Zone Deployment Optimization: Wingers

Throughout the previous week we've shown the effects of zone deployment and how to use it to the Jets advantage. We continue this in depth study, now turning our focus to wingers

The Winnipeg Jets have quite a few players who will be playing on Bryan Little, Mark Scheifele, Mathieu Perreault, and Jim Slater's wings. Unfortunately due to sample restrictions, most of the analysis focus will be limited to Blake Wheeler, Chris Thorburn, Michael Frolik, Andrew Ladd, and Evander Kane. This means limited analysis for Matt Halischuk, T.J. Galiardi, Eric O`Dell, and Anthony Peluso and no analysis for NHL rookie Adam Lowry.

Galiardi comes in as a new addition for the Jets, so his numbers are skewed from playing under a different system, with different teammates, and also being compared against other players.

Note: All numbers are for 5v5 events and 2013-14 unless otherwise indicated. Much of the data and graphs have been derived with the help of Muneeb Alam's, Adam Stringham's and Corey Sznadjer's amazing talents. Please give them a follow and support them in their work.

Contextual Nuances

Before we start discussing how these players performed against each other, we should offer some context. After all, there are nine other players on the ice and these players affect the results as well.

Player	TmCF/20	TmCA/20	OpCF/20	OpCA/20
Michael Frolik	19.62	18.51	19.14	19.02
Blake Wheeler	20.08	19.43	19.19	19.21
Eric O'Dell	19.03	19.32	18.94	19.42
Anthony Peluso	19.05	19.38	18.66	19.04
Chris Thorburn	19.27	19.68	18.95	19.37
Evander Kane	19.37	19.82	19.12	19.20
Andrew Ladd	19.25	19.77	19.27	19.18
Matt Halischuk	17.89	19.41	18.95	18.74

TmCorsi is how a player's linemates perform with the player not on the ice, while OpCorsi is the same but for opponents. The players are listed in descending order of TmCD/20 (CD = CF-CA).

We can see here, as an example, that both Ladd and Frolik played against top players, but Frolik mostly played with linemates who performed well even away from him.

Defensive Zone

Maurice stressed defensive zone performance as a priority for the Jets. The goals against average needs to come down. After all, Ondrej Pavelec's performance obviously cannot be assessed until team defense improves.

Post Defensive Zone Face Off Shot Attempt Rates:

The lines represent the rate of shot attempts. Black is shot attempts for with the player on the ice, which you want to have higher than red (shot attempts for with the player off the ice). Blue is the shot attempts against with the player on the ice, which you want lower than green (shot attempts against with player off the ice).

The difference between Ladd and Kane is unexpected. Kane actually ends up with more shot repression, while Ladd generates more offense. Ladd comes away with the better differential in the end.

Microstatistics - Zone Exit:

Defensive zone effectiveness goes beyond defending a defensive zone face off. Most of the game is played in fluid motion. A major component to defensive zone effectiveness comes from a player's ability to successfully breakout with control of the puck.

Player	Touches	Success%	Turnover%	Icing%
T.J. Galiardi	199	52%	4%	2%
Evander Kane	315	50%	6%	1%
Matt Halischuk	141	47%	10%	1%
Blake Wheeler	391	42%	4%	2%
Michael Frolik	346	40%	5%	1%
Eric O'Dell	42	38%	14%	2%
Andrew Ladd	358	37%	7%	0%
Dustin Byfuglien (F)	40	30%	3%	8%
Chris Thorburn	168	25%	8%	4%
Anthony Peluso	76	22%	11%	1%

There are a few non-surprises. Kane performed superbly in breakouts, as with Wheeler. Meanwhile, Thorburn and Peluso floundered at most everything.

There are also a few surprises with Halischuk and Galiardi having success in moving the puck, and with Ladd and Byfuglien struggling.

Neutral Zone

The eye-test struggles to evaluate players on neutral zone play, most likely because humans psychologically depend on direct impact for memory and tends to miss indirect impact. However, there is evidence that the neutral zone's importance is very important, and therefore highly underrated.

Post Neutral Zone Face Off Shot Attempt Rates:

The lines represent the rate of shot attempts. Black is shot attempts for with the player on the ice, which you want to have higher than red (shot attempts for with the player off the ice). Blue is the shot attempts against with the player on the ice, which you want lower than green (shot attempts against with player off the ice).

On the left wing, Ladd kills it for neutral zone face offs, while Evander Kane plays the face off as a high event game.

The right wing lacks any standout players. Thorburn has had moments of out shooting and out shot, while Wheeler and Frolik have experienced different levels of being out shot.

Microstatistics - Zone Entry:

Neutral zone microstatistics first started with zone entries. Eric Tulsky (et al.) found that a major driver of shot production came from play in the neutral zone, specifically in entering the offensive zone with control of the puck.

Player	# of entries	Shots per entry	# of carry-ins	Shots per carry-in	% of entries with control	# of dump-ins	Shots per dump-in
Evander Kane	291	0.54	165	0.73	56.7%	126	0.29
Dustin Byfuglien (F)	31	0.42	17	0.59	54.8%	14	0.21
Blake Wheeler	304	0.45	150	0.73	49.3%	154	0.18
Michael Frolik	231	0.47	109	0.72	47.2%	122	0.25
Andrew Ladd	278	0.51	121	0.84	43.5%	157	0.25
Eric O'Dell	27	0.41	11	0.55	40.7%	16	0.31
Matt Haslichuk	105	0.51	35	0.71	33.3%	70	0.41
Anthony Peluso	52	0.38	16	0.75	30.8%	36	0.22
Chris Thorburn	80	0.41	21	0.81	26.3%	59	0.27

There is a Jets twitter meme that goes: Keep Kane and Carry-in. Kane kills it for gaining the zone. Kane leads in gaining entries with control and also generating shots from his entries. Byfuglien's and O'Dell's shots from controlled entries is quite low; it will be interesting to see if that rises or if it is from something they do.

Differences in icetime skew the numbers, so we can add context by adding that 5v5 TOI:

Player	Entries per 60	Shots per 60 from entries	Shots per 60 from controlled entries
Evander Kane	18.5	10.1	7.7
Andrew Ladd	15.5	7.9	5.7
Blake Wheeler	15.9	7.2	5.7
Matt Haslichuk	13.4	6.9	3.2
Michael Frolik	12.3	5.8	4.2
Chris Thorburn	9.9	4.1	2.1
Anthony Peluso	10.3	4.0	2.4
Eric O'Dell	5.9	2.4	1.3

Not much surprise in the order of players with large sample sizes. Kane, Ladd, Wheeler are strong offensive players, while Thorburn and Peluso value derives not from their offense.

Microstatistics - Denial of Opposition Zone Entry:

The obvious byproduct to research in zone entries came the discovery of the importance in denial of opposition zone entry.

Player	Targets	Carry%	Break-up%	-Rel. Carry%
Anthony Peluso	2	0%	0%	+44%
T.J. Galiardi	5	40%	40%	+40%
Evander Kane	12	25%	33%	+19%
Chris Thorburn	17	29%	18%	+14%
Matt Halischuk	8	38%	38%	+6%
Michael Frolik	15	40%	27%	+4%
Blake Wheeler	14	50%	29%	-6%
Andrew Ladd	16	56%	31%	-13%
Dustin Byfuglien (F)	5	80%	0%	-36%

The sample size of these numbers are extremely low, as expected by being forwards. It is anecdotally intriguing that most of the Jets offensive players are poor in denying entry and most of the Jets bottom players are not. If this difference is somewhat significant, is this due to differences in play style? Or is this because these bottom players are strong defensive players? Or is this from differences in line matching with the the top players facing top line matches more often?

What is interesting to see is targets relative to icetime. Thorburn has the largest target numbers with very little icetime. Are opponents singling him out because they view him as an easier one-on-one target?

Microstatistics - Overall Neutral Zone Score

Tuslky (et al.) combined the effects of zone entries and zone exits -with the inclusion of events not directly impacted by the individual- in a formula to evaluate the overall neutral zone effectiveness in gaining and preventing zone entry.

NHL Rank (N=388)	Names	NZ score	NZ Score Relative	% of On-Ice Z.E For	Control% For	Control% Against
10	Kane	51.6%	4.2%	51.2%	48.2%	46.0%
104	Galiardi	48.7%	1.5%	49.8%	45.0%	50.3%
106	Wheeler	49.5%	1.4%	50.4%	44.1%	48.4%
172	Ladd	48.9%	0.5%	49.7%	46.0%	50.1%
205	Frolik	48.5%	0.0%	49.1%	44.7%	47.4%
321	Thorburn	46.6%	-2.2%	47.6%	35.0%	39.6%
340	Haslichuk	46.2%	-2.6%	47.3%	39.9%	45.0%

While Kane struggles with neutral zone face offs, his neutral zone performance during open play is outstanding. This is what makes Kane the 5-on-5 beast he is and why he will always produce extremely well without much help from line mates. Kane, Little, and Perreault are the only Jets with a neutral zone score over 50%. This will need to improve in the future if the Jets are to grow into contenders.

Using neutral zone data, we get a glimpse into why Halischuk has one of the worst Corsi% regulars in the NHL while still finishing at high rates. His neutral zone score is terrible while he still generates a large volume of shots from the entries he gains.

The ranking on the left column is in terms of relative score.

Offensive Zone

Relatively speaking, the offensive zone tends to be easier to grasp the most effective players, but still there can be some surprises.

Post Offensive Zone Face Off Shot Attempt Rates:

The lines represent the rate of shot attempts. Black is shot attempts for with the player on the ice, which you want to have higher than red (shot attempts for with the player off the ice). Blue is the shot attempts against with the player on the ice, which you want lower than green (shot attempts against with player off the ice).

Ladd and Wheeler unsurprisingly perform best for ofensive zone face offs. They both sustain large shot volume rates for extended periods. Kane and Frolik are not far off, but still not at the same level.

2011-14 Point Production

Player	Points/60
Blake Wheeler	2.07
Andrew Ladd	2.04
Evander Kane	2.00
Matt Halischuk	1.80
Michael Frolik	1.51
Eric O`Dell	1.31
T.J. Galiardi	1.20
Anthony Peluso	1.10
Chris Thorburn	1.10

Wheeler, Ladd, and Kane have consistently performed at or above top line rates for point production. Frolik's production was stifled playing a fourth line defensive role for the Chicago Blackhawks. Removing those two seasons puts Frolik's production since 2007 at 1.73 per 60 minutes.

Closing Thoughts

Combining this information with our last post on centres, we would suggest:

Ladd-Little-Frolik/Byfuglien: Ladd, Little, and Frolik are quite the utility players. They can be placed in almost any situation or zone and expect to perform anywhere from well to exceptional, especially when together. They can take on the tough minutes allowing to the other top six line's benefit. This creates a top six system very similar with Boston's with the Krejci and Bergeron lines.

Kane-Scheifele-Wheeler: While Maurice says he doesn't have concerns for this lines defensive game, he should have enough concern that he tries to avoid deploying them overly in the defensive zone. For key face offs Maurice might want to throw one of Perreault or Slater who have performed not just well in FO% but also post-face off Corsi%. Otherwise, the line can be used be used interchangeably with Little line as primary option for neutral and offensive zone.

Lowry-Perreault-Frolik/Byfuglien: Lowry remains an unknown, although his defensive performance in the AHL was above average. An optimist would also call Byfuglien an unknown, although Byfuglien's poor performance as a forward last season did seem to be stable. Perreault, however, has competently performed in most zones, especially in the defensive zone against lower competition. The Perreault line can be deployed often in the defensive zone, especially if Frolik is the right winger, while being strong secondary options for neutral zone and offensive zone.

Halischuk/Galiardi-Slater-Peluso/Thorburn: Slater has shown value in defensive zone deployment. Halischuk gives the line scoring for offensive zone deployment while Galiardi gives strong underlying numbers allowing the line to take more meaningful minutes and zone deployment. Peluso and Thorburn struggle to provide on-ice value but do offer "intangibles" and "toughness"; your subjective value pending on those things. Outside of Halischuk's scoring and Galiardi's decent two-way play, the fourth line's primary value extends from Slater's face off capabilities. Because of this, you'd want to avoid deploying this line in the neutral zone, where a face off win has less impact.

Extra Thoughts on the Future of Hockey Analytics

You may notice Ladd's results are very similar to that of Bryan Little. This brings up one area in analytics that needs to be addressed further. WOWYs work for separating players but difficult when you have guys like Little and Ladd who have spent over 75% of their icetime together.

The future in hockey statistical data is in finding ways to determine who is moving the needle earlier than Corsi currently is able to. Corsi's largest advantage over goals is that it becomes relevant far quicker than goals do in determining who is helping your team outscore more. These looks into microstatitistics help us determine why and how you can optimize your roster, but also find who is pushing the needle and why.