I interrupt what is surely a thrilling series of September hockey blogs covering the advanced stats of the Coyotes roster to talk about something that came up earlier today in the comments section:
How much do zone starts matter?
Friend of the blog, Camfor, pointed out to me that a lot of the players that I tend to criticize for their horrid possession numbers (Nick Grossmann - with two n's - Dan Girardi, Luke Schenn, etc.) have really tough zone starts, which he suggests is the reason why their possession numbers are bad.
I was writing out a comment when it became apparent that I was writing a blog, so why bury an hour of my time on the fourth page of a blog probably like four people read, when you can instigate a discussion that might actually interest more people? (Hey it's September, can you PLEASE break down the NHL's least popular team's five through seventh defenseman?was the general consensus about yesterday's piece).
But before I begin, I'd just like to say that after work today I went to the Mall. The Mall is basically my kryptonite and besides bullets, knives, disease, nuclear waste and penicillin, one of my main weaknesses.
Of course, I went there for the only socially acceptable reason: to go to the record store. Problem is I always forget what a crap-hole HMV has become now that it only sells Mall-Pop and kitchy items with Kiss or the Beatles pasted on them. They played horrible music and had only two of the six or ten things I wanted to buy. (How does a record store not carry New Order or any of Prince's back catalog?).
But I did get new records by Wilco and Beach House, so it was somewhat worth it. I don't know if either record is any good yet, but at least they put some $ into the packaging, so that's a good start. The single off the Wilco record (if you can call it that) is one of their best songs in years, so here's hoping the rest of the record lives up to it.
Now, onto the Zone Starts Issue:
I constantly see people using zone starts as an excuse for a player with bad numbers or as a reason why another player got good numbers. Much like face-offs, hitting, blocked shots, turnovers and fighting, I think that zone-starts are another thing that gets blown far out of proportion, in terms of it's importance. Without bogging the whole thing down with numbers and graphs, here's my main reasons why:
(Keep in mind that I'm not arguing that zone starts have no bearing or importance. If you started every single shift in one zone or the other, you'd end up with either an extremely high or low possession rating. But realistically, no one starts that high of a percentage of their shifts in one zone or the other).
Using Stats.HockeyAnalysis.com, I went through and just looked at a whole bunch of random players percentage of total draws that they had in each zone. Most players come in where you expect (roughly 33% in each zone) but even the extreme ones aren't that "extreme." Take Matt Cooke from 2013-14 who took 42% of his faceoffs in the dzone. It seems like a big number, but he's only taking an extra nine faceoffs in the dzone vs. the average for every 100 faceoffs that he takes.
Since he had about 1000 face-offs in a whole season, we can figure it takes about eight games to be on the ice for 100 faceoffs, which is about one face off per game in the dzone extra. Or, more succinctly: not that big of a deal.
It matters, of course, but the point is that 42% jumps off the page, where as in context, it's not such a crazy amount that he's helpless out there.
Here are several problems with zone starts:
1. Assuming that we're going by where the faceoff is taken and calling that a zone start if your on the ice for it, a large amount of zone starts are because of goalies freezing the puck and causing faceoffs. It's not uncommon to see two or three faceoffs by the same players if the goalies keeps freezing the puck.
If you are a bad player, you'll have more shots against you, your goalie will save and freeze the puck and you'll start again in your own zone. Now it looks like a super high percentage of your faceoffs are occurring when you are in the dzone, but its not exactly true. Furthermore, if you suck at clearing the puck, this can keep happening. Especially when we're talking defensemen and a coach is liable to put his "defensive" guys out or keep them out in this situation, then more zone starts vs less zone starts becomes more about who is a better player and not an excuse to justify why stats make one player look bad.
Ergo: Girardi doesn't have bad stats because he's always playing tougher minutes, he actually forces himself to play tougher minutes by being bad.
2. If zone starts matter so much, explain this:
Over 3 years, McDonagh has played 75% of his 5v5 time with Girardi. Together they post a 49.5% CF.
In the other 25% of the time, McD jumps over 5 percentage points, which keeping in mind the range of normal Corsi is about a 25% jump in overall effectiveness.
Without McD. Girardi drops to 47% or about 15% worse of a player then when they're together.
75% of the time these guys play, it's with each other in the same zones and the same competition. There is barely a difference in the total percentage of faceoffs each took in each zone over a three year sample.
So if zone starts are the reason why guys like Grossmann and Karlsson have their respective numbers, why is McDonagh posting such significantly better numbers than Girardi? There is no way it is possible that for that 25% of the time when they aren't together that Girardi's minutes are so much tougher, because the actual percentage of zones where they take their faceoffs remain similar.
3. The range isn't as great as you think. While many like to use the relative numbers, a much easier thing to do is just look at the total % of faceoffs taken in each zone. They almost always end up close to 33% in each zone. Players who are really good tend to have more Offensive faceofffs, as explained in point #1
4. "Defensive zone start" Doesn't mean right in front of your goalie.
5. Players shouldn't be penalized for how a coach deploys them. A good coach is going to deploy his player where he will get the most value out of them.
Just because a player has an exploitable skill doesn't mean his ratings should be handicapped.
Overall, I believe that defensive zone starts make a defenseman's job harder, but just not to the degree it's made out to be. The main reason to discount them is because they're often used as an excuse or to justify something when in fact it seems logical that a players play will have more effect on his zone starts than zone starts will have on his play.
Blogger • 
Join the Discussion:
»