embed from getty images
welcome to the second annual list of the top 50 hockey players on the nhl roster.
The main reason I put this together last year (you can see it here) was as a basis for comparison to the other top 50 most famous player lists. The annual list is a season preview staple for TSN and TNH and the rankings are often a bit controversial. both lists are created through a survey of various people within hockey who are generally very smart people but also prone to old school thinking and value is sometimes shaped by recency bias, reputation and a winning pedigree.
this list is a bit of the opposite, as it mostly comes from outsiders, people who study and analyze the game in the public sphere. that’s not to say they’re necessarily smarter people, they just approach the game from a different angle based primarily on underlying trends and numbers over more traditional stats and what you immediately see on the ice.
The most interesting thing will be how much the list created by outsiders differs from the one created by insiders (tsn list). Is anyone higher on this list being underrated or overrated by the analytics community? Is anyone who isn’t ranked here but on the tsn list worthy based on something missing from the numbers? that won’t be answered here, but it’s something worth thinking about.
now to the list. last year 15 people submitted lists, this year we have 18, which is three (3) more chances of being wrong.
Because these types of posts tend to get a lot of views, I asked each person responsible for the list to share a link to a work (theirs or someone else’s) that they think deserves more recognition and most of them did exactly that. . here is the list of contributors along with some great work everyone should read:
@mannyelk: k, a total value statistic using a composite custom regression model
@steveburtch: tactalytics: defending the pass
@andrewberkshire: Top 20 by the numbers c lw rw d g
@seantierneytss: pk flyer analysis: systems and tactics
@ghostofnyles: neutral zone playstyles
@robvollmannhl: dubnetwork: whl statistics database
@hockeyanalysis: do we predict the future or analyze the past?
@shane1342o: Predicting Playoff Series Using Logistic Regression
@ryanstimson: Is aggressive pk play worth it?
@kikkerlaika: under the microscope: pittsburgh-washington playoff series
@cane_matt: measuring the productivity of a single game
@madeupcall: a fun introduction to hockey analysis
@bsh_charlie: Evaluation of the impact of the deployment in the suppression of individual shots in the pk
@dtmaboutheart: nylon calculation
These are who they collectively believe are the 50 best players in the nhl for 2016-17. what was interesting was that the different groups of players were grouped into levels and there were some clear drops after certain players. I’ve illustrated it a bit with some line breaks in the list below.
In the event of a tie, I opted for the player who appears in a higher percentage of the top 5, then the top 10, then the charts, then the highest rank on any individual list.
10 thoughts on the list
1. Sidney Crosby being the best player in the league is something almost everyone can agree on. he was unanimous on the tsn chart and almost unanimous here, finishing first on 15 of 18 charts. Best D-Man was a bit more decisive as Erik Karlsson finished at the top of this list and drew Doughty ahead at TSN’s. Karlsson was the best d-man on 16 of 18 lists here and second overall, while Doughty finished fifth behind Karlsson, Hedman, Subban and Giordano.
2. Crosby’s heir to the throne is very much in evidence, as only one other player got more than one first-place vote: Connor McDavid. He was in the top five on half the charts and in the top ten on all but four, both numbers third after Crosby and Karlsson. after electrifying the world for half a season, he makes a huge jump from 43rd last year to 3rd this season.
3. Victor Hedman got a lot of love here finishing as the second highest D-Man and finishing 11 places higher than the TSN chart. he’s one of the guys here that the numbers crowd feels he doesn’t get as much love as he should. Others on that list (10 spots or more above TSN’s list): Patrice Bergeron, Joe Thornton, Taylor Hall, Oliver Ekman-Larsson, Filip Forsberg, Cory Schneider. Players who made the cut who weren’t on the TSN roster: Mark Giordano, Jakub Voracek, John Klingberg, Max Pacioretty, Hampus Lindholm, Tyler Toffoli, Marc-Edouard Vlasic, and Anton Stralman.
4. the flip side of that? jonathan toews, patrick kane, bravely drew, steven stamkos, ryan getzlaf, nicklas backstrom, duncan keith, roman josi were ranked lower while corey perry, braden holtby, mark scheifele, shea weber, artemi panarin, alex pietrangelo, aaron ekblad got votes but didn’t make the cut. jonathan quick got zero votes, as expected.
5. goalkeepers actually got far fewer votes in the year overall. only three goalkeepers made it to the list, although a further five received votes. two contributors had no goalies on their roster entirely as a testament to their perceived value relative to skaters, the variety of talents, and the overall uncertainty at the position. centers dominated the list, while wingers and d-men were equally represented.
6. As for team representation, San Jose leads the way with five guys on the roster while Tampa Bay was next with four. pittsburgh, dallas, washington, chicago, nashville and los angeles each had three. basically a list of the best teams in the league. teams with no one on the roster: carolina, columbus, minnesota, toronto, vancouver.
7. there were 13 players who were on last year’s roster who didn’t make the roster this year: tuukka rask, corey perry, pavel datsyuk, phil kessel, rick nash, zach parise, jake muzzin, tyler johnson, jeff carter, ryan johansen , henrik zetterberg, shea weber, kevin shattenkirk.
8. players 51 and 52 appeared on 56 percent of the lists, meaning that more than half of the panel thought the duo belonged in the top 50, but they were beaten by too many votes in the 40-50 range. players were daniel sedin and dustin byfuglien. ryan o’reilly finished in 48th place despite appearing on only a third of the charts.
9. the top five proved very difficult to decide. Only one player (Crosby) was in the top five for each person and only one (Karlsson) was in more than half (McDavid was exactly in the middle). Karlsson finished second and his average rank was exactly five, with only six players having an average rank below 10. In total, 19 different players were in someone’s top five and 25 were in someone’s top 10. the top of the nhl is so close now that it’s no surprise that the question “who are the top five players in the nhl” gets very different results.
10. there were 112 different players on the 18 rosters. Opinions on who the best players in the league are differ wildly, and while a composite roster like this is great, it’s likely to be different from the one you’re thinking in your head. the variance between how one person sees one player versus another is insane. I made a visualization of where each of the top 50 players got a vote (and that doesn’t take into account how many times they weren’t ranked) to show how many opinions differ.