dig deeper than just one number.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s