The Oilers made a sensible deal last week acquiring forward James Neal from the Calgary Flames in exchange for Milan Lucic. For a minor fee, the Oilers were able to bring on a forward that has a better chance of bouncing back, considering the drop in his shooting percentage last season compared to his career average, and they also gained some much needed cap flexibility in the future if they need to buy-out the remaining years of the contract. The Flames slightly reduced their cap-hit and also shed some real dollars, but I still can’t understand why they were so desperate to move Neal that they would take on a heavy, burdensome contract like Lucic’s.

Below are Neal’s even-strength (5v5) stats since 2008/09, including his total goals, assists, points, shots, shots per hour and individual shooting percentage.

Season	Team	GP	TOI	G-A-P	Shots	Shots/60	Sh%
2008/09	DAL	77	949.37	15-11-26	126	7.96	11.90
2009/10	DAL	78	1064.82	22-19-41	148	8.34	14.86
2010/11	DAL/PIT	79	1119.63	16-18-34	157	8.41	10.19
2011/12	PIT	80	1134.65	22-26-48	208	11.00	10.58
2012/13	PIT	40	514.70	11-7-18	93	10.84	11.83
2013/14	PIT	59	802.70	14-17-31	153	11.44	9.15
2014/15	NSH	67	950.97	16-13-29	170	10.73	9.41
2015/16	NSH	82	1248.35	24-13-37	197	9.47	12.18
2016/17	NSH	70	978.22	15-9-24	148	9.08	10.14
2017/18	VGK	71	992.12	17-12-29	156	9.43	10.90
2018/19	CGY	63	784.55	5-8-13	108	8.26	4.63

What we know about Neal is that his point production has gradually declined, with his most recent season being his absolute worst. What especially stands out is his individual shooting percentage of 4.63% last season, an anomaly over the course of his career and well below his career average prior to the 2018/19 season (11.11%). Knowing that individual shooting percentages typically regress towards the mean over time, there’s a decent chance Neal’s bounces back next season. How much of a bounce-back depends of course on the player’s aging, as well as the situations (i.e., teammates, competition, zone starts, etc) that the player will be in.

While the Oilers hope that Neal’s 2018/19 season was an anomaly, it’s worth noting that his overall play has gradually declined even prior to his arrival in Calgary. Looking at his relative-to-team stats over the last five years, we see that his more recent teams have done better without him at even-strength than with him. The graph below includes relative to team Corsi For% (CF% – proportion of shot attempts, a proxy for possession), Fenwick For% (FF% – proportion of unblocked shot attempts, a proxy for scoring chances), Scoring Chances For% (SCF% – as defined by Natural Stat Trick), and High Danger Scoring Chances For% (HDCF% – as defined by Natural Stat Trick).

For instance, in his single season in Vegas in 2017/18, the Golden Knights posted a 50.51% Corsi For percentage at even-strength with Neal on the ice, but a slightly better proportion of shot attempts without him. Relative to the team, Neal was a -0.51 Corsi-rel, which must have been a little disappointing for the Vegas coaching staff considering he received top-six minutes (finished fourth on the team among forwards in ice time per game with 13:58) and he always posted solid numbers with and without star players in Pittsburgh and Nashville. Vegas also did better without Neal when it came to scoring chances, a trend that continued and got worse in Calgary.

What’s interesting about Neal’s stint in Calgary is that he finished seventh on the team among forwards in ice time per game (12:27), most often playing in the bottom six with Mark Jankowski and Sam Bennett. His reduced playing time might have to do with the fact that against elite players, his on-ice numbers had been in decline. (Data: Puck IQ)

Season	Team	TOI	TOI%	CF% (Rel)	DFF%	GF%
20142015	NSH	335.98	35.42	53.6 (+5.40)	51.9 (+0.90)	57.10
20152016	NSH	470.25	37.74	55.8 (+8.00)	60.2 (+9.50)	71.40
20162017	NSH	315.08	32.34	48.4 (-1.60)	48.1 (-2.50)	39.10
20172018	VGK	378.37	38.05	44.9 (-6.40)	43.4 (-12.70)	41.70
20182019	CGY	223.90	28.54	52.4 (-0.60)	47.9 (-5.10)	50.00

The table above includes Neal’s numbers against elite talent, as defined by Puck IQ. Included is Neal’s on-ice Corsi For percentage and relative to team mates number, as well as Dangerous Fenwick (DFF%), which is a “weighted shot metric using shot distance location and type of shot to give each shot a danger value”, and goals-for percentage (GF%).

In Nashville, and prior to that, Neal was playing often against the other team’s top end players, spending over 35% of his ice time at even-strength in these situations and faring quite well relative to his teammates. In 2016/17 however, while playing 32.3% of his time against top end talent and finishing second on the team among forwards in ice time per game, he posted a 48.40% Corsi For percentage in these situations, which was a drop relative to his teammates (-3.73). When he was left unprotected by the Predators in the expansion draft and went to Vegas, things got even worse as he again played often against elite talent but posted a very poor 44.90% Corsi For percentage, or a -6.40 relative to his teammates. The Flames coaching staff may have been aware of his declining performance against high end talent and his poor goal-share as they deployed him far less frequently against elites (28.5% of his even-strength ice time spent against them) and did see him post a very solid 52.40% Corsi For percentage (-0.60 relative to teammates) and a 50.0% goal-share. Unfortunately, this reduced playing time meant less time with skilled forwards, which may have played a role in his 4.63% individual shooting percentage.

It will interesting to see how the Oilers manage Neal, who because of the lack of scoring talent will likely get plenty of opportunities playing with McDavid, Draisaitl or Nugent-Hopkins. Hopefully the Oilers are aware of Neal’s declining on-ice numbers and how poorly he’s fared against elite talent when he was getting top six minutes. It would be in the Oilers coaching staff’s best interests to have a deployment plan in place to mitigate any risks Neal’s acquisition brings to the team and get as much offence as they can from the player.

Data: Natural Stat Trick, Puck IQ

Related:

• How many goals will James Neal scoring in 2019-20 – Boys on the Bus (2019, July 23)
• A deep dive behind the numbers to gauge expectations for Milan Lucic and James Neal – The Athletic (2019, July 22)
• Lucic-for-Neal swap brings Ken Holland’s vision for 2019-20 Edmonton Oilers into focus – Cult of Hockey (2019, July 23)

Also posted at The Copper & Blue.

Following the Oilers signing of Markus Granlund to a one-year, $1.3 million contract, we were immediately informed of his experience and success on the penalty kill in Vancouver. And that it was a reason why the Oilers signed him.

The Oilers were 30th on the PK at 74.8 per cent with Chicago 31st at 72.7 per cent last year. Granlund did his best work on a shutdown line with fourth-line centre Jay Beagle and helped the Canucks 11th overall penalty kill with centre Brandon Sutter. (Source: Edmonton Journal)

When you get information like this – that’s obviously being distributed by the management group that just invested in a player, and probably the player’s agent as well – it’s always a good exercise to suss out just how much of an impact an individual player had on the overall success of the team.

In this case, we know that the Canucks had a decent penalty kill last season, finishing 13th in the league allowing 6.86 goals against per hour. This was due in large part to their ability limiting shots-on-goals against, ranking 9th in the league allowing 48.3 per hour. Had their goaltending been league average (their team save percentage ranked 21st in the league with 85.80%), their penalty kill very likely would have finished in the top ten. And we also know that Granlund led all Canucks skaters in total ice time on the penalty kill, and was fourth among forwards when it came to average ice time per game. So naturally, one could connect the information and assume that since Granlund played significant minutes on the penalty kill, he must have had a positive impact.

But when you dig into Granlund’s on-ice numbers, you realize that that’s not the case at all.

In the team-leading 183 penalty kill minutes that Granlund was on the ice, the Canucks allowed a rate of 9.16 goals against per hour. To put things into perspective, the Oilers penalty kill allowed a rate of 9.21 goals against per hour, good for 30th in the league. In the 236 penalty-killing minutes that Granlund wasn’t on the ice, the Canucks allowed a rate of 5.07 goals against per hour, which is right around the rate the Lightning and Coyotes posted as top penalty kill units last season.

Now the rate of goals against are heavily influenced by the goaltenders performance, which as mentioned above was below league average, so you can’t put everything on Granlund. Having said that, the rate of shots against – which players do influence as it’s part of their job – also saw a jump when Granlund was on the ice compared to when he was on the bench. Last season with Granlund on the ice, the Canucks allowed 54.31 shots-on-goals against per hour; without him that number dropped significantly to 43.63. What’s alarming is that even the season prior (2017/18), Granlund’s penalty kill numbers were just as poor. The team allowed the second lowest rate of shots against (49.8) in the league; with Granlund on the ice they allowed 60.71 shots per hour and without him they allowed 45.29. For someone being touted as a penalty kill option, it’s strange that his former team had so much better success limiting shots without him on the ice.

It’s also worth looking into the impact Granlund had on his teammates, specifically the defencemen, to see if someone else was potentially driving up his on-ice rate of shots against. Below is a graph showing what the rate of shots-on-goal against were when Granlund was deployed with the various Canucks defencemen last season, and how those defencemen did away from Granlund. Included is the team’s overall rate of shots-on-goal against (grey line across). Again, I focused on shots as it’s something that the players can influence, while the rate of goals against are more reliant on the goaltender’s performance.

Across the board, the most commonly deployed Canucks defencemen saw their own on-ice rate of shots-on-goals against drop on the penalty kill when they didn’t have Granlund on the ice with them. For whatever reason, each defencemen’s numbers would jump when Granlund was deployed with them, making you wonder why that would happen and how the Oilers think they’ll mitigate the issue. This of course is assuming they know about Granlund’s past penalty kill numbers – they did just sign him to a  contract and penciled him in as a penalty kill option.

Considering how poorly the Oilers performed on the penalty kill last season, and the massive impact it’s had on their overall goal-differential, you would hope that the Oilers have identified their needs and have a plan in place to bring in the right personnel and tactics. So far, it doesn’t appear to be the case.

Related: Penalty Kill Expectations – The SuperFan (2019, May 30)

Data: Natural Stat Trick

Also posted at The Copper & Blue.

UPDATE: 2019, July 5 (11:50 PM)

Couple important points I want to add based on some of the feedback I received.

First, we can assume Granlund was on Vancouver’s first penalty kill unit based on his total ice time. However, as I mentioned in the article, his average ice time per game was third among forwards last season, meaning when the roster was healthy and Brandon Sutter was available, Granlund was moved down to the second powerplay unit where he’s probably better suited.

Another question I received was how Granlund’s on-ice numbers compared to the top penalty killing forwards on other rosters. Below is a list of 31 forwards (one from each team) and their on-ice rate of shots and goals against relative to their teams.

One thing that stands out is that because this group of players likely played against the other team’s top powerplay unit, their on-ice rate of shots against were higher relative to their team – the average among the group is +5.45 shots against per hour. So it should be no surprise that Granlund’s numbers were poor. Having said that, compared to his peers, Granlund’s on-ice rate of shots against (+9.36), as well as goals against (+4.04) relative to his team were one of the worst.