Does Alain Vigneault sit back on leads? Yes

Last night, the Vancouver Canucks nearly blew a 3-0 lead, hanging on for a 3-2 victory against the St. Louis Blues. They were not the only team to nearly waste such a lead.

Los Angeles had leads of 2-0 and 3-1, but Phoenix pulled to within one on both occasions. Boston lost 3-1 to Winnipeg after taking a 1-0 lead. Washington took a 1-0 advantage against Pittsburgh, but lost 2-1. Buffalo blew a late 2-goal lead against Montreal, only to win 3-2 in overtime. Columbus took a 4-1 lead over Nashville with 8 minutes left in the 3rd but hung on for a 4-3 win. The Islanders had a 3-1 lead over Ottawa, but lost 5-3.

A night earlier, Dallas had a 3-0 lead over Calgary but they pulled to 3-2. After Dallas scored a late goal to make it 4-2, the Flames eventually made it 4-3. Anaheim was up 4-1 over San Jose, but the Sharks made it 4-3 before the Ducks scored an empty net goal. That prompted this comment from fearthefin:

Advertisement - Continue Reading Below

If you haven’t noticed a theme, it’s that blowing leads isn’t something that’s singular to the Vancouver Canucks. Quite the opposite, in fact. This is the human element of sports, and the toughest to wrap my head around. The odds tilt in favour of the losing team, no matter what sport you seem to be playing. It’s particularly noticeable in hockey. Basketball has a fixed number of seconds a possession can last. Football has a fixed number of yards a team can cover before eventually possession goes back to the other team. In hockey, one shift can last 20 seconds or one shift can last a minute and 20 seconds.

I didn’t pull out Moneyball last night from the storage locker to simply quote one passage, but I’m going to use it as reference until concepts stick. The key isn’t to look at all problems on a micro level. I agree that the Canucks under Alain Vigneault don’t play as well with the lead as they should. However this was also the case with Marc Crawford, Pat Quinn, and the case of virtually every single coach who has ever coached a sport.

Advertisement - Continue Reading Below

Michael Lewis quotes Bill James’ 1983 Baseball Abstract, writing that “there exists a level of negative momentum,”

which acts constantly to reduce the differences between strong teams and weak teams, team which are ahead and teams which are behind, or good players and poor players. The corollaries are:

  1. Every form of strength covers one weakness and creates another, and therefore every form of strength is also a form of weakness and every weakness is a strength.
  2. The balance of strategies always favors the team which is behind.
  3. Psychology tends to pull the winners down and push the losers upwards.

It’s fairly philosophical, but the beauty about sports is that they’re played by humans, and no human is so flawlessly built that in athletic competition, they can’t be exposed in some manner. In the NHL over a six-month period, the difference between an elite team and a lottery team is 20 wins. That’s three wins per month, or a little less than one extra win per week that between the largest gap in the standings.

It’s like how the difference between a 20-goal scorer and a 30-goal scorer is one goal every two weeks. Any idiot can appreciate the difference between scoring 20 (a good player) and 30 (a great player) but visualizing alone isn’t enough to cover the difference. This is why I like to look at things at a macro level rather than a micro level. It’s not enough to analyze the Vancouver Canucks inside and out, and the way the Canucks use their players, the way they are coached and managed as it pertains to simply the Canucks. To really get a sense of the Canucks, you need to compare what they do to other teams in similar situations, or other NHL teams. How does each team operate, and what are the differences? For the differences we can quantify, are they positive ones or are they negative ones?

So do the Canucks play more conservatively while ahead, and does this negatively affect the group? There was an interesting post on Hockey Buzz today from Travis Yost, a pretty sharp dude:

Some of the more irritating commentary is how broadcasters and writers will launch shit at the wall when it comes to both playing with the lead, and playing from behind. If a team gives up a goal on a lucky bounce, they’re not mentally tough enough to win late/close hockey games. If a team strings together a couple of statistically unexpected come-from-behind victories, they’re suddenly a team that thrives from playing behind.

Travis goes on a bit about “score effects”, but then he gets to this section:

Vancouver’s a weird team. Why is this team so bad ahead, and so great from behind? This probably requires some in-depth analysis into player usage and game strategy; my guess is a team that once traded Cody Hodgson (et al.) for Zack Kassian (et al.) kind of gets away from their bread and butter. I’d like to see what their numbers look like after forty-eight games — will they normalize a bit, or is this really what Alain Vigneault wants?

The Canucks are 16th in puck-possession while they have the lead, but 1st in puck-possession while trailing. What accounts for the discrepancy?

Advertisement - Continue Reading Below

I’ll admit the following research looks at the discrepancy a different way because most of the grief I’ve heard on Twitter is how Vigneault sits on 2-goal leads. That statement is true, but not exclusive to the Canucks. has five seasons worth of data of puck-possession percentages, calculated by the rate of unblocked shots a team takes, sorted by game scores “tied” “up1” and “up2” or whatever.

Look at the correlation between puck-possession for teams up by 2 goals and tied games. All but five teams in the past full five seasons give up more possession when ahead than behind:

Those that aren’t by the way, are the 2008 Edmonton Oilers, 2008 Calgary Flames, 2011 Edmonton Oilers, 2012 Montreal Canadiens and 2009 Toronto Maple Leafs. That’s a 2:1 ratio of lottery teams to playoff teams, and I’m guessing the discrepancy exists because of sample size.

Advertisement - Continue Reading Below

So I am willing to say that the Canucks’ problems with the lead this year are a sampling issue. There simply isn’t enough evidence to indicate that past Canuck teams consistently play with less possession when expected when up by two goals as opposed to tied.

I subtracted “up 2” score from “tied” score for every team between 2008 and 2012, and looked at each team who was outside one standard deviation from the median, which was a net loss of 6.22 percentage points for the changed game state:

Year Team Tied Score Up 2 Score Difference
2010 Columbus 48.67 33.44 -15.23
2010 Edmonton 44.61 29.61 -15.00
2008 Vancouver 48.41 34.67 -13.74
2010 Atlanta 49.30 36.55 -12.75
2009 Carolina 54.44 41.80 -12.64
2011 Pittsburgh 54.01 41.73 -12.28
2012 Florida 51.24 39.18 -12.06
2011 Philadelphia 51.61 40.09 -11.52
2012 Boston 52.93 41.76 -11.17
2009 San Jose 56.63 45.54 -11.09
2010 Carolina 48.95 37.89 -11.06
2011 Nashville 52.06 41.13 -10.93
2009 Nashville 48.78 37.85 -10.93
2012 Minnesota 44.54 33.67 -10.87
2008 Buffalo 50.94 40.64 -10.30
2011 New Jersey 53.53 43.48 -10.05
2011 Buffalo 51.65 41.69 -9.96
2012 Tampa Bay 47.89 37.93 -9.96
2009 Edmonton 46.80 36.84 -9.96
2012 Washington 51.26 41.36 -9.90
2011 Anaheim 46.12 36.27 -9.85
2012 Calgary 47.15 37.47 -9.68

The Canucks appear on the list just once, back in 2008. But to see if there was any year-to-year repetition to the extent teams sat on leads, I looked at the 16 teams pre-2012 and compared them to the next season’s results (I didn’t do this with 2012 teams because 2013 isn’t a full season of data) to see if they correlated at all:

Advertisement - Continue Reading Below

Honestly I don’t see enough evidence that this is repeatable, if there’s any.

The Canucks issues with the lead lately stem from sample size. They aren’t one of the teams that regularly get outperformed when “up 2” compared to “tied”. I think this is an issue of people who don’t like Vigneault’s coaching style trying to pick apart things that he does wrong. Sure, the Canucks sit on leads and that’s a problem, but everybody does.

In the macro sense, this isn’t something the Canucks do differently than anybody else.

Advertisement - Continue Reading Below

  • Fred-65

    Too many fans blame coaches and their style of play.

    To me, its pretty obvious that officials favour the trailing team to keep games close/competitive/exciting.

    Done any work on the timing of penalty calls?

    Nice work btw

  • Fred-65

    AV is a great Coach.He is intelligent,works his rear off, is meticulous and focused ,AND, backs his players publicly.Getting rid of Alain would be a mistake.A BIG ONE. I like him, and isn’t that what counts?

    • I think it has more to do with the team that’s trailing taking more risks, and generally risk-taking is the right thing to do in sports. There’s some exceptions, but in hockey, skating the puck in is more effective than dumping the puck in, but late in the game it’s just a mentality to make the “safe” play if you’re ahead and the “risky” play if you’re behind.

  • I’m a fan of Alain Vigneault and I fully agree that people don’t give enough credit to score effects (and instead credit/fault the team). Having said that, based on those behindthenet numbers, doesn’t it seem like the Canucks are a little worse with the lead than they should be?

    Unless I’m mistaken, here are the Canucks’ ranks in “up 2” and “tied” scenarios for each of those seasons:

    Up 2: 10th
    Tied: 7th

    Up 2: 15th
    Tied: 4th

    Up 2: 17th
    Tied: 16th

    Up 2: 25th
    Tied: 17th

    Up 2: 30th
    Tied: 23rd

    That seems like a bit of a trend to me. I also glanced at “close” and “up 1” scenarios and they all seemed to support that trend as well with the exception of the 2008-2009 “up 1” scenario (where the Canucks were especially good).

    Thoughts anyone?

  • Each of those seasons is within one standard deviation but is that still true when considering all 5 seasons (not to mention similar results so far this season)?

    Not trying to make you crunch more numbers; those results just surprise me. Maybe it’s just surprising because I underestimate the variability as you suggested.

  • Put it this way: Assume you have a half chance of out-performing your tied score and half a chance of being out-performed.

    Five coin flips, essentially. Well… landing “tails” five times in a row is a 3.125% chance. Multiply that by 30 teams, and there’s a chance it should happen to at least one team.

    Dice and coins have no memory.

  • An article from Travis Yost from yesterday. Maybe human nature can be manipulated by coaching?

    What Teams Drive Possession Ahead/Behind? by Travis Yost

    “There’s something to be said about the collection of teams near the top {of this ranking} who continue to drive possession while leading. I don’t think the players are inherently less risk averse with those clubs, but I do think a decent portion of the praise can be allocated to the coaching staffs who don’t let their units dial it down while ahead.”
    Vancouver is not near the top….they are number 16, and1% above an “average” possession team while leading. The Kings rank #1 and they are about 8-9% above the average.

    “Vancouver’s a weird team. Why is this team so bad ahead, and so great from behind? This probably requires some in-depth analysis into player usage and game strategy; my guess is a team that once traded Cody Hodgson (et al.) for Zack Kassian (et al.) kind of gets away from their bread and butter. I’d like to see what their numbers look like after forty-eight games — will they normalize a bit, or is this really what Alain Vigneault wants?”
    Vancouver is #1 on the list for possession while trailing, and 5% above an average possession team.