Talkin’ faceoffs

A quick thought on faceoffs here, since Randy Carlyle is scratching his head about whether or not the Boston Bruins cheat in the dot. Frankly, they may, but has that really had an effect on the series? 

I’ve written at both The Leafs Nation and Canucks Army with thoughts about how overvalued faceoffs are. “Faceoffs” of this year seems to be the “shot blocking” of last year, when everybody was discussing the success of the Washington Capitals and New York Rangers blocking shots and playing stifling, defensive hockey.

I’ve had the luxury of sitting in one of them fancy NHL press boxes this post-season and one thing I’ve noticed, other than the trend of writers dutifully copying down goal and penalty announcements by hand as if there were no place to find that information, is how much printed paper gets handed out at the end of a period.

This isn’t a new thing. I don’t have too much use for scoresheets since any useful information is available upon reference because’s game sheets are pretty detailed, and the play-by-play chart can be downloaded and scraped for information. One thing I quite like is the Event Summary, and it’s very useful after the game. It has attempted shot numbers for every player (shots on goal, attempts blocked, and missed shots) and breaks down shooting by strength for a team. There is, however, lots and lots of crap on the page. 

Anyway, the event summary, at least in Vancouver, gets handed out at the end of each period. White for the first period, blue for the second period, and green for the third. Along with the useful shot numbers, a number of other categories show up: the RTSS stuff that I dismiss because it’s so inconsistent building-to-building: hits, giveaways, takeaways, and then, consistent, yet hardly useful, is “faceoffs”.

After a period or so, faceoff stats look fairly disproportionate. Say there are 16 total draws in a period, and the home team wins 6 of them. On the scoresheet, that’s printed up at 38%, and there’s quite a discrepancy between teams after that amount of time. 

38% on the draw does not look very good, and oftentimes writers (I do this all the time) tweet out statistics that appear to be illuminating when we try to determine why teams are winning and why teams are losing (the best example after a single period is “small sample size”). 

But while 62-38 looks pretty bad, the difference between 38% and 50% for the home team at that point is quite literally two faceoffs.

Now, the Leafs have gotten crushed on faceoffs this series. They’re 56-of-129 if you discount neutral zone draws. (One Vancouver writer, Tony Gallagher, wrote an interesting anecdote about how Brent Peterson would almost purposefully lose draws in the neutral zone because they didn’t really matter and he didn’t want opponents learning about his style) That’s 17 away from 50%.

17 is lots, at this point. The team is 43.4% in the series on draws in the offensive and defensive zone, but that’s still a long way from costing the team a goal. New GM of the Edmonton Oilers Craig MacTavish quoted an analysis he read as “every 40 draws lost in the defensive zone costs you a goal”. I’m not sure who he was quoting, but that’s similar to numbers I’ve seen in the past, so we’ll go with it. 

Now, there are examples this series of goals going in directly off of faceoffs, but what you need to consider is even the best faceoff men will lose a draw 45% of the time. That’s hardly distinguishable from a coin flip when you’re looking at maybe 40 events over the series. If faceoffs were truly that important, they’d show up in areas beyond the margins: they’d correlate higher with wins and puck possession. 

Just as a thought exercise, I went through my scoring chance data for Maple Leafs centres and broke it down by game, along with faceoffs won and lost in the offensive and defensive zones (this is another cool app, the ‘Faceoff Report‘. I mainly used it to find out which lines the home team tried to match up against, and in which zones, before apps started working again this season).

Game 1 Chances F Chances V Faceoffs W Faceoffs L
Tyler Bozak 3 4 12 8
Mikhail Grabovski 1 2 3 5
Nazem Kadri 0 2 1 2
Jay McClement 0 4 2 3
Game 2 Chances F Chances V Faceoffs W Faceoffs L
Tyler Bozak 8 6 11 17
Mikhail Grabovski 3 2 3 3
Nazem Kadri 5 1 2 2
Jay McClement 0 0 3 6
Game 3 Chances F Chances V Faceoffs W Faceoffs L
Tyler Bozak 4 4 11 13
Mikhail Grabovski 1 2 5 5
Nazem Kadri 1 3 3 6
Jay McClement 0 3 0 3

Jay McClement is getting killed this series, but the fourth line has been killed all season, even when McClement was winning more draws than he was losing. Tyler Bozak’s worst game in scoring chances came in Game 1, which was his best in faceoffs. Nazem Kadri and Mikhail Grabovski have both chance numbers and faceoff numbers that are pretty indistinguishable game-to-game.

Point is, if there’s a trend, I’m not really seeing it. The Leafs won 36% of the draws in Game 2, which was their best game overall, and 54% of the draws in Game 1, which was their worst overall.

The message isn’t “lose faceoffs” but it’s moreso “worry more about other aspects of the game”.

  • Miker

    It’s a micro vs. macro thing, basically.

    On a faceoff-by-faceoff basis, you want to be winning as many as possible. When you send out two centres for a D-zone draw in the final minute of a 1-1 game, you want to win that draw.

    But on a macro level, a franchise worrying about collecting guys with a high FO%, or blaming their record on poor faceoff stats, is focusing on the wrong areas.

  • jasken

    Statistics don’t lie, but they also don’t tell the whole story.

    Frequent O-zone faceoff wins, if they don’t lead directly to a scoring chance, at minimum open the door for “something lucky to happen” (i.e. deflection on a point-shot)… and luck is a huge part of hockey, or any sport, really.

    • This is true.

      But you won’t find a player who will win you 100% of the faceoffs.

      You can find a player who can win you 55%. The difference between the best and worst faceoff guys in the game ends up being two or three per game, or one per period.

      • Agreed, though I think all the hoopla surrounding faceoffs in this series stems from team rate – not individual. And the Bruins having a 3:2 success rate is huge.

        That said, I agree correlating faceoff wins to goals is low-grade apophenia since a billion things can happen between those two events.