Statistics predict which teams will benefit most from moves

Every offseason, each front office across the NHL has specific priorities to address from its unique set of needs. Some are looking for help on special teams, others want to improve their goaltending situation, and there are those searching for another option in the shootout.
Modern hockey analytics can help identify a team's needs, and how successfully those needs have been addressed during the offseason. From a statistical perspective, here are the teams that have improved the most this offseason in six categories: coaching, power play, penalty kill, shootout, hitting and goaltending.

Coaching

Biggest improvement: Dallas Stars
There is a way to establish just how valuable Ken Hitchcock can be for the Stars, who hired the 65-year-old for his second stint as their coach on April 13.
By comparing how a coach's team has performed in terms of points in the standings and then subtracting the team's statistical expectations, a coach's impact can be reliably measured over the long term. As explained in my 2013 book "Hockey Abstract," these expectations can be set in a variety of ways, most simply by taking the team's points from the previous season regressed 35 percent toward the League average.
Using this approach, Hitchcock has added an average of 5.2 points per season to the standings.

Power play

Biggest improvement:New York Rangers
New York signed Kevin Shattenkirk, the most dominant defenseman on the power play in quite some time, to a four-year contract July 1.
Over the past four seasons combined, Shattenkirk has scored an average of 6.21 points per 60 minutes at 5-on-4. To put that in perspective, the second-ranked defenseman is John Carlson of the Washington Capitals at 5.17 (source: Hockey Analysis).
If he continues at that rate, Shattenkirk will be the best Rangers defenseman on the power play in their history.
Since ice-time data is available back to the 1997-98 season, Shattenkirk's scoring can be placed in a historical context by comparing it to former Rangers defenseman Brian Leetch, who was elected to the Hockey Hall of Fame in 2009. From 1997-98 to 2000-01, Leetch's average scoring rate on the power play was 3.81 points per 60 minutes.

Penalty kill

Biggest improvement: Arizona Coyotes
It's fair to say the Coyotes have room for improvement on the penalty kill. In traditional terms, they've finished each of the past four seasons no higher than 26th out of 30 teams in penalty-killing percentage. In terms of shot-based metrics, they have allowed a League-worst 106.5 shot attempts per 60 minutes while shorthanded.
The arrival of defenseman Niklas Hjalmarsson in a trade from the Chicago Blackhawks should help, but the key player to watch could be center Derek Stepan, who was acquired in a trade from the Rangers on June 23. Over that same four-season time span, New York allowed 84.67 shot attempts per 60 minutes when Stepan was on the ice at 4-on-5, 19th among forwards (minimum 200 minutes), compared to 93.76 when he wasn't (source: Hockey Analysis).

Shootout

Biggest improvement:Buffalo Sabres
The Sabres were 4-13 in the shootout the past two seasons combined, having scored on 10 of 54 shots (18.5 percent). Their opponents scored on 25 of 53 (47.2 percent).
To address this issue, the Sabres added two shootout specialists, forwards Jason Pominville (in a trade from the Minnesota Wild) and Jacob Josefson (as a free agent). A player's value in the shootout can be measured by the number of goals he has scored over and above a League-average shooter, who will score on 31.4 percent of his shots. From that perspective, Pominville has provided an extra 6.6 shootout goals in his NHL career (15th among active NHL players), and Josefson has added 4.4 (23rd).
What exactly is that worth? According to
NHLnumbers.com
, each shootout specialist can increase a team's chance of winning a particular shootout by 8 percent and is worth an extra 0.68 wins per season. Statistically, this impact would be even greater on a below-average team like the Sabres.

Hitting

Biggest improvement:Chicago Blackhawks
For the fifth time in the past six seasons, the Blackhawks were last in the NHL in hits. To become more physical, they added two of the League's best hitters among free agent forwards, Tommy Wingels and Lance Bouma.
The latest way to measure a player's hitting is in a fashion that controls for manpower situation and ice time, and adjusts for the different rates at which hits are recorded in NHL arenas. From this perspective, Bouma was 17th in the League with 13.0 adjusted hits per 60 minutes over the past three seasons combined, and Wingels was 29th with 11.4 (source: "Stat Shot: The Ultimate Guide to Hockey Analytics").

Goaltending

Biggest improvement:Winnipeg Jets
Ever since the location of shots was added to NHL game files in the 2002-03 season, the ultimate way to measure goaltending performance in
hockey analytics
is with a version of save percentage that takes the difficulty of each shot into account.
The original version of this statistic was developed by Alan Ryder in 2004 and involved assigning a weight to each shot based on the average probability that it would go in, based on its location and other quality-related factors. Goalies could be evaluated by comparing the actual number of goals they allowed to this calculated number of expected goals.
Based on the latest version of this statistic, which was provided to me by the author of the
Don't Tell Me About Heart
website, a League-average goalie would have a .9318 save percentage at 5-on-5 when facing the same volume and quality of shots as Steve Mason over the past three seasons combined. Subtract that from his actual even-strength save percentage of .9388, and Mason's difference of .0070 was seventh among the 64 goalies to have faced at least 1,000 shots in that time.