The law dictates that any story that so much mentions the NHL All-Star Game in Ottawa must be accompanied by a picture of Daniel Alfredsson

There was a lot of talk this weekend pointing at the All-Star Game in a sort of derogatory sense, that true hockey fans were somehow above the game. Of course, more hockey fans I know watched hockey’s All-Star Game as opposed to football fans watching the Pro Bowl, and while I understand I’m looking at a small sample, there’s always one or two real cool things that pop up in hockey’s game that make it worth watching.

With the weekend behind us, we can focus on playoff races, the unpredictable and off-season-storyline making playoff races. An entire team’s season will be judged on a fraction of games that will allow old-timey pressbox scribes to determine whether a team was clutch enough or had enough “it” factor to win in this league. Percentage-induced hot and cold streaks will be furiously debated during the coming summer months.

What I always like about this time of year is how there’s also a segment of hockey media that like to discuss “how the standings would look without the shootout” or “how the standings would look if regulation wins were worth more”.

(If you’re a stat-head, it’s worth checking out a series of posts by Driving Play’s JaredL on win expectancy at certain stages in the game. Looks like he’s out to prove that the current point format the NHL uses is flawed, but we didn’t really need a textbook for that. It’s some cool stuff regardless)

Pavel Datsyuk, surprisingly not the NHL's best shootout specialist, salutes a very large Joe Louis Arena crowd after another victory

I believe, generally, that a shootout is a coin-flip way to determine a winner. The NFL takes this to extremes and actually flips a coin if a game is tied after an allotment of time. I’m not sure that being really really good at a shootout as opposed to average will help a team at all in the standings. Put it this way: the New Jersey Devils, the best team in the shootout since its inception, average 11.7 shootouts per 82 games. They’ve won 7.6 of those 11.7. If they were “average” as opposed to the “really really good” they are now, they would lose, on average, 1.7 points per season. Being the best at this isn’t even worth a win over the course of 82 games for this team.

So it’s funny when I hear “skills competition” when people refer to a shootout. Flipping a coin is not a skill. You may get lucky placing all your money on black at the roulette table, but you wouldn’t rely on black to provide you with a sustainable income. If the shootout were a true skills competition, I think we’d be able to easily see which players were the best at this sort of thing.

I posited the question last night “which player is the best at the shootout” not expecting that anybody would correctly guess a player in the Top 5 at this sort of thing (actually, one person did). Guesses were pretty well split between Pavel Datsyuk and Jussi Jokinen, who are both no slouches at the shootout, certainly, at 9th and 11th respectively, among players who have taken 30 shots in their careers.

The only player in the Top Five named after ten minutes? Ales Kotalik, but I think the guy was making a joke when he said that.

I used to always pick Michal Handzus in playoff pools for some reason. Why didn't any of you stop me?

It just so happens that two All-Stars are in the Top 20 all-time in shootouts, and while we can gawk at endless Youtube clips of sweet hands on the guys we saw play in Ottawa on Sunday (like Dan Girardi), the fact of the matter is that many of the shooters that we saw at this weekend’s Breakaway Challenge or Elimination Shootout were totally average at this thing. The shootout isn’t for All-Stars, maybe.

Last week, San Jose and Edmonton were playing a game nationally televised in Canada that went to a shootout. Fans from the Eastern time zone who had managed to stay awake seemed confused that Michal Handzus, whoever he was, was shooting for the Sharks. He missed in rather brutal fashion and a few yuks were exchanged online. This was funny to me, because I’d seen Handzus beat the Canucks in a shootout in person just a week-and-a-half earlier, and his miss against Edmonton dropped his career average to 50%, which is tied for 3rd all-time.

The same thing happened when Randy Cunneyworth in Montreal sent Tomas Kaberle out for a shot in one shootout and Scott Gomez in another, which drew a lot of criticism (the Gomez choice in particular had some creative guy from the Montreal Gazette write: “Cunneyworth must be the only person in Montreal who actually thought Gomez might score”).

In fact, the only healthy Montreal Canadiens better than Gomez in the shootout in their careers are David Desharnais and Kaberle himself. Desharnais actually shot in that shootout, and his result was the same as Gomez’s, but he didn’t draw any mocking for some reason, but that can’t possibly be because Montreal media and management are horrible at judging players.

Alexander Ovechkin, oft regarded as one of hockey's most skilled, has scored just 18 times out of 60 in the shootout.

There are players, who, despite putting up success rates of less than a 37% league average, continue to earn a lot of shots thanks to their star power. Vincent Lecavalier, Patrice Bergeron and Alexander Ovechkin are three players in particular who any fans would love to see compete in the post-game gimmick, but neither of which are very good at this.

I see an abject failure in fans and media to properly understand the shootout or its implications. Last season, the Carolina Hurricanes missed the playoffs because they were unlucky enough to go 5-5 in the shootout, while the New York Rangers, who finished just two points ahead, went 9-3. Had they gone 6-6, Carolina are the ones who get stomped by Washington in the first round. In the West, Dallas went 5-7 while Los Angeles went 10-2. Reverse the records, and that’s Dallas facing off against San Jose in the first round.

But were those teams “good” at the shootout as opposed to lucky? Ask the Kings, who are now 4-6 in the supposed skills competition this year, while New York are 2-3. Getting really good at this particular aspect of the game is probably hard, but, if you ask any Devils fan, they’d probably trade a few shootout wins for a few more in regulation. This little gimmick causes us so many math problems during the season to figure out the standings. It’s an extension of the curse of close games that trick us into thinking that teams are better than they really are. It’s an optical illusion, and the extra points over a small sample can make all the difference in a team’s season.

But there’s still so little understanding of the event. Sure, maybe if I asked 100 people, one of them would have correctly guessed that Frans Nielsen, with 19 goals in 32 attempts, is the best at shootouts. And maybe one of them would have also correctly guessed that Erik Christensen was second among players who had taken 30 career shots.

It’s not a skills competition because nobody can guess who is the best at this sort of thing.

Comments (19)

  1. So wait, who are the top 5 (or top 20) in shootouts all times?

  2. of guys who’ve taken more than 15 attempts:

    1. Frans Nielson (59.4% – 32 attempts)
    2.Eric Christensen (52.2% – 46 attempts)
    3. Ales Kotalik (50% – 44 attempts)
    4. Jonathan Teows (50% – 44 attempts)
    T5. Michael Handzus (50% – 30 attempts)
    T5. Jarret Stoll (50% – 30 attempts)

  3. Ugh. This is like those old bogus statgeek claims that no such thing as shot quality exists despite the fact that it obviously does.

    Some players are good at shootouts. Some players are not. A team with 3 good players at shootouts will win more of them than a team with 3 less-good players. A team with a better goalie at shootouts will win more than a team without (some goalies are much better at defending this than others). Not all that complicated and certainly not a coinflip. You can’t change the probabilities of a coin you’re flipping by acquiring more specialists than the person betting it lands tails.

    And yeah, everybody who knows who Erik Christensen is knows the only reason Erik Christensen hasn’t washed out of the NHL is because he’s good (good, not lucky) at shootouts. This has been documented extensively in every market he’s ever played in and covered by the likes of TSN and the New York Times. You can find articles from his rookie season talking about how good he is at shootouts. If he’s still second in the league, percentage-wise, six years later, that seems an awful lot like he’s got a sustainable skill, not a ton of blind luck.

    • The claim that statgeeks claim there is no shot quality is also bogus… or at least, over-simplified. The statgeeks say two things about it: one, that it is currently impossible to reliably measure shot quality over the long-term; two, that the difference among elite players and teams is generally so narrow that in any given individual instance, it is impossible to reliably predict an outcome.

      If I sent Frans, Kotalik, and Toews in a shootout against a goalie not known for his shootouts – say, Nicklas Backstrom of Minnesota with his .570 sv% in the skills competition… and you sent three average-to-poor shooters at Henrik Lundqvist (.764, fourth-best all-time among goalies with 100 or more shots faced)… you can expect to win more than 50% of those contests over the course of a few years, but in any one of them, nobody could tell if you would win or lose.

      • I beg your pardon… that should read “reliably measure shot quality over the SHORT term.” In any short stretch, anyone might run into hot keepers or cold luck and miss 30 shots in a row, while a relatively-unskilled player gets a bunch of easy tap-ins or lucky bounces off defenders. A lot of what goes into scoring in the NHL is outside the direct control of the shooter and the goalie.

      • “The claim that statgeeks claim there is no shot quality is also bogus… or at least, over-simplified. The statgeeks say two things about it: one, that it is currently impossible to reliably measure shot quality over the long-term; two, that the difference among elite players and teams is generally so narrow that in any given individual instance, it is impossible to reliably predict an outcome.”

        http://www.arcticicehockey.com/2010/9/20/1696352/team-shooting-percentage-and-shot

        The thesis here is that, while some individual shots will go in more than others, getting better shots is mostly a matter of random error and results in nearly no wins. In addition, this same website held a contest declaring shot quality irrelevant and brazenly daring anybody to prove them wrong which, iirc, they lost.

        Arctic Ice hockey was formerly SBN’s behindthenet blog, so it’s not like I’m pulling the random ravings of some yahoo out of the most obscure corner of the internet.

        • That’s how things move forward. You put up a theory. You present all your data. Other people test it and either prove the theoryas is, or improve it. I mean, you’re quoting something they wrote 15 months ago -things have advanced since then.

          • “I mean, you’re quoting something they wrote 15 months ago -things have advanced since then.”

            No kidding. That’s why I compared this “shootout as a coinflip” theory to “old bogus statgeek claims that no such thing as shot quality exists despite the fact that it obviously does.”

  4. I would be shocked – SHOCKED, I tell you – if nobody from this blog named Nielsen as the best in shootouts. His signature move is the namesake of this blog!

    I’m also a little shocked that Jussi Jokinen isn’t on that list, either.

  5. The danish backhand of judgement. 60% of the time it works every time.

  6. I would’ve nailed Stoll because of his crazy good season last year, and everyone would’ve nailed Christensen (I hope). Stoll being scouted a bit more by opposing goalies is pretty obvious, as they’re playing him a LOT differently this year.

    (It’s always hilarious when the Kings get into a shootout on the road and the opposing commentators on the TV broadcast wonder why they’re sending Stoll out there. Way to do the research, guys!)

    Nielson I didn’t know about. But when you’re hitting on that percentage of your attempts, it goes far beyond “it’s random.”

  7. Carey Price actually alluded to another factor during the All Star game that impacts this – video tape. The goalies watch tape of the players they face regularly and generally know their ‘go to’ moves.

    This could mean that in order to be sucessful a player needs luck (shoot outs end up being against goalies that havent watched their tape) or that they need to continually work on new moves (I am not sure that is worth the effort). If that is the case, randomly selecting players so that the goalie doesnt know what they are going to do might be the best idea.

    • True – which is why teams might actually be smart to work a lower-percentage player in there occasionally, or at least tweak the order once in a while.

  8. Yeah if you have one of these guys on your team you know it. Any halfways serious Sabres fan knew about “automatic” Al Kotalik being almost unstoppable in the Shootout. As a player, he peaked low and early, but I never saw him miss after overtime in 2005-06. Forehand deke to…wait for it…Backhand Shelf every time. Watch some tape sometime goalies! The announcers would joke about it being “his” move. Like Zoolander, he only had “one look”.

  9. Only us Islander fans know of the Danish Backhand of Judgement.

  10. Toews is the best shootout player in the World hands down..

    I do believe he single handidly won a Gold Medal for team Canada scoring 4 times in a row against Team USA in the World Jr’s…

  11. Also TJ Oshie has been atrocious this year I think he’s like 1 for 7 this year, but still Tied for 5th best in the NHL with 26 attempts (50%) in his short career..

  12. unfortunately this article completely ignors a few facts;
    1. that teams also have goalies….the other half of the equation

    2. that there is now more than ever increadible incentive for teams to go into overtime….think about it if the shootouts basically a coin flip, a team going to shootout 82 straight games should finish with about 123 points….there is something seriously wrong with that.

    3. that 30 shots is an obsurredly small sample….i’d say this is like judging an MLB player on 30 at bats….you certainly at least need to consider goalies faced at this small of a sample size, because it simply will not come out in the wash as is relied on for baseball stats

    4. and finally I simply don’t believe you that no one guessed jonathan toews…and you go ahead and put frans out there I’ll take Toews any day, whens the last time frans scored 4 shootout goals in the same game….never

Leave a Reply

Your email address will not be published. Required fields are marked *