clock menu more-arrow no yes

Filed under:

Keep rolling, rolling, rolling those lines: Part 1

New, comment

Julien's thrown out an old, bad strategy that doesn't work.

Bob DeChiara-USA TODAY Sports

Over the years, many have praised the Bruin's ability to "roll 4 lines". It's gotten to the point where the praise is just slightly ridiculous. Let's examine this Chris Kelly quote that came after a 2-1 win against Arizona in 2014:

"That's our strength, our depth and our ability to roll four lines," Chris Kelly said. "I think a perfect example was the last game where we were able to roll four lines the whole time and I don't know if Phoenix did." - per Metro West Daily News

Now, there's depth, there's wanting everyone on the roster to contribute, and there's what the Bruins used to do. In Claude Julien's old system, "rolling 4 lines" meant EVERY line played in EVERY situation. Down a goal, there's 5 minutes left in the third, and the other team's worst line has just iced the puck? Play Campbell! Up a goal with 5 minutes left, defensive zone face off, against the Malkin-Crosby line? Put out Seguin, and if he screws up we trade him! ALL SITUATIONS, GUISE.

But if you've been watching the Bruins over the last couple seasons, you may have noticed that Claude seems to have thrown that strategy straight out the window. It seemed to start with Torey Krug. For a while Krug and whoever his partner happened to be at the time were the only players allowed to start +40% of their shifts in the offensive zone. The next big change came from the addition of David Pastrnak and Ryan Spooner. In 2014-15, David Pastrnak started more shifts in the offensive zone than any other player with over 300 minutes. The Lucic-Spooner-Pastrnak combo just flat out didn't play good defense, and still managed to get ice time.

This wasn't just one line getting more offensive starts. There was an appreciable spread in the way lines were used. Despite the next day rants and photoshops from DoY, even the Bidet himself was subject to this trend. Per war-on-ice, in the lockout shortened 2012-13 season, Gregory Campbell started 5.65% less of his shifts in the offensive zone compared to the average Bruin. In 2013-14, it was 7.04% less. And last year? 11.45% less.

This stratification isn't limited to the Bruins. It's been a league wide phenomena that the Bruins are quickly catching up to. Per stats.hockeyanalysis.com, here's the offensive zone start standard deviation for the Bruins and the rest of the league over the past few seasons, minus the lockout shortened 2013 season.

Season LeageSigma BruinsSigma
1415 5.230821824 4.816063372
1314 4.898055676 3.943280218
1112 4.486226122 2.326564083
1011 4.168142687 2.405895537
0910 3.696180682 2.545667113

And a graph.

And here's the standard deviation for all teams in the 2014-15 season.

Team Zs% sigma
Nashville 10.43720908
Chicago 9.460579867
Edmonton 6.94572
Minnesota 6.378564278
NY Rangers 6.261334985
Detroit 6.110435684
NY Islanders 5.835143937
Arizona 5.765190738
Calgary 5.359375408
Montreal 5.318886477
Tampa Bay 4.833489324
Boston 4.816063372
Philadelphia 4.186599939
New Jersey 4.18386311
Florida 4.181750188
St. Louis 4.118475425
Dallas 4.012871606
Carolina 3.972824352
Washington 3.962382641
Toronto 3.844952262
Pittsburgh 3.557875235
Ottawa 3.484747222
San Jose 3.229043999
Buffalo 2.955387587
Vancouver 2.712920026
Los Angeles 2.55276694
Columbus 2.401843628
Winnipeg 2.298789537
Colorado 1.856704861
Anaheim 1.744304275

If you haven't gotten it yet, Julien is moving away from his previous strategy of "every situation" deployment for his lines. If you take out the heavy outliers of Nashville and Chicago, the Bruins were actually above average in player zone start standard deviation last season. This should interest Claude haters. If you think he's stubborn, well, here's evidence that he's thrown out an old strategy that was highly praised, but sucked. If you want Claude to start using more young, offensive minded players, then this deployment system gives them roles to fill. Spooner and Pasta may not have had the success they did last season if they hadn't started nearly 45% of their shifts in the offensive zone.

It's also worth noting that the standard deviation of all player's zone starts has been rising steadily as well. That's because, well, the strategy Claude's implementing works well. Chicago and Nashville are probably the two best examples of this system working wonders. To be fair, a part of it is the players you have at your disposal. You try having Marcus Kruger or Paul Gaustad on your team and not deploying them in every defensive zone start! But coaches and GMs are a team, and a GM will certainly prefer players who fit their coaches system.

In part two, we'll look at just how successful this strategy has been and can be.