2016 Stanley Cup Final: Analytics predict winner
When Pittsburgh Penguins Captain Sidney Crosby and his San Jose Sharks counterpart Joe Pavelski were at home wondering what Santa would bring them last Christmas neither had much cause for optimism for the remainder of the hockey season.
Both players’ teams were on a virtually identical and seemingly unalterable path to yet another disappointing season. Crosby’s Penguins were in particular trouble. After cutting ties with coach Mike Johnston a couple of weeks before, the team dropped its next four games under Mike Sullivan before finally winning one against the struggling Columbus Blue Jackets. The Pens were still tinkering with a stacked top six and apparently ignoring our advice from six weeks earlier that Phil Kessel should be dropped to a “third line” with Nick Bonino.
Sitting 20th in the overall standings and out of playoff contention, Pittsburgh looked like a rudderless ship facing increasingly longer odds of even making the postseason, let alone doing anything once it got there.
Pavelski’s 18th-place Sharks had slightly more cause for optimism. Not that they were having a better season (they had one more point than the Pens but had also played one more game), but at least they were in the NHL’s weakest division. Somehow, despite their meager 17-15-2 record, the Sharks managed to still find themselves in second place in the horrifically bad Pacific.
While no one expected Anaheim to remain second to last overall, as long as San Jose could keep ahead of league punching bags Calgary, Edmonton, Vancouver and Arizona, they were assured another spot in the playoffs, at which point hockey writers could change the date on any number of “Joe Thornton is a choker” pieces from the past decade.
So here we are at the end of May with two teams that ought to have been golfing by now, and instead, one of them will be hosting a victory parade soon enough. Barring an extraterrestrial invasion, a massive asteroid hitting the planet, or some other catastrophe, either Thornton or Kessel is about to shed the “he can’t win” label.
Penguins vs. Sharks
Prediction: Penguins (58.1%)
If you’ve watched the Sharks pummel their opponents thus far and consider the fact that to get to this point they went through tougher teams than the ones the Penguins beat, you might think that 58.1% is a typo. We checked it five times and it’s not.
One thing that’s at work in our model is the fact that it looks at each of the eight possible outcomes for this series. As it turns out, most of the work in differentiating these two teams comes from the fact that a short series (i.e. four or five games) heavily favors Pittsburgh; whereas, if the series goes longer, the odds tighten up considerably. The two most likely outcomes are Penguins in seven (21.0%) and Sharks in seven (17.2%).
The other thing to keep in mind is our model is based on regular season performance. So if you think the Sharks are a different and improved team in the playoffs than they were in the previous 82 games, that might change things as well.
Since we’ve now seen each of these teams through 18 playoff games, let’s set aside the regular season stats and take into account that pretty decent sample of how they’re playing right now. As the table below shows, these teams have put up very close numbers to get to this point.
|Team||Fenwick For %||High Danger Scoring Chance +/-||Sv%||Sh%||PDO||PP%||PK%|
The Penguins have generated more unblocked shot attempts and have been a slightly more dangerous team than San Jose during 5-on-5 play, and while the Sharks have gotten strong goaltending from Martin Jones, they’ve steamrolled their way to this point on the basis of a high shooting percentage that may very well be unsustainable. This final makes for an interesting matchup in large part because San Jose and Pittsburgh both play a high-octane offensive game.
There’s another nuance here, however, and that has to do with which players are generating each team’s offense.
When we dreamed up the idea of pairing Kessel with Bonino back in November (and before that, the idea of the Penguins acquiring Kessel at last year’s trade deadline), the justification for doing so was simple: Pittsburgh’s offense lacked depth, and any capable team had pretty much cracked the code for beating them, namely slow down Crosby and Evgeni Malkin, thereby neutering their top four wingers and forcing a terrible group of bottom six forwards to step into a scoring role.
Coach Sullivan stumbled into one of hockey’s best lines—Hagelin, Bonino and Kessel—when Malkin went down with an injury. Sullivan discovered what we already knew—by having three lines that are a major scoring threat, the Penguins would be a far more difficult team to defend against.
As the table below shows, the strategy has worked. All of Pittsburgh’s top 12 forwards have been able to generate big scoring chances, and the offense has been spread around. When he’s not on the power play, Crosby isn’t getting on the score sheet much, but is there any question that the defensive effort required to shut him down has opened things up for the rest of the Penguins forwards?
Now take a look at the same chart for the Sharks …
San Jose's top six forwards (Hertl, Thornton, Pavelski, Couture, Marleau and Donskoi) are undeniably fearsome, but among their bottom six only Chris Tierney and Joel Ward have been at all productive. While it’s true that the Sharks' offense is unique in that one defenseman is enormously productive (Brent Burns has generated 2.06 points per 60 minutes), this attack looks a lot like the Penguins of yore in that it is very top heavy.
Shutting down those top two lines is no small feat, but if the Penguins have paid any attention to the bitter medicine pretty much all of their playoff opponents have dealt them over the past several seasons, they will do just that. The result will be a war of attrition, but they are still likely to win on depth of forwards.
Defense is a different story.
Whereas the Penguins have basically gone with Kris Letang and whatever else they can find, the Sharks have a significant advantage on the blue line. Burns is basically an extra forward and has contributed immensely on the offensive side of the ledger, but Marc-Edouard Vlasic has quietly put up a Conn Smythe MVP worthy playoff without getting much credit for it.
As the table below shows, Vlasic and defense partner Justin Braun have dominated their opponents to date in terms of both puck possession and shot quality.
|Player||Fenwick For % (score adjusted)||HD Scoring Chance +/-||Points/60||Even strength TOI/Game|
If it wasn’t obvious from watching the games, the table above makes abundantly clear that the Sharks are heavily reliant on their top four (and Vlasic in particular). At the same time, the bottom pair of Roman Polak and Brenden Dillon, although basically placeholders, log significant ice time each game.
One key for the Penguins in this series will be to try to get out one of their three very dangerous scoring lines in situations where they can get through the neutral zone with speed and challenge that bottom pair. Because it’s so difficult to find sheltered minutes against Pittsburgh, San Jose coach Peter DeBoer may lose the luxury of resting his top four guys and find himself forced to shorten the bench.
As for the Penguins, injuries to Olli Maatta and Trevor Daley have complicated matters, but looking at their current top six a couple of things become clear.
First, while their bottom pair isn’t the same glaring liability as the Sharks’ is, and they aren’t getting the same offensive production off the blue line as San Jose. As good as Letang is, he’s not putting up numbers that are anywhere close to those of Burns and has actually been less productive than Vlasic, who is generally known more for his elite shutdown abilities than his offensive prowess.
Second, the loss of Daley (18:09 of even strength ice time per game and high danger scoring chance +/- in the playoffs of +12) is likely to be felt as the final wears on.
|Player||Fenwick For % (score adjusted)||HD Scoring Chance +/-||Points/60||Even Strength TOI/Game|
Pittsburgh’s defenders have been surprisingly good so far, particularly in the second round series against the Capitals. And while that series included a one game suspension for Letang and the loss of Maatta for three games, it’s fair to wonder how the current depleted group will fare against a San Jose offense that is firing on all cylinders. The Sharks managed an insanely good 27.0% power play in the first three rounds, so it goes without saying that staying out of the penalty box will be key to preserving the legs of an already shaky defensive corps and Letang in particular.
We’re going to go out on a limb here and guess that DeBoer is going to be looking for opportunities to get his two big lines out when Letang isn’t on the ice.
Regardless of the result, this one has all the makings of a memorable final.
If you want to see Ian’s five-year-old daughter’s pick for the final, you can find it here.
The Department of Hockey Analytics employs advanced statistical methods and innovative approaches to better understand the game of hockey. Its three founders are Ian Cooper (@ian_doha), a lawyer, former player agent and Wharton Business School graduate; Dr. Phil Curry (@phil_doha), a professor of economics at the University of Waterloo; and IJay Palansky (@ijay_doha), a partner at the law firm of Armstrong Teasdale, former high-stakes professional poker player, and Harvard Law School graduate. Please visit us online at www.depthockeyanalytics.com
Dr. Mikal Skuterud (@mikalskaterud) is a professor of economics at the University of Waterloo.