Hits and Giggles, and Faceoffs

Announcer 1: “If the Oilers are going to dig themselves out of this hole, they really need to ramp up their physical game”

Announcer 2: “They, especially Connor McDavid, need to start winning some more faceoffs. 48.5% just isn’t good enough. Also, how the heck do you win half a faceoff?”

OK, we’ve heard from the Brainiacs. Are lack of hits and losing faceoffs really what ails the Oilers?

(note: data is all situations, as of January 13th; downloaded from corsica.hockey)


hits and giggles 1

This chart shows the relationship for every NHL team between hit differential and goal differential. Mostly there is no relationship, but to the extent we can find one, it is mildly negative. Teams that outhit their opponents show a mild tendency to be outscored by those opponents.

That certainly seems to be true for the Oilers, who’ve spent most of the season with a poor goal differential and in the bottom 5 of the standings, but are best in the league at hitting.

C and C Statistics Factory

Now before anyone goes all galaxy-brain on this as proof that hitting is bad, let’s remember that correlation is not causation. The fact that there is a slightly negative relationship here doesn’t mean that more hitting means you’ll score less.

It’s just as plausible, probably moreso, the other way round – that teams that score a lot tend to have the puck a lot … so they tend to be too busy skating, passing, and scoring to worry about hitting people. However, since they have the puck, they do tend to themselves get hit.

Even while being careful to avoid the correlation/causation trap, the fact that there’s really only this weak negative relationship does mean we can conclude a couple of things.

First, we know that getting back in a game means scoring more. So when it comes to hitting, unless the hitting is purposeful and focused on reacquiring the puck when you don’t have it, the idea that “we need to hit more to get back in this game” is probably garbage advice.

The second conclusion is a little more subtle. It’s important to recognize that even though the observed relationship is mildly negative, it doesn’t mean hitting is inherently bad.

As a matter of fact, if all else is equal, personally I’d rather sign a rugged 70 pt winger who hits a lot over a soft 70 pt winger who doesn’t hit a lot.

“All else” is almost never equal though. Which means the second key lesson we should take away is that swapping (be it in signing or trade) a “soft” 70 pt winger for a “rugged” 50 pt winger is probably going to make your team worse.

That’s because the loss of those 20 points will make scoring goals harder, but the increase in ruggedness won’t compensate for that because it isn’t really related to goal scoring.


So if hits aren’t the key, maybe faceoffs are!

hits and giggles 2

Ooops. Not faceoffs either.

That isn’t to say that certain individual faceoffs don’t matter. Of course they do. It really sucks when you lose a faceoff and get a goal scored on a few seconds later. Then again, that might have happened if you’d won the faceoff too. It just seems a little less likely. So it’s generally better to win a faceoff than not.

But what the chart does tell us is that, league-wide, faceoff prowess does not have a relationship with the ability to score goals.

Of course, as with hits, all else being equal, you’d prefer a good faceoff guy to a bad faceoff guy. But once again, “all else” is rarely equal. From a roster management perspective, if you find yourself trading goal scoring skill for faceoff winning skill, you’re probably hurting your team.

So it’s not hitting. And it’s not faceoffs. Presumably, the ability to score goals is what really matters when it comes to scoring goals.


That’s the kind of razor-sharp analysis you pay me the big bucks for, folks!

Why *Do* They Call the TV the Idiot Box?

Shockingly however, it appears that with their fixation on hits and faceoffs, neither Announcer 1 nor Announcer 2 are providing us with the kind of meaningful analysis and deep insight we’ve come to expect from those folks. Obviously, we need them to get to the next level. Shred the red, ferda!


There’s nothing profound here. The non-relationship between hits/goals and faceoffs/goals that I showed you for this season is quite consistent year to year.

And hitting and faceoff winning are by no means bad things. It’s just that they don’t impact the overall outcome of the game to anywhere near the extent the TV guys would have you believe.

In the big picture, swapping (whether it’s roster or ice time) actual goal scoring and/or prevention ability for hitting or faceoff winning is going to be a losing proposition most of the time.

To quote Lowetide quoting Arbour & Torrey: get good players, keep good players. Just make sure you’re paying attention to what makes good players actually good.

6 thoughts on “Hits and Giggles, and Faceoffs

  1. Good post. I think you’ve come to the right conclusion, but hit data should largely just be thrown in the garbage. The variance arena to arena is worse than just about any stat. Especially when it only counts as a hit when the player is separated from the puck. All those “Finish your check” hits, don’t count for anything. Same goes for checks that don’t separate the player from the puck. It’s too subjective to really have any accuracy.


    1. There’s some truth to that. Some people stick to using away data only to minimize that issue (idea being that the relative under/overcounting mostly cancels out, which I think is true). I wouldn’t use that data for doing individual level analysis, but it has enough season to season repeatability to suggest that it’s far from random. I think it’s fine for big picture conclusions.


  2. While I agree with the general premise of the article, I am somewhat skeptical when talking about hits. Physical play is more than just hitting and the way our roster is built, we are most effective when the players are engaged physically.
    We have seen a few games this season (mostly afternoon games idk why) where our players just coast around doing nothing. (I thought Lucic was supposed to stop that but what do I know ?). If this were a roster built with speedy skilled guys, we would not have come away thinking of their effort as listless. However a team of Lucic, Maroonm Strome, is not going to outskate the other team and they need to be assertive and aggressive using their size to win board battles and like to be effective.
    Hence when an oiler broadcaster says that this team needs to engage physically to get back in the game, he is unintentionally describing the roster makeup if oilers while just reading from the broadcaster’s manual.


    1. I find it tough to agree with that argument for the simple reason that I think it reverses the causation.

      I agree with the part of the premise that this is a Chiarellified roster with less skill and speed and more physicality. So naturally, when they’re engaged in the game, one of the things you see is hits. Also possession. Puck battle wins. Goals.

      When they’re not engaged you don’t see many if any of those.

      I think it’s mistaken to suggest hits drive engagement rather than engagement driving hits. The broadcaster’s manual gets cause and effect backwards!


      1. I agree with the cause and effect. Effect of being engaged is the physicality. My point is just this: Oilers possession, goals etc will always be accompanied by physical play. The way this roster is constructed, physical play can be linked to give-a-shit meter. If guys like maroon or JJ aren’t being physical, how engaged are they in the game?
        Other than 97 and 93, most of these guys are big bodies who need to be using it for puck shielding, retrieval etc. So when a broadcaster, who isn’t that bright a fellow to begin with starts associating lack of physicality with lackadaisical play, I am wiling to give him a pass.


      2. Ha ha, you are more forgiving of broadcasters foibles than am I! As squeaky and odd as he was, I miss the days of guys like Howie Meeker, who actually understood the game at a deep level and did his best to educate fans. Modern broadcasters (esp. the Oilers crew) discourse isn’t really distinguishable in quality from “drunk guy at the end of the bar yelling at the TV”


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 )

Connecting to %s