Sunday, August 31st, 2014

A Game of Puck Position

22

Greg Ballentine over at The Puck Stops Here explains why winning faceoffs doesn’t correlate with winning hockey games.

The only reason we talk about faceoff proficiency is because the statistic is collected. All evidence – and common sense – tells us faceoffs are not worth talking about. Nobody talked about them before the NHL published the data but once the numbers were out there, the media manufactured a myth to justify talking about it.

Conventional wisdom argues that hockey is a game of puck possession.

This is a relatively new piece of conventional wisdom, something that came along at about the same time that hockey announcers were handed the new statistic to deliver up to fans. “Hockey is a game of puck possession,” declared every colour analyst, “And puck possession starts with the faceoff…”

But what exactly do we mean by this piece of conventional wisdom? I don’t think the description is apt at all. Of course it is usually better to have the puck than for the opponents to have it. That’s a given, but that doesn’t mean the game is about puck possession. In fact, hockey is a game of territory.

Basketball is a possession game and it is easy to see the distinction. Hockey teams would much prefer to see their opponents in possession of the puck behind their net than have the puck themselves behind their own net. In basketball, teams prefer to have the ball everywhere. In hockey, teams voluntarily turn the puck over in exchange for territory many dozens of times a game. In basketball, teams never voluntarily give up the ball.

Calling hockey a game of puck possession makes zero sense. A couple of years back, Kelly Hrudey showed a 40 second Igor Larionov shift that saw Igor recover the puck for the Red Wings on five different occasions in five different spots on the ice. Aside from the fact that the puck was following Larionov around it was an unremarkable sequence. It was a 40 second clip that nobody watching thought was unusual yet the puck turned over eleven times. How can something like that be squared with the conventional wisdom?

Offense in hockey is about finding (or creating) open ice and moving the puck through it to the opponent’s goal. Give any NHL player open ice and he will carry or pass or shoot the puck through it. Defense in hockey is about denying open ice to the offense. The key is the ice – open or not – not puck possession. If the ice is open, the puck is moving through it, and if the ice is not open the puck is usually going to get turned over. The position of the faceoff is more important than who wins it. The puck marks the amount of the ice controlled by one side or the other, much like the football marks the amount of territory controlled by football teams.

Hockey is a game of puck position, not puck possession.

Be Sociable, Share!

Comments

22 Responses to “A Game of Puck Position”
  1. Cookie La Rue says:

    I think so too. It’s a game of transition.
    If you can change your defensive play without the puck into offensive attacks relatively fast you have the most wide open spaces.

  2. Tyler says:

    I agree with his assertion but think he’s wrong about why this doesn’t matter. He kind of comes close to why it doesn’t matter when he notes that the puck changes possession so frequently on giveaways and takeaways. Faceoffs are like hits and blocked shots-they don’t tell enough of the story to matter.

  3. baroose says:

    While I agree with the contention that puck position is more important than puck posession, winning faceoffs is important. At the point of the faceoff, there is a substantial amount of open ice. The team that initially controls the puck off the faceoff has the initiative in using all that open ice for a shot or a breakout or whatever.

  4. Egil says:

    Would a team rather the other team have the puck in their zone or would they rather have the puck behind their net? Faceoffs in the offensive or defensive zone are important, the difference between the two outcomes is striking. Winning offensive zone faceoffs on the PP is also extremely important. Offensive/Defensive zone faceoffs dictate puck position as well as posession.

    Neutral zone faceoffs are a different matter though. As an example, take a faceoff just outside the blueline. The two primary outcomes are:

    Team A wins the faceoff back into their zone, were they proceed to clear the zone pretty quickly.
    -or-
    Team B wins the faceoff and then dumps the puck into Team A’s zone, who will then be able to get it back out (starting from standing still isn’t the best way to generate a forecheck).

    In both cases, the outcome is very similar. But, if Team B is on the PP say, then that faceoff is slightly more important. Powerplays, the good ones at least, are a game of puck possession. HAving the puck is always preferable to them having the puck.

  5. Tom Benjamin says:

    Faceoffs in the offensive or defensive zone are important, the difference between the two outcomes is striking. Winning offensive zone faceoffs on the PP is also extremely important. Offensive/Defensive zone faceoffs dictate puck position as well as posession.

    It depends on what you mean by important, I guess. It certainly isn’t more important than a battle along the boards for a loose puck. In that case, the winner of the battle emerges with the puck and the loser loses position as well as the puck.

    Losing a faceoff in the offensive end is not the end of the world even in a power play situation. Unless a mistake is made, there is no clear path for the puck to take to the net. In the case when the faceoff is lost and a mistake is made, the blame for the scoring opportunity rests mostly – 95%? Higher? – with the mistake.

    Similarly, winning the defensive zone faceoff is better than losing it, but even when the faceoff is won, the puck has to be cleared. It is probably better to lose the faceoff cleanly than win it and fail to clear the puck. In the first case, everyone is in good position. In the second, the failure to clear will almost always catch somebody going the wrong way. Again it is the mistake that is critical, not the fact that the draw was won or lost.

    To me the most fascinating part of the controversy (if we want to call it that) is not whether faceoffs fall into the “not very important” category or “somewhat important” category.

    A piece of trivia has been inflated way beyond its actual importance even for someone who thinks faceoffs are “somewhat important”. The fascinating part to me is why – I think it is simply because a statistic has been published. It is the flip side of the reasoning error that ignores information simply because it can’t be measured. Just because something is easily measured does not make it important.

    The second most interesting facet is because winning faceoffs does not correlate with winning, the statistic a myth about hockey being a game of puck possession is required to justify it. That’s the really misleading part. The game has been defined incorrectly because the number was published.

  6. Roberto says:

    To me, the only faceoffs that really matter are during a penalty kill. Win a faceoff while on the penalty kill, and it’s usually good for knocking 20-30 seconds off of the other team’s power play. Everywhere else, it usually comes out in the wash.

  7. Mike W says:

    I suppose it’s true that no one means what they say when they talk about puck possession, Tom. Otherwise, you wouldn’t hear coaches talking about “dump and chase” and an “aggressive forecheck” as if it were a good thing.

    Poor hockey commentator diction is to blame and I guess I’m guilty of using it, too. What we really need is a time-in-the-offensive zone type stat currently found only in NHL video games.

  8. Tom Benjamin says:

    I suppose it’s true that no one means what they say when they talk about puck possession, Tom. Otherwise, you wouldn’t hear coaches talking about “dump and chase” and an “aggressive forecheck” as if it were a good thing.

    Yes. We also wouldn’t applaud great efforts to get the puck out of our end either. We’d moan about an easy turnover. The dumpout is a good play because it forces the offense to clear the offensive zone. Teams happily trade the puck for a few feet of ice.

    What we really need is a time-in-the-offensive zone type stat currently found only in NHL video games.

    They did this for quite a while, but it doesn’t show what we would think it should show. The territorial battle is too subtle to show up even in this statistic.

    Even within the defensive zone, the tradeoff is made. Most teams now collapse to the net in a tight know. They will permit teams to cycle around the perimeter. They won’t give up the front of the net – territory – to pursue the puck aggressively.

    There are lots of games where the losing team seemed to have the puck most of the night. (I’d put the Calgary-New Jersey game in that category.) The team that plays the best without the puck almost always wins.

  9. magnum44 says:

    One other reason I think face-offs are important is because winning a faceoff increases the teams chances of getting one of their set plays to work, since off the faceoff everyone is positioned exactly as they are on the blackboard and are all on the same page at that moment. I agree it is a over rated stat but it does hold some importance. Every Edmonton game I watch I have to listen to them drone on about how great oilers are on the faceoff, but every edge you can get helps, and for a team like the Oilers that one aspect of their game could be the difference between making the playoffs and not.

  10. grampapinhead says:

    You have written an excellent piece that puts things back in perspective.

    Conventional wisdom argues that hockey is a game of puck possession.

    Of course it is true, and the first rule of jounalism is to state the obvious and you will achieve respect for being concise. However it is important to apply the law of inverse proportions to any truism. The more times you possess the puck in your net, the more likely you are to lose. The more likely, also, that the media and analysts will overanalyze any available stat in order to come up with scapegoat. Next time you watch a game, count the # of times a faceoff is won and puck possession is lost after one player touches it. you may be surprised at how the faceoff stats mean little, and the important fact is the skill of the players surrounding that faceoff win/loss. Faceoffs are important, that is why the coaches go to great lengths to control who is available in case the first choice is bounced from the faceoff. Faceoffs are important, faceoff stats are not.

  11. James Mirtle says:

    Faceoffs are important, faceoff stats are not.

    I think Gramps has it here. Perhaps for individuals, there’s a difference between a guy like Brind’Amour, who wins 65% of faceoffs, and Crosby, who wins 40%. Between teams, however, the difference are negligible.

    I would still guess, however, that the team in a particular game who wins more faceoffs is more likely to have a higher shot total.

  12. Egil says:

    You mention an important point. Faceoffs are somewhat easily measured, and is now a published stat. But the same can be applied to goals, assists, +/- and PIM’s. All are easily measured, but none of these stats is perfect.

    Even something like the assist is horribly flawed. Often, the guy who creates the goal (the player screening the goalie) gets no credit, whereas the player who made a simple pass gets an assist.

    No hockey stat measures how good of a player someone is. They are all flawed!

    However, some stats that would be useful:

    Shooting % should be calculated based on shots, missed shots and shots that get blocked. Not just on shots that hit the net.

    PP and PK percentages should acount for shorthanded goals. Each shorthanded goal should count for -1 goals when calculating PP%, and +1 goal when calculating PK%.

    Assists should be given out to the 2 players whose play helped cause the goal the most. If that is the guy screening the goalie, or the guy who pressured the defenseman into coughing up the puck, or the guy who made the sweet pass to the goal scorer.

  13. Rob says:

    Face offs are am important component of creating scoring chances and elliminating scoring chances particularly with the excessive amount of PP and Pk time in todays RHL (referees hockey league).

    I would prefer to see stat that tells the number of scoring chances created and how many against.

    Winning face-offs, winning the battles on the boards, (does that still exist?) effective forechecking and breakouts help to create scoring chances for and decrease scoring chances against.

    Does all of that add up to puck possession?

  14. DonK says:

    This reminds me about reading discussions about the value of stolen bases in baseball. Just as stolen bases in general are an unimportant offensive stat (but a particular steal at the right time can be very valuable), so, too, with faceoffs — winning a particular faceoff can lead directly to a goal (or a clear on a penalty, etc.), overall faceoff numbers don’t appear to have a lot of value. Also, the number of faceoffs is down about 10 a game (fewer offsides and icings — it’s why game times are about the same even with all the added penalties).

    Also, having watched Isles-Oil in person tonight, faceoff wins for individuals can be deceiving. Many times tonight, the centers tied each other up and the team (and the center involved) “won” the faceoff because its second man on the puck was able to wriggle it free.

    Given all this, I’d still rather win 60% of the faceoffs than 40%. You still can’t score without the puck.

    The NHL used to put time in zone on the computerized scoresheets, but dropped it a few years ago. In most cases, it was interesting but didn’t really prove much.

  15. Tom Benjamin says:

    This reminds me about reading discussions about the value of stolen bases in baseball. Just as stolen bases in general are an unimportant offensive stat (but a particular steal at the right time can be very valuable),

    The comparison doesn’t really work for me. I don’t think anybody suggested that the stolen base statistic was unimportant. The argument was about the value of the stolen base and whether attempting lots of steals was a good idea. Nobody was suggesting the statistic itself was meaningless or should not be published.

    In this case, I’m arguing that this statistic is so trivial, it’s harmful. It’s like counting foul balls and attaching value to the players who hit a lot of them. There probably is some value in it, but it is also surely trivial. It’s harmful if a rational explanation is invented to establish its importance as in, “Everybody knows that wearing out the starting pitcher is a key to winning baseball, and nobody wears them out like Freddy Foulball.”

    so, too, with faceoffs — winning a particular faceoff can lead directly to a goal (or a clear on a penalty, etc.), overall faceoff numbers don’t appear to have a lot of value.

    I’d argue that winning a particular faceoff can’t lead directly to a goal. It still usually takes a mistake in the defensive coverage and it is the mistake that leads directly to the goal.

    All that happens after winning a faceoff in the other guy’s end is that the team has control of the puck in the offensive end. That happens dozens and dozens – hundreds? – of times in every game anyway.

    The difference between the having the best versus the worst (a 60-40 split instead of 50-50) is one faceoff in ten. Assuming about ten offensive end faceoffs in the game, that gives a team one extra possession in the offensive end in a game. The team with the best faceoff guy in the league gets dozens and dozens of possessions plus one over the mere dozens and dozens of possessions the team with the worst faceoff guy gets.

    If you are the defensive team it means you merely have to play it correctly to get it out. If you lose the draw, you have to first win the puck and then you have to play it correctly. That’s an edge, big enough for coaches to strategise around because they will take every tiny edge they can get. But it is not a huge edge in a game where you have to win the puck from the other guy dozens and dozens (hundreds?) of times in a game.

    Even on the power play it is not a big deal. The typical power play gets set up several times in the offensive zone and a successful PK usually involves getting the puck and clearing the zone several times. Winning the faceoff and clearing the zone right off the bat is nice because you’ve won the first of several battles you have to win to kill the penalty.

    Given all this, I’d still rather win 60% of the faceoffs than 40%.

    I think there is 100% agreement on this point.

    I see it as a duffer preferring to play white in a chess game against another duffer. If I’m magically able to pick the hand with the white pawn 60% of the time am I a better chess player?

    You still can’t score without the puck.

    Steve Smith. (I couldn’t resist.)

  16. Tyler says:

    Steve Smith. (I couldn’t resist.)

    Game winning goals put into your own net in crucial game sevens correlates very well with Stanley Cups in the two years preceding and following the goal. I believe it’s a correlation of 1, although I could be mistaken.

    Who says stats are useless.

  17. InsultComicDog says:

    When a guy like Lindros or Brind’Amour would win a faceoff, it was another way of gaining possession without having to be a particularly fast skater. These days, there are usually speedy guys on the ice on both teams so winning the faceoff is even less of a factor than it used to be.

  18. Tom Benjamin says:

    Winning the faceoff and clearing the zone right off the bat is nice because you’ve won the first of several battles you have to win to kill the penalty.

    Thinking about this point, I wonder whether it isn’t sometimes better to lose the faceoff to start a penalty. Isn’t it better if the team on the PP runs the puck around for 30 seconds, loses it, and then it gets whacked down the ice? The PK changes and the team is set to use three sets of legs against two PP units. If you win the faceoff and immediately whack it down the ice, you are more likely to get caught with tired penalty killers when the PP makes its second attempt.

  19. will says:

    In other words, the important stat, as far as face-offs go, is where they are taking place? I think this would be interesting, maybe they already do that in Canada, but in Socal, when they break the stat down, they still only tell you who is winning the offensive or defensive face-offs, rather than where they are happening.

  20. Tom Benjamin says:

    In other words, the important stat, as far as face-offs go, is where they are taking place?

    I don’t think so, although that is more important than who wins them. It’s simply not an important stat at all. If the NHL had never started counting them, we would know more about hockey.

    We wouldn’t believe that hockey was a game of puck possession.

  21. Tom Benjamin says:

    A friend pointed out to me that there’s been a noticeable lack of ‘shot’s off the draw.’ This appears to be true and I suppose supports your claim. What’s most important is the quality of puck possession: you most want possession of the puck when you’re team is skating towards the opponents net, with speed, and preferably leaving opposing players behind. How’s that for conventional wisdom?

    Yes, but I think this begs the question. How is this achieved? Finding the open ice to fly with the puck through the neutral zone is the hard part. NHL players are really, really good. The worst of them can make a good pass if there is a clear path to his target.

    The key to offense is creating that open ice. In great games that open ice can be created in transition. It happens because a creative player creates it with a clever move or a threaded pass or a deceptive change in speed. Most frequently it happens because a defender makes a mistake, misreads the play and gets caught. It’s a game of mistakes – a piece of conventional wisdom I do accept.

  22. Name says:

    Disagree with this. Any statement that goes “Hockey is a game of…” and isn’t followed up with about 100 different nouns does not tell the full story. There is some wisdom in this post, and some truth, but “faceoffs aren’t important, and weren’t until the stat was published” is not true. An offensive zone faceoff won by the offensive team cleanly back to a defenseman allows that team to immediately shoot at goal, when they wouldn’t have been able to do so had they lost the faceoff. If one in 50 shots from the point are successfully deflected into the net by one means or another, for example, then every faceoff won offensive zone faceoff won that is shot at the goal from the point is an extra 2% chance at scoring. If one team wins 10 more offensive zone faceoffs in this matter than the opposition in a given game, that’s a 20% chance of scoring, or basically a .2 goal advantage to that team.

    Not to mention an offensive zone faceoff won on a powerplay, vs a loss one, is a 15 second or longer powerplay, or however long it takes to re-enter the zone after the puck is cleared. If you win the faceoff, you get to set up your powerplay immediately, at the 2 minute mark, and start attacking. If you lose it and the puck is cleared, you won’t get it set up until 1:50 at minimum, if not longer. I don’t have to explain why longer powerplays are an advantage.

    Faceoffs do matter, unfortunately. They shouldn’t, they’re the most boring, stupid part of the game, but as it stands, some teams are better than others at cheating at them, and manipulating the current, flawed faceoff system to their advantage. The whole point of faceoffs is to just get the puck back in play, flipping a coin as to who gets it. That’s why they should always be 50-50, ideally. When they’re 50-50, they are a nonfactor, as they should be. The way the game is now, with teams like the San Jose Sharks having a special video coach to help break down and manipulate the faceoff system, the way certain referees drop the puck, all of that, to manipulate it to their advantage, now faceoffs are a factor, when they shouldn’t be, and they give an unfair advantage to teams like the Sharks, when no advantage has been earned through play.

Speak Your Mind

Tell us what you're thinking...
and oh, if you want a pic to show with your comment, go get a gravatar!