Jake Gardiner and the “Mikhail Grabovski Expectation”

Something different.

The chart after the jump will be something called “The Mikhail Grabovski expectation”. It comes from a few days of sleuthing around every single Jake Gardiner statistic I could find, looking for some kind of lead, or some kind of reasonable explanation for why a plus-Corsi defenceman on a minus-possession team like the Leafs could be potentially expendable in a trade for Roberto Luongo.

Not that I dislike Gardiner, he’s just a fascinating case, and I want to learn more about him.

So what’s the “Mikhail Grabovski expectation”? If you’re a reader of the blog, you’ll know exactly how much I value Grabovski. He’s the most under appreciated athlete in Toronto sports right now. It’s weird to me that he doesn’t get his due. He’s a 20-goal scorer who is an absolute beast in the defensive zone, or, would be a beast if he ever spent any time in the defensive zone.  

In the last three seasons, Grabovski has been first on the Leafs in a stat called “Corsi”, a shot-differential metric that counts all shots, including blocks and misses. Despite the Leafs being horrifically out-shot last year in all instances, with Grabovski on the ice, the Leafs actually out-shot the opposition 30 to 29 at 5-on-5:

(Leafs Corsi numbers counted here)

Why count Corsi? It’s best explained by this old post over at an old blog called Irreverent Oiler Fans:

That post is accompanied by a table that shows which teams in 2002 were the worst in zone time, back when the statistic was tracked by the NHL. It shows two things: That teams with favourable zone time make the playoffs, and that teams that take more shots than their opponents have favourable zone time. Players like Mikhail Grabovski, who can do it so well, are a rare breed. 

Back to Gardiner: I was looking at his Hockey Analysis page from David Johnson, which shows how players did in possession when on the ice or not on the ice with most teammates and certain opponents. It also shows ice time, and I noticed that Gardiner spent 316 minutes on the ice with Grabovski, 755 without. He did much better on the ice with Grabovski (59.2% of puck possession was estimated to be in the opponents’ end) than without (just 51.2% of shots were in the opponents’ end).

I got around to thinking… is a player’s Corsi rate determined by the amount of ice-time any Toronto Maple Leaf spends with Mikhail Grabovski? It’s a simple question that’s simple enough to be answered, by way of this handy graph that confirms my hypothesis. The x-axis is the percentage of 5-on-5 time any player spent with Grabovski, while the y-axis represents the players’ overall Corsi rate, or possession percentage:

(For you math types, the r-squared is 0.40 among skaters who played at least 500 minutes.)

The line in the middle is the Grabovski Expectation. Blue diamonds, representing Leafs, above the line did better than expected, if a players’ possession rate is simply a function of how much time they spent on the ice with Grabovski. If you count, you can see eight blue diamonds above the line, and, if you’re a reader of mine, you could probably guess which Leafs made the cut: Clarke MacArthur, Carl Gunnarsson, Dion Phaneuf, David Steckel, Phil Kessel, Joffrey Lupul, Jake Gardiner, and one other. 

Here’s the full table:

Of course, there are caveats with this. Grabovski’s a good player to use for the expectation because he’s a) such a dominant possession player and b) because he plays fairly moderate minutes in terms of difficulty. Gardiner, of course, plays easier minutes, starting more shifts in the offensive zone relative to the defensive zone than any Leaf, but you’d expect that out of an offensive defenceman.

Now, who is the other defenceman over the Mikhail Grabovski expectation? Eagle-eyed readers may have noticed it. Another offensive defenceman who plays fairly easy minutes, although not as young, Cody Franson.

If the Leafs don’t get Franson re-upped when the season begins, and wind up trading Gardiner, I think that those are the types of decisions that have got the Leafs into a lot of trouble in recent years. With so many young defencemen, Stuart Percy, Morgan Rielly, Matt Finn and Jesse Blacker, coming up through the system, you really hope that Brian Burke doesn’t find both of his possession-influencing young defenders.

I’m still skeptical on Gardiner’s future ability (as I am of Franson’s, no doubt) but I’m re-assured seeing that his play wasn’t entirely determinant on playing alongside Mikhail Grabovski as initially suspected.