What We Learned: Tuukka Rask and the Bruins' workload problem

Toward the end of the Boston Bruins’ doomed campaign last season, people seemed to be harping on Tuukka Rask's workload with some frequency.

Specifically, they felt it to have been far too big. He played 70 games, 67 of which were starts, and it was the largest number of his career by a good 20-plus percent. Indeed, in the last two seasons, he nearly doubled the total number of games played he'd racked up in his career, and was in fact busier than when he appeared in 36 of Boston's 48 games in the lockout-shortened season of 2013.

What's interesting about this, though, is the fact that despite the cries for a more reasonable handling of Rask's office hours, there simply isn't a lot of evidence to say that his playing 70 games was the reason his game appeared to take a step back.

And indeed, let's consider for a moment the monumental ask laid at Rask's feet. Not the 70 games, but the fact that people were complaining to some extent about a goaltender “only” carrying a .922 save percentage over 70 games. On more than 2,000 shots last season, Rask allowed just 156 goals, which is just an outrageously strong number, and somehow didn't even warrant a single Vezina vote(???).

Another issue here, of course, is that Rask lost nine games in a shootout, which is a very high number.

It is widely acknowledged that shootouts are effectively coin flips over which no player wields a significant amount of influence, but here's the really crazy part: Those nine shootout losses came despite the fact that he carried a save percentage in shootouts of .755, well above the league average of 698. He carried more than his share of the water in the skills competition, and his teammates decidedly did not.

If the Bruins' shooters improve that number even slightly, Rask probably has at least two or three more wins on the season, the team makes the playoffs, and we're not having a discussion about, “Did playing 70 games hurt Rask's effectiveness?” How he would have done in the playoffs is anyone's guess at that point, but we have plenty of evidence to suggest that overworked goalies tend to see their wheels come off in the postseason. (I would argue that it's perfectly okay to throw out data from playoff losses, particularly early ones, because you're playing so few games against such good teams that the water is probably going to be pretty muddy.)

And further, that lack of goal-scoring prowess is what led the Bruins to play so many one-goal games in the first place, which is obviously what necessitated Rask playing so often in the first place. I leave it to you to determine whether the team actually did anything this summer to go out and address that rather grave concern.

In addition to all this, the same is probably true if the Bruins' backups last season hadn't been so middling, or at least perceived as such. Niklas Svedberg appeared in 18 games last year, four of them in relief of Rask. He had a .918 save percentage which, while not as good as Rask's, is pretty damn credible and probably indicates he could have been relied upon a little more heavily and still delivered results.

Yet conventional wisdom obviously states that you don't want goalies playing 70-plus games in a given season. That's a lot of work not only physically but mentally, if you choose to believe in that sort of thing, and certainly doesn't seem conducive to all that much success. Last season, Rask became the 17th goaltender since 2005-06 (what I'd consider “modern” hockey in all its aspects) to make at least 70 appearances, but it's a feat that has been accomplished 33 times in that 10-season span. That obviously doesn't include those who worked the equivalent of a 70-game season in 2013 (about 40 appearances in 48 games), which nine guys accomplished.

But the thing about that conventional wisdom of giving your goalie more than the very occasional night off is that there doesn't seem to be any real statistical correlation between the number of minutes a guy plays of his team's total number available, and the percentage of pucks he stops.

I ran the numbers several different ways — guys making 40-plus appearances, guys making 60-plus appearances, shots against per minute for guys with more than those numbers of appearances, all career numbers of guys making at least that many appearances, etc. — and couldn't find any kind of even remotely significant r-squared correlation, to the point that from a mathematical standpoint it might not even be worth discussing.

These numbers can, admittedly, get a little weird. Obviously great goalies are going to play more often even than good ones, and it's rare to see anyone you wouldn't consider elite clear 60 games, for example. But even if you look at those guys individually, things get a little goofy. I looked at everyone who played 60 games at least once in the salary cap era — there were 41 such goaltenders, with a combined body of work of more than 300 seasons between them — then looked at every season in which each of those guys played 30 or more games and compared that with the total number of team games available. And once again, there was basically no correlation on a collective basis, an r^2 of .0543, or almost non-existent. But it is positive, which is interesting.

(The explanation here is a little convoluted, I know. Blame the seasons shortened by lockouts.)

That's not to say that some guys don't see their numbers slip as the amount of work they do goes up, because some can clearly run out of steam. Among this group, the average correlation (.09509) is about 75 percent higher than the collective number seen above, which further shows positive correlation between the amount of games played and save percentage. Again, that tends to skew in favor of elite goalies with heavy workloads — and usually, longer careers — which is why it's important to keep that the median in mind here as well.

It is, again, very different for everyone. For example, the more ice time a guy like Craig Anderson gets (and he has only four seasons in which he appeared in at least 60 percent of his team's games), the worse he tends to be, relatively. Others, like Martin Brodeur, seem to thrive under the extra work. Brodeur played 70-plus games 12 times in his career, and his best season ever was 2006-07, in which he went .922 despite playing 78(!!!!!) times.

But as with the overall correlations, many more guys tend to see little connection between their workload and their save percentages. Miikka Kiprusoff is one of the more famous cases of a guy going 70-plus games with regularity — he did it seven straight seasons — but his numbers fluctuated wildly in that stretch. These numbers, by the way, work out to a minor negative correlation. The difference in work between these seasons, though, was usually just a handful of starts.

All this may simply point to an area where “eye test” and “numbers” diverge; longstanding hockey wisdom is that you simply don't play guys that much, and yet it doesn't seem to have much impact on performance, at least in the regular season. We have to keep in mind here that even a handful of goals allowed can have a decent-sized impact on season-long save percentage, and so if a guy allows a goal that appears to be one he stops when “well-rested,” would probably show up over and over if he's that worn out by the 15-20 extra games being asked of him. But then again, you might just subscribe to the theory that goalies are inscrutable (see the above Kiprusoff numbers) and don't play by many mathematical rules from one season to the next.

The thing with Rask is that we don't know one way or the other into which of these three goaltending categories he falls.

Does he suffer with more work, or was his “meager” .922 save percentage just a blip? We just can't be sure yet. And regardless, if his bottom-of-the-barrel seasons are going to be .920-plus, does it really matter that much? If you need your goalie to turn in .930 seasons for you to make the playoffs — and the Bruins just might this season — then you have much bigger problems than how often you're trotting out your No. 1.

The trick for Julien, I think, is to figure out what he has in the goaltending trio of Malcolm Subban, Zane McIntyre or Jeremy Smith actually gives him to work with. One would assume Subban will be the backup in Boston while the other two bus it in the AHL, but one also never knows with this team. Subban's body of work in the AHL — .921 over two seasons — speaks for itself at this point. But even still, Julien has to actually trust him more than he did Svedberg last year if all the hand-wringing over Rask's playing time is to be avoided.

Maybe it doesn't matter at the end of the day because Rask can handle the amount of work. But then again, maybe the Bruins should try scoring more goals so he doesn't have to find out one way or the other whether it matters.

What We Learned

Anaheim Ducks: This is a team that looks very much like it could be the best in the West this season. Some issues with the roster, sure, but that's true for basically everyone in the league (except maybe Tampa?).

Arizona Coyotes: Arizonans can now buy Coyotes license plates. Dibs on “MVNSOON.”

Boston Bruins: Do people not think Ryan Spooner is an everyday NHLer at this point? That's pretty clearly what he is.

Buffalo Sabres: Yo so why hasn't Cody Franson signed that two-year deal yet?

Calgary Flames: And now, some bad news for the Flames.

Carolina Hurricanes: That's a nice little two-year extension for Elias Lindholm, and now the team has to confront the realities of the Eric Staal situation.

Chicago: At this point is there anyone within an hour of Chicago who hasn't had his or her picture taken with the Stanley Cup?

Colorado Avalanche: The Avs are not alone in not being able to bring the NCAA players they've drafted to rookie camp in a few weeks. Some NCAA players aren't happy about it, but oh well buddy.

Columbus Blue Jackets: Periodic reminder that Columbus is paying David Clarkson a ton of money, and they're doing so voluntarily.

Dallas Stars: Please don't try to defend the Cody Eakin contract.

Detroit Red Wings: The new Wings rink is moving right along as planned. It's still gross! “Olympia has a stated goal of 30 percent Detroit business participation and 51 percent Detroit resident employment in construction of the arena.” Wow so generous to set those goals in exchange for hundreds of millions in public money.

Edmonton Oilers: Ahhhhhhhhh. Ahhhhhhhhhhhhhh. Ahhhhhhhhhhhhhhhhhhhhh.

Florida Panthers: This Florida Panthers roster from the 1993 expansion draft is hilarious. Especially because they were getting castoffs from only 24 teams, not the 30 we'd see today. They got John Vanbiesbrouck and not a whole lot else.

Los Angeles Kings: Russian surnames can sometimes be a little tough to pronounce, even for those who know you're supposed to put the stress on the second syllable most of the time. This new Kings signing, though, is particularly baffling.

Minnesota Wild: “It would be nice not to have to go 27-9 again there at the end,” says Devan Dubnyk. No kidding hey?

Montreal Canadiens: Carey Price is only one point better than Jonathan Quick. NHL16 sounds like another great game mailed in by the folks at EA Sports.

Nashville Predators: Wow, did you even remotely remember the Preds signed Barret Jackman?

New Jersey Devils: The operative word here? “Try.”

New York Islanders: How defensemen like Travis Hamonic are handled by their coaches is one of the most fascinating things about hockey to me.

New York Rangers: One has to wonder just how many pushes the Rangers can logically take before it all slips away from them. This might be their last, best effort before Lundqvist's skills start to really diminish.

Ottawa Senators: Imagine you're a Senators fan who thinks someone other than Erik Karlsson is the most important player on the team?

Philadelphia Flyers: Samuel Morin seems intent on cracking the big club's roster this year. No one tell him they have six guys on one-ways already.

Pittsburgh Penguins: The Pens already gave Daniel Sprong a contract. They seem real excited about him.

San Jose Sharks: Oh come on.

St. Louis Blues: This is actually pretty not-bad. But then again, if Scott Gomez actually makes your team...

Tampa Bay Lightning: This is just about the only kind of hit you'd want to see associated with Stamkos.

Toronto Maple Leafs: Don't do it, Kyle! Think of your cred!

Vancouver Canucks: Jim Benning read this with great fascination, then bookmarked it. “So that's how you pronounce Hamhuis?” (And oh my god can they invite Mike Gillis to this?)

Washington Capitals: The Capitals’ potential starting lineup this season.

Winnipeg Jets: This would be less interesting than the biography of an actual, paid mascot.

Gold Star Award

Wayne Gretzky was hanging out with Donald Trump this weekend. Yooge and luxurious.

Minus of the Weekend

This week in “Please Like My Sport.”

Perfect HFBoards Trade Proposal of the Week
User “WhiteLight” is bringing some white heat.

To Ottawa:

Jeff Skinner

To Carolina:
Jared Cowen

Signoff
Y'know what we have for dinner in my house? Nothing! Sometimes chicken.

Ryan Lambert is a Puck Daddy columnist. His email is here and his Twitter is here.

(All stats via War On Ice unless otherwise noted.)