The Sedins are getting "old." They've both just reached their 34th birthdays. On the same day, no less. Because they're twins. Who play in the NHL together. Still a crazy thing.
We don't talk enough about the complete game of either of the Sedins or their individual tendencies. We should. Regardless, the sole narrative that persisted throughout the off-season had to do with their decline in offensive production and whether, at their advanced age, they are capable of bouncing back from their lowest point totals in nearly a decade. At the risk of beating the topic to death, I want to focus this article on, hopefully, dissecting that issue a bit better.
The Age-Related Decline
Don't worry too much about the chart below. Just take a look and read on. We'll come back to it.
Several studies have examined, in varying degrees, how aging affects offensive production. The debate typically revolves around three generalized issues: (1) when a player reaches their peak of offensive productivity; (2) how to measure a player's declining production from that peak as they age; and (3) whether there exists a safe baseline expectation for that decline.
Any study in population trends over time requires a sample selection. The sample selection that we would be studying when discussing the three issues above should be all NHL skaters playing from their 20s through their mid to late 30s.
However, there is a problem at the outset of the study. The sample selection becomes increasingly more suspect as the players age. This is because those players who reach the peak of their personal performance, but only perform at a league-average or below level, typically fall out of the league before their decline can be fully measured. So by the end, we are only looking at players who are good enough that even in their declined state, they can perform at or above league average and, therefore, stay on an NHL roster.
The declining veterans who performed at or below league average last season are often the types of players we see on tryout contracts hoping to stick around in a league testing out its new shiny prospects at training camp. Quite simply, as an average player declines, he becomes below average and utterly replaceable. And such a player is inevitably replaced by a newer, younger, and often cheaper (or more contract controllable) model. So we rarely get to see what 60% of the declined average or below average player's production ever looks like, or when it's most likely to occur.
The result is that we only get to see and measure the truly exceptional players playing into the bottom of their decline. This changing sample makes it difficult to say whether the above-average player declines at the same rate as any old average or below-average player. That's where the variables do crop up. Is part of what made some of them great their exceptional conditioning and athleticism? Do elite players who rely on "tools and skills," such as stickhandling and awareness, decline at the same rate as elite speedsters, or elite shooters? Could those who relied on conditioning in fact defy conventional logic and decline slower than the "naturally gifted" because natural gifts decline as much as physical ones? The questions are endless. These alone could spawn 30 different other studies and articles.
Luckily for us, we at least know that the Sedins are above-average offensive producers. However they might be defined. So we might at least be able to start here by creating a baseline measure for them against comparable age-related declines of above-average producing players who play into their mid to late thirties.
I think the best method I've found for factoring age into a discussion of performance, and projecting performance, is loosely expressed in this well-written piece by Eric Tulsky. He then ever-so-usefully follows it up with this specific look at how NHL scoring rates change with age. I personally think you should read them both. But if you just can't stomach the stat-tastic goodness, definitely check the second one out.
It's ok. I'll wait.
::hums "Hooked on a Feeling"::
Oh, you're back! Welcome. Let's move forward.
If you already believed the generalized theory that star-boy players just kind of start declining precipitously around 30 years of age, or falling off the proverbial "cliff," you now have a bit of a lifeline. Tulsky presents what I believe to be a sound, statistically-backed theory regarding a sample of players', mainly good and great ones', declining offensive production. In light of this we can accept that, as Tulsky puts it, "on average, players retain about 90% of their scoring through age 29, but the drop from there is pretty sharp." Now that you've had a primer, let's revisit the handy visual reference from earlier:
So our expectation is that above-average players will drop to about 80% of their peak by around 30-31, 70% around 32-33, 60% around 34-35, and so on. And yes, as players get into their 30s, their production will eventually decline at some point, regardless of whether they follow the above curve or not.
But I am not convinced that, with where we are at in nutrition and training, the greatest athletes have to start their decline at any particular magical age. It's simply a substantial trend worth noting and considering, while accepting the sample selection and variable issues that Mr. Tulsky and I have detailed.
Henrik and Daniel Have Been in Decline for Some Time Now
The Sedins have an entire career behind them. A really good one. So sure...maybe it's just their time. But the narrative sure seems to be that they fell off a cliff last year. Due to age. Due to some of the dump and chase aggressive 2-1-2 forecheck relentlessly employed by former coach John Tortorella. Due to tougher zone starts. Due to a combination of those things and the extra workload they took on in total TOI and PK work last season.
The above reveals that at even strength (5v5 only), and removing variables such as power play time, the Sedins' points per 60 minutes rate has been in decline since their aged 29 season. Their peak came a bit later than other players playing into their mid-late thirties. It's also noteworthy how closely their fates intertwine (we will get to that in a bit). The problem is, the decline from that late peak has been steeper, and resulted in much less of a percentage of their peak, than their aging contemporaries. Take a look:
A few things are revealed here. First, they peaked at 29, which is relatively late. The decline in the season after their peak, when compared with their performance in the seasons just prior to their peak, tells me that their career highs in that year set an unreasonable expectation for production. If box-plotted, their 2009-10 season is outside the interquartile range and nothing before or after it really comes all that close.
Both Sedins benefited from insane percentages that season. In sh%, PDO, everything. Even for them, and they are uber play-drivers regardless. So the season following their peak is a bit of course correction for their late and extremely high peak in 2009-10. Expecting at or near 100 points per season was never really that reasonable, even though they did it that season. The point per game pace they had before and after was much more sustainable and achievable year over year.
After their age 29 season, Henrik declines at relatively normal percentage rates, as compared with contemporaries through age 31 and 32 (lockout season). Daniel, however, falls all the way from 80% of his peak to just 58% of his peak and stays about there through the lockout season.
Henrik experienced a significant 10% drop in production last season from the lockout shortened season prior. But Daniel drops a whopping 18% in production last year. No bueno. What the hell happened?
While Henrik has shown an ability to produce on his own as a result of his exceptional setup and playmaking ability, he is also obviously very tied to his linemates' success at scoring goals. This is because he nets a vast majority of his points from assists and you can't have one without the other. In 2013-14, the Canucks scored at just 7.93% while Henrik was on the ice at even strength. This represents a significant drop-off from the 10.16% sh% the team averaged while Henrik was on the ice in the 6 prior seasons.
A big reason that both Sedins have experienced dropoffs is that Daniel's shooting percentage has nose-dived. Especially in the high probability scoring areas around the front of the net. Areas where, in the past, he has made a living walking out of the corners or the half-walls for one-timers and beautiful snipes (usually setup by his brother).
As you can see, Daniel's S/60 hasn't really declined too much prior to last season, but his ability to convert his chances has. Couple the over 2% drop in sh% from his last full season in 2011-12 (not counting the lockout because SSS), with the .5 drop in s/60 and that shaved 4 even strength goals off his total, despite 60 more minutes of EV TOI.
Here's what his shooting looks like from 2008-2013:
Now take a look at his last full season prior to 2013-14:
And last season:
Any questions? Daniel simply wasn't scoring at the same rate from his bread and butter areas around the crease and low slot. He also took fewer shots from just under the hashmarks within the faceoff circle. That was previously his preferred locale for curl-outs, give-and-gos and one-timers.
If you want a straight answer, I don't know how the Sedins go about rectifying the problem of Daniel's sh% other than systems adjustment (discussed below). But at least some of it is bad luck, as is reflected in his PDO.
For the uninformed, PDO is a measure of the team's sh% + sv% while the player is on the ice. It can be represented as a percentage or number, but it typically regresses to around 1 (or 100%). Ignore the name, it stands for nothing. But the premise is relatively basic. If those two numbers are higher or lower than they usually are over a given sample size, they are likely due for some kind of regression back to their mean average of 1, or 100%. Thus, if in a single season the PDO is higher/lower than normal (the team was scoring more/less while player x is on the ice and saving more/less shots while player x is on the ice) it's probably the result of some good/bad luck, because player x can't control factors of play independent from him and the team is performing at a level that is above/below normal averages when he is on the ice.
Just as Daniel's 2009-10 PDO (105.9%) was higher than the surrounding average PDO (102.1%), Daniel's 2013-14 PDO was 99.9%, his lowest since 2007-08. If the Sedin's have averaged performance a few percentage points above 100% PDO, it's reasonable to expect some bounceback in this department.
The addition of Vrbata, or another competent and performing winger opposite Daniel, should help alleviate some of the volume shooting burden and allow for greater opportunities to pick spots, receive uncontested passes and shoot from uncontested lanes. This alone is likely to raise his sh%. Vrbata's propensity for face-up volume shooting should also be a recipe for rebound goals, another sh% booster.
The Tortorella Effect
For starters, much conjecture has been made about the fact that under Torts, the Sedin's total TOI went up, as did their time spent on the PK. I have no desire to add or detract from any of that because, quite simply, I don't know how fatigued they were or were not. And I don't care to pretend to know.
Beyond that, at even strength, John Tortorella imposes a zone defense that relies almost entirely on loose puck battles won, much more dump or chip and chase zone entry, and hard 2-1-2 forecheck duties. None of these things, in a vacuum, are all that conducive to offensive production. Last season was no exception, and the Sedin line was not immune to his systems by any stretch.
This may have been most evident in Henrik's substandard even strength defensive zone exit numbers (h/t @shutdownline). Henrik only managed controlled possession 33.1% of the time when exiting the zone, leaving Daniel (46.3%), Alex Burrows (45.8%), Jannik Hansen (42.1%) and Ryan Kesler (35.1%) to do the heavy lifting when they played with him. He was 4th amongst Canucks centers with 150 or more touches in this department, revealing that he may have struggled with gaining puck control coming out of the zone defense and/or winning individual loose puck battles.
He also seemed to struggle in finding support lanes for puck movement. He was just 3rd amongst Canucks centers in total passes out of the zone, despite his TOI and low carry out rate. Though he didn't often turn the puck over (3.5%, lowest on the team), this all resulted in only a 34.9% exit rate when the puck touched his stick in the D zone. I can tell you, this was not very Sedinery. It was 11th on the team amongst forwards with 150 or more touches.
The Sedins also saw less offensive zone starts under Torts than in previous seasons, at about 60 Off. ZS%. Compare this to the 63% for Henrik and 66% for Daniel in the lockout season, and the near 80% for each in 2011-12, their last full season. Combined with the struggles that Henrik had in acclimating to his center role in the breakout coming out of Tortorella's "defensive shell" and you have a recipe for reducing offensive success down ice on nearly every shift where the puck ends up in the defensive zone.
Because Tortorella typically employed a hard dump and chase with an aggressive 2-1-2 forecheck, the team's carry-in rate on offensive zone entries suffered last season. At just 44%, it was lower than the 48% that teams facing Vancouver maintained, just for comparison purposes. The difference can certainly be chalked up to system style and depth personnel. Given this, Henrik (58%) and Daniel (55%) still carried in at a rate better than every teammate with more than 100 zone entries, other than Santorelli (60%) and Kesler (60%). But their shots per carry-in (Henrik .64, Daniel .68) were lower than what you would expect from skilled players.
Nobody on the team benefited from the hard dump and chase entries, as team shot rates were just .30 per dump-in. Nobody. Except the other team.
Henrik (.31 per dump) and Daniel (.32 per dump) didn't fair any better than their teammates, in spite of their superior overall possession metrics. And, given the percentages above, there were a lot of these self-defeating attempts to gain the zone leading to offense-crushing loose puck battles behind the goal line. Nothing more.
The problem is, the Sedins also really struggled with the entries where they had to chip and chase post-possessed-entry and then employ Tortorella's aggro-low attack. It's just not their style. They would rather initiate a cycle from a tempered and passive attack. The aggro approach simply negates continuing to carry clean possession to the half-boards for that cycle, even when it works to keep the puck in the zone.
And when the aggro stuff doesn't work to keep the puck in the zone?
The style relies on a banger or speedster (or both) to get low, and for the third forward in to not get sucked down to areas where a missed opportunity at keeping the opposition pinned means a rush going the other way. Daniel isn't exactly the fastest or most physical player for such a forechecking role. Not having a competently performing linemate for large parts of the season only exacerbated the problem.
Everything indicates that Willie Desjardins intends to reinstate man-to-man defense, increase the twin's OZS%, and utilize a more tempered and possession-based approach to zone entry. The Sedins should immediately benefit from all of this. To what degree is uncertain. But logic and history tell us that they will likely be able to exit their zone more often and more cleanly and see improvements in the number of shots produced from controlled entries and rush shots, which typically carry a higher overall sh%. Because these are systemic changes, they should help recover some of their lost production from last season, despite any further age-related decline.
As the twins have declined, so has Alex Burrows, who is coming off a particularly rough and injury riddled 2013-14 campaign in which he only played 49 games and managed a paltry 15 points. While Burrows possession peripherals were just fine (56.5 CF%), he struggled mightily with his scoring. His 5v5 sh% was an abomination, at a mere 3.8%. This coming from a guy who's 6 year average up until last season was 12.64%, showing he was capable of sustained high sh%.
It's hard to go too crazy over the insane cliff-jump there, because it was really a lost season for him and the team. But this, along with the lower than average sh% of linemates on the ice with Henrik and Daniel, certainly contributed to a decrease in production. New acquisition and notable volume shooter Radim Vrbata should help cure this problem, provided the Sedin's find chemistry with his face-up shooting approach and work in the high slot. However, if the Sedins truly want to return to form, I believe they benefit most from a player who can play in front of the net with size and physicality, providing a distraction for the Sedins to work from the corners and half-boards. If Burrows returns to form this season, I would not at all be surprised to see him on a wing with the Sedins again. I think it's the best fit, from a traditional standpoint.
The Power Play
I'm not going to get into the hows and whys of what was a forgettable season with the man advantage. A lot of it was simply inexplicable setups designed to create low percentage chances, and the wrong personnel on the ice to shoot those chances. And, quite frankly, we are going on 3000 words here.
But it is a major area where the Sedins lost overall production, and can likely gain it back. Not only did Vancouver do an inferior job of drawing penalties (as many Tortorella teams do because of the systems described above), but the Twins overall production went down. I just want to highlight the degree of that, visually:
I left out the lockout season because it was shortened and strange. But even if I extended out the lockout season numbers to an 82 game rate, Daniel and Henrik would have put up 21 and 22 points respectively. I'm not telling any Canucks fan anything they don't know. And both Sedins did miss games, which contributed to the reduction. But a good power play, and namely a right-handed trigger man, are vital to their overall offensive production. Always have been, and necessarily will be if there is to be a clear return to form in overall numbers.
And of Corsi (puns!), their possession numbers haven't really gone anywhere.
When they are on the ice, the Sedins are still driving play at a world-beating level. Daniel at 57% and Henrik at 55.7% were both head and shoulders above all but the very best, and, along with Burrows, remained head and shoulders above their forward teammates in CRel. They are soundly out-attempting their competition and raising the CF% of every teammate who plays with them.
It's more so how play was being driven within the system, and the percentages related to that, which took a big hit. And, quite frankly, both were banged up last season, which didn't help them any.
If they can get back to "normal," they can get production out of their third-wheel (currently Vrbata), and the Canucks can muster a more productive power play, even in consideration of the natural decline with age, there is no reason to believe that they can't easily surpass their 2013-14 overall production totals, and likely come relatively close to their 2011-12 numbers. Takers on 70+ points anyone? I don't see why not.
[Statistics compiled from the following sources, with many thanks:]