Whether you form your hockey opinions by looking at the standings; base your analysis on corsi, fenwick, and PDO; or rely on the much-maligned (revered?) “eye test” method, there’s no mistaking that the New York Islanders regressed and regressed hard during the 2013-14 NHL season.
The Islanders (34-37-11, 79 pts) finished the regular season eighth in the Metropolitan Division, 14th in the Eastern Conference, and 26th in the NHL. And as befitting a team that ranked among the league’s worst, the underlying numbers weren’t good: bottom half of the league in goals scored per game, among the three worst teams in goals allowed per game, and dead last in the NHL in team save percentage.
(Yikes. My apologies to Isles fans for not prefacing that section with #NSFW.)
The fancy stats—while similarly grisly from a team perspective—can offer us insight into where the team faltered. Or more importantly, how players performed in the situations in which they were utilized by the coaching staff. Enter: player usage charts.
Player deployment is a part of the game that’s receiving more and more attention as coaches and general managers learn that how players are deployed is just as important as how players perform. (You know, all that stuff about maximizing potential outcome and mitigating risk and whatnot.)
This is a gross oversimplification, but: a coach can’t reasonably expect a league-average player to put up stellar numbers when continually matched against above-league-average competition. This mindset already factors into the matchup decisions coaches make during games; whether coaches know they’re relying on principles like zone starts and quality of competition when barking out their line changes is up for debate (in most cases).
Regardless of how he manages the game, a coach generally wants to get his top forward line on the ice against less-talented opposition, and his top defensive pairing matched up against the other team’s best players. This isn’t rocket science, and the way teams are gathering data to achieve the best possible matchups is being codified right before our eyes.
What does this all boil down to? Get used to seeing player usage charts in your hockey coverage, basically. (If you’re not used to seeing them already.)
Hockey Prospectus founder Rob Vollman (Hockey Abstract, Bleacher Report, ESPN) is best known for his development of the player usage charts which have since become a staple in the hockey analytics community. (You can read his latest usage analysis of each NHL team here.) He’s the guy behind the player-names-in-circles thing, which means we have him to thank for those visual representations of player deployment.
Before we get to the Islanders’ chart for this past season (courtesy of Vollman, of course), let’s first learn how to read it. From Vollman:
Player usage charts are explained in rich detail in a number of places, but here’s a quick cheat sheet of the most important points to remember.
- Blue is good; white is bad. Blue means the team is doing better than usual possession-wise with that player on the ice; white means they’re doing worse.
- The size of the circle indicates the size of that advantage or disadvantage.
- Shutdown players are at the top, leaning to the left, whereas sheltered players are on the right, leaning down.
- Players at the top of the chart are facing tough, top-line opponents, whereas those at the bottom are facing mostly depth lines.
- Players on the left are tasked with starting most of their shifts in the defensive zone, whereas those on the right have the advantage of a lot of offensive zone starts.
- Asterisks denote players who either began (prefix) or ended (postfix) the season elsewhere.
Remember: context is everything. Those playing tough minutes will probably have white circles, while those enjoying favorable minutes ought to have nice blue ones.
Now, for the chart itself:
Using Vollman’s criteria, there are a few things we can determine right away: 1) the Thomas Vanek—John Tavares—Kyle Okposo line was good, 2) the Josh Bailey—Frans Nielsen—Michael Grabner line quietly locked down the opposition, and 3) Eric Boulton will knock you out if you mention this chart to him.
The Vanek—Tavares—Okposo line generally played tough competition (as indicated by their position well above the chart’s median y-axis) and generally out-possessed the opposition (two out of three blue circles ain’t bad). Pretty standard first line stuff here, but still a good indication at how lethal that combination was, especially when they led the NHL in scoring from mid-December until mid-January. (Ah, memories.)
On the second line, Bailey took a lot of heat for a lackluster offensive year—and rightly so, the kid was drafted at No. 9 overall in 2008—but he was a big part of how defensively sound this trio was. Bailey—Nielsen—Grabner might not have gotten the headlines it deserved, but this line played the toughest competition night in and night out, while still outplaying what we can safely assume were the opposing teams’ top lines.
A couple of other important takeaways regarding how head coach Jack Capuano deployed his players…
The Lubomir Visnovsky concussion was bad
Which: duh. Look at his circle on the chart: it’s the largest of any defenseman and blue, meaning that while facing above-average competition, Visnovsky was able to control the puck and do so in a way that boosted the Isles’ performance more than any other player on the roster. It’s a small sample size, sure—he only played 24 games—but his track record proves that his team is much better when he’s on the ice.
The kids were good
Anders Lee, Brock Nelson, and Ryan Strome were drafted with the intent that they one day would form the core of the Islanders’ forward group (alongside franchise cornerstone Tavares), and it appears that all three are well on their way toward doing so. Lee and Nelson played tougher competition than Strome, but all three players boosted the club’s possession rate when they were on the ice. Somewhere, Islanders GM Garth Snow is smiling.
The fourth line was suspect
Having an “energy line” is important, but when most of that energy is spent backchecking it doesn’t do the team much good. The line of Matt Martin—Casey Cizikas—Colin McDonald took a definite step backward compared to their performance during the lockout-shortened 2012-13 campaign, and the chart reflects its struggle this past season. All three players were outplayed by inferior competition, which no coach wants to see. Capuano can’t expect his fourth line to dominate possession on a consistent basis, but he at least would like to see them get back to pestering opposing defensemen on the forecheck rather than chasing the puck in their own defensive zone next season.
The defense is in flux(?)
Calvin de Haan had a good year. Travis Hamonic? Not so much. Both players faced roughly the same level of competition and were tasked with starting more than half of their shifts in the D-zone, but de Haan was able to drive possession at a much higher rate than Hamonic was. When de Haan was on the ice, the Isles out-chanced the opposition; the reverse is true of Hamonic’s shifts. Does this mean de Haan increasingly sees more top-pair time next year? With the impending arrival of uber-prospect Griffin Reinhart, there might be an opportunity to pair de Haan with Reinhart, thereby bumping Hamonic to the second pair where weaker opposing competition likely awaits.
Player usage charts shouldn’t be taken as the be-all, end-all of statistical analysis. What they allow is for coaches and fans to place a given player’s performance in perspective by combining puck possession with situational deployment. Should a coaching staff rely solely on these charts to gauge how the season went? No. Can a coaching staff make better-informed personnel decisions by using them? Absolutely.
Follow me on Twitter (@MichaelWillhoft) for objective* hockey analysis and hilarious** jokes. (*biased, **mediocre)