clock menu more-arrow no yes mobile

Filed under:

Bottom Roster Players, Charts, Shot-Based Metrics, and You

Strap in for a lot of charts focusing on the Ducks' lower line forwards.

This is some nerd sh*t
This is some nerd sh*t
kid ish

This is going to be a chart-intensive post, so I will let all of that do most of the talking except to point out what I see. It will be long. Take many breaks.

There's a bit of haranguing about how signing (or potentially signing) older forward players is blocking the "youth movement" in Anaheim, so I wanted to look at a few of those young Ducks players. By and large, they are only an average group of players, and any team signing "known quantities" (even with some risk, a la Dany Heatley) is basically doing its job to ensure against that fact. There are a couple exceptional young talents here, and it doesn't seem like they are being blocked at this point.

A quick word on shot-based metrics again. I value them not just as statistical measures of what has occurred but also as a window into what a player is bringing to a team. The aim of hockey is to score the most goals in the game, therefore winning it. Goals are rare. Shooting a lot is less rare and is also the only way to make the rare goal occur, so any player that affects a team's ability to shoot more should be considered "better" at the aim of the game: winning.

There's more to hockey than shooting the puck, obviously. There's a wide variety of skills to consider. A player who doesn't help his team take more shots isn't "bad," per se, and could be providing value in other ways. But to me, the ultimate value a player can provide his team is "offense." This includes defensive-minded players, because they contribute toward offense more than points totals show. Offense isn't just goals and assists alone, because again those are rare events. Offense is shooting the puck, ie "being on offense."

All of that is to say:  the young players I am highlighting are by and large average (one or two above, most below) to this point in their careers as judged by shot-based metrics. What I mean is, regardless of the individual skills or finishing talents in question, none of them really help the Ducks' offense overall or at least consistently enough to be relied upon (yet). This isn't an analysis of why that is. This is more to say that even though we like these guys, they aren't putting the Ducks over the top or anything.

I am using data between the 2011-12 and 2013-14 seasons for all players, whenever possible. Some players don't go as far back, like Etem, or have holes between seasons. I'll try to note that in the text.

Kyle Palmieri

I did a post on Palmieri already, but this is further illumination on this player, in my opinion. Let's start with his forward WOWY (with or without you Corsi) numbers in chart form, minimum of 100 minutes played together. (How to read: blue is Palmieri in this case and player listed on chart together, red is Palmieri alone, green is player listed on chart alone.)


Here's why Palmieri's numbers can get clouded: his top four is split between the very best (Getzlaf, Perry) or the very worst (Bonino, Beleskey). In fact, his minutes with Bonino and Beleskey have done an awful lot to hurt all three players. I'll get back to this when I get to Beleskey.

His WOWYs with defenders, same minutes.


While his numbers with forwards suggest murkiness, the numbers with defensemen reaffirm my thought that Palmieri's defensive zone play is really shoddy. Defensemen tend to contribute toward "shots for" more than we might think, and forwards tend to contribute toward preventing "shots against" in the same way. Seeing all the defenders bump up without Palmieri on the ice implies a weakness in that.

How does that look overall, as a trend? Here's a look at his averages between 2011-14.


This chart is all of the data in each respective field (generation, suppression, etc.) added together with points production to illustrate the player's on ice trend. I like it because it lends itself well to cross-player comparison, as can be seen throughout this post.  [Ed. Note:  For those of you, like me, unfamiliar with this type of chart, I found this example of three different types of players - Corey Perry: good prevention/generation w/ great point production, Justin Williams: Great prevention/generation w/ decent production and Patrick Marleau: Almost perfectly balanced - illustrative and useful. ... and now back to your regularly scheduled Palmieri Analysis. -CK]

I am encouraged that his raw shot suppression numbers aren't wildly out of sync with his shot generation, although I would expect his generation to be much better. His raw Corsi is pretty crummy, but he's also played with some wildly fluctuating possession players and is not elite in that area on his own. Palmieri likely turns a corner as a player if he puts himself in a better position to generate more; with his finishing skill, he could become a solid and dependable depth contributor.

Lastly, here's a trend map of his Corsi relative to teammates against his zone starts relative to teammates.


As I noted before, his payout for cushy starts isn't high enough (yet) to have earned him more minutes from his coach.

Matt Beleskey

I like Beleskey's "game" (he skates hard and hits people) but have been working out how often I find myself questioning it by his numbers. On the forecheck, he's sensational. His neutral zone tracking appears solid enough, although he tends to either be too aggressive and behind his man or too conservative and well "behind the puck" defensively (prevents transition turnovers). Defensively, he leaves a lot to be desired. But that's pretty typical of a third or fourth line winger.

But he turns play toward offense at a worse clip than nearly anyone on the team. While I previously believed this had to do with who he played with, I am becoming more and more convinced that the problem is Beleskey himself. And that is precisely why stats like this matter: they reveal what we aren't seeing readily or don't want to believe once we see them.

Forward WOWYS, minimum 100 minutes.


Any player seeing this much time with Parros will have deflated numbers, and we see how much better Beleskey was away from the likes of him and Gordon. The real interesting thing here is his numbers with Palmieri. From Kyle's viewpoint, he's better away from Matt; from Matt's viewpoint, he's better away from Kyle. Two considerations for this: 1. each were primarily with Bonino when together, so when apart they were also apart from him. (And neither player worked with Bonino centering them.) 2. When apart, they were usually jumping up to the team's best players (Getzlaf, Perry, etc.) and not like-fourth line talent. Usually.

His defensemen WOWYs, same minutes.


Sami Vatanen is so, so, so stupid good. Anyway, same trends as Palmieri really, but more pronounced. It makes sense with this though.


Consider: Palmieri is younger, plays on the opposite side, has slightly better numbers, and has played with better linemates overall. (Not spending 100 minutes with Parros is definitively better.) Beleskey has only SEVEN fewer points than Palmieri since 2011-12.

Both players need to generate far more shots to be effective, but it isn't like the talent isn't there for either guy. What has hurt both is the line of them paired with Bonino. Looking at it from Bonino's side, it was largely Beleskey who dragged them all down. This is worth exploring in the future, I think. Who were the defensive pairs when this line was out? Who was on the left side with Beleskey? I'd have to do more work to check that out.

But if I had to guess, I'd say Beleskey is the player we have before us. In a limited capacity, in primarily a forechecking role, he can likely still provide some value. But he doesn't add strength to the best when put with them, and that's the key difference between he and Palmieri. Getzlaf and Perry haven't been strong numbers-wise with Beleskey, and his experiment there should end. (Boudreau was using him with the Twins against the Kings before he was hurt, so let's not hold our breath.)

That being said, this should reaffirm how largely average (or below) both have been. Perhaps time away from Bonino (and each other) will improve both players.

Here's Beleskey's possession relative to zone start trend.


Beleskey is Palmieri but without power play time and fewer P/60. Also: average. Palmieri is younger, however, and has more added value when put with better talent.

Devante Smith-Pelly

Devo's data is from 2011-12 and 2013-14, as he didn't play for the NHL team in 2012-13. That means there's less of it available than Palmieri and Beleskey. "Sample size" is real here because it tends to amplify numbers toward more extremes - the "regression to the mean" thing said about PDO is very true about these types of stats as well. Larger samples are always better.

Because there's less data for Smith-Pelly, I just dropped it all into one chart. It is a bit harder to follow, but that's life. This is a minimum of 60 minutes played.


Once again, <3 Sami. One thing that pops up with Devo and Etem (in a sec) is how good Steckel makes his wingers. (Steckel is very interesting because he literally makes wingers look sensational and defenders look terrible. His own raw numbers are below-average. I've heard it said that he plays an AHL game and just never translated it up to the NHL. That's likely why teams have signed him to two-way deals for forever. He's a sensational mentor to have but not a very effective NHLer.)

So Steckel and Perreault (who we know is great) aside, Smith-Pelly was more or less a drag on most players. He played well with Cogliano and Palmieri, two guys who play on the right side. If there can be a good connection between him and Palmieri, perhaps we'll see some better numbers from both.

I'm not overly concerned about any of it yet considering he's appeared in less than one season worth of games. Too soon to really judge him as a player, but we can certainly look at what's there to show that his minutes have been largely below average for the Ducks. Both are true things.


Again, sample size and all that, but hopefully the trend of his suppression being so godawful ceases. One thing Devo said he worked on with the Admirals is playing harder in all three zones, so going forward this should normalize some. We look at his playoff production and see hope, but let's not forget that he was also massively inconsistent during those games. He was negative one game, positive the next two, negative the next two, so on. His scoring of the clutch goals isn't maturity to his game so much as luck, which is important to have in sports. But counting on it to continue is foolish.


Here's a silver lining to Devo's numbers: while his trend line is still negative, it gets noticeably better the nicer his starts got. There's a bit of optimism in thinking he can succeed in friendlier minutes and the right linemates in the future.

Emerson Etem

Etem has slightly fewer games than Smith-Pelly from which to pull, so everything there applies here. Oh and also Etem played 2012-13 to 2013-14, so that's a little different. This is over 60 minutes together.


You can see the Steckel effect again and also another curious thing: Cogliano played well with this player. Sort of separately, I looked a little more at Cogliano because he's made an effective pairing with a surprising amount of players, especially other forwards. The only player, so far, that Cogliano didn't make better when they were together is Beleskey, who like I said is probably the drag on possession to begin with.

Also once again, SAMI IS THE BEST.

But small sample size.


This chart is the reason I actually split out Etem from Smith-Pelly. There's a smaller gap between his suppression and generation numbers that give me a happy thought. That's a much better trend to recover from than Devo's, which is more lopsided. If Etem can learn to play a similar game to Cogliano with his speed, I think there's a lot of positive he can bring.

Again, like with Smith-Pelly, while we aren't judging Etem as a player yet we can say his additions to the roster to this point have been largely useless in terms of consistent offense.


The second reason I split Devo and Etem. While there's a better suppression trend in Etem, he hasn't rewarded his coaches with better offense relative to zone starts over his teammates. That kills team-wide possession numbers, because those starts need to go to forwards who can generate better when they start closer to the other net.

Let's get off these scrubs and get to some heroes though. I'm tired of negativity.

Patrick Maroon

The big dog. He actually doesn't have a ton of data points. He's basically gotten Etem-esque minutes/games from 2012-13 to 2013-14. His first year was pretty sketchy, but then last year he turned into a totally different player who is amazing and grows great facial hair. (Also, Lindholm ending up in the forwards chart is an authorial oversight so whatever.)


Perreault and Maroon had just about one of the best all-around pairings I've ever seen for the Ducks, to be honest. Getzlaf and Perry are the only ones I've seen that's similar or better, in fact. It really is a shame the team let Perreault walk, I feel like there was a great match there.

At the same time: look at how good Perry and Maroon played together. LOOK AT IT. Perry makes a lot of people better, namely Getzlaf, so this isn't really a new thing. Adding Maroon to that top line makes them really, really big and good. So if Perreault wasn't qualified because the team wants Pat up top more often, that's ok by me thank you very much.


Hey look who is better away from Vatanen and also made Vatanen better? It's Pat! One interesting thing to note here, largely on guesswork and usage from my end: the WOWYs with Fowler tell me he's less effective in the shutdown role Boudreau fed Cam last year. So Maroon's not really here to be a checking line guy, which could spell trouble if we see him in that role with Kesler. (It may not, however, because Kesler is also very good.) This is ok if Maroon makes the top line a Superstar Elite Awesome Offense Line though, so who cares.


Look at that: a depth player who generates more shots than he allows when on the ice. This is terrific. Think about how much better he will be in suppressing shots when his other winger isn't a 40-plus retiree who consistently cheated the zone and ditched his line defensively!


A proper trend of possession relative to teammates relative to zone starts. Two things influenced this positively that weren't just Pat being great: 1. Selanne was very effective when starting close to the other net in terms of generation. 2. Perreault was extraordinarily effective at generating shot attempts off offensive zone draws, won or lost. On won draws, Maroon was either going straight to the net or to the wall to cycle the puck, so he played his role. On lost draws, and this is the key, Maroon was a hound in preventing a clean breakout and still generating offense for that line.

I bookend this with the same thing I start Smith-Pelly's section with: a small sample shows extremes a lot. So there's still a gamble in Maroon that he's not as good as he was last year. That's all valid and why I don't go gaga over Maroon. Unlike...

Jakob Silfverberg

I included him in here not because his minutes really fit, as he was decidedly playing second line minutes a lot last year. He's here because his games played is very similar to our first two players. HE'S ALSO HERE BECAUSE I LOVE HIM.

I split his WOWYs by team, the Senators in 20-whateverwhocares and the Ducks last season. Let's start with the Senators since most of you know nothing about them and rightfully so because they're assholes. Over 60 minutes here, since it is split by team.


Here's what you want to take from that chart: Silfverberg was a positive player on a largely positive possession team. He also made a lot of really good players even better. For example, when defensive wunderkind Karlsson was on the ice with Silfverberg, other teams didn't touch the puck. The real data points you want to keep in mind are his numbers with Turris and Zibanejad, two extraordinary centers. Both are better than Silfverberg in raw open play possession, but he didn't really drag either one down in a noticeable way. That's very important.

Now, the Ducks.


Love love love love love love. (I am singing this, you see. Sing along. Make up your own tune.)

Big difference? The team around Silfverberg wasn't as good at creating offense. And yet Silf is there, standing tall like a possession monster, and I love him.

Silfverberg was at his worst when paired with Koivu, which makes sense because the latter fell off a cliff after his injury. (Fowler and Lovejoy played behind the Koivu line the most, which explains that connection.) The same is true of his play with Cogliano, which was also on the line with Koivu. Here's where I try really hard to stay brief...

I think there was an intentional coaching shift after Koivu was nearly murdered in Columbus [Ed. Note: by his own stick -CK]. The way the team opened the year was hot fire, fast action, and tons of possession and corresponding chances. I believe that injury tamed Boudreau's system because he wanted to protect his players but also because losing the up-until-then very productive (offensively in terms of shot control) Koivu hurt his depth at a key position. The team set up shop defensively much more often, slowing the pace down, blocking more shots but not retrieving pucks as well and sending play up the ice as fast.

There's some logic to this too. In raw percentage terms, the longer a team holds the offensive zone, the lower their chances are of scoring off a proper scoring chance. (Scoring as a random event increases because the puck stays closer to the net.) Remember, the correlation to possession but also to scoring chance tracking is why we like shot-based metrics. That's the real reasoning behind coaches saying "push them to the perimeter." It isn't so much that there's a designed system to do this so much as the natural occurrence of 10 skaters in a small space eventually pushes the puck out to the sides. What high event teams (like Dallas or Ottawa) do is retrieve it quickly and play more of the game off transition. This means less bodies in the zone, which also means the puck can slide through it easier and open up goalies. Low event teams (Los Angeles or New Jersey) dominate the zone with a lot of body movement and perimeter passing and shots from everywhere in the hopes that eventually something happens. One day I'll get back to this topic.

Boudreau has hinted before that he's very aware of "fancystats." Whether he brings them to his staff or the room, I've heard him mention them in a few interviews and quotes. I've also seen him do things that indicates he gets the principles of what it all means. Zone time is good but less important than chance creation. The prevention of chances against is more important than giving up the zone. And his systems off faceoffs is absurdly smart. What he sees in hockey is offense-minded, so it stands to reason that his principles are aligned with shot-based metrics.

Going back to his Washington days, he took a dominant possession team and tamed it trying to play more cautious. Since being here in Anaheim, he's improved the team's underlying numbers but not to a significant degree, and last season suggested it's by intent. The Ducks roared out of the gate and were killing teams with the puck. That Koivu injury started a very consistent plummet in possession, and I think he played it safe by giving up the zone and by locking down the middle (with bodies) as a focus. Go watch the Anaheim-St. Louis game immediately following the Olympic break to see this in action the most.

All of this is to say, I believe Boudreau saw a strong possession player in Silfverberg and intentionally paired him with the struggling Koivu in attempts to keep that line afloat. And privately, now that the older veterans are gone and some young blood and legit second line talent are here, I hope Boudreau opens things back up and lets the Ducks fly the way we saw them start the year (or when trailing in games).

When Jakob was away from Saku, Cam, and ol' Benny Love, he was more or less the same strong player that his career numbers show. He played well with Bonino and Perreault and even brought some respectability to Beleskey's possession game. That takes talent.


Angels are now singing with me, a heavenly tune. Where Silfverberg goes, shot generation follows.


Start him closer to the opposing net, more shots will happen. Yay!

If anyone is deserving of a slot alongside Kesler, Silfverberg and Cogliano top the list in my mind. But it wouldn't shock me if Bruce uses Jakob to bring along any younger centers if needed, like oh say my favorite Rickard Rakell. The Ducks should give Silfverberg a ton of money, by the way.

This is getting out of hand but I have one more just for comparison.

Nate Thompson

He's a fourth line center. I have no idea what he'll bring but I'm not extraordinarily hopeful he'll be turning play with any degree of success. Still, depth at center is depth at center, and the organization wants that right now.


He slowed Martin St. Louis down a lot, but that happens when two weaker possession players get paired up.


The Hedman and Bergeron pairings intrigue me because Anaheim has some comparable defensemen in terms of playing style. So there's some hope Thompson can at least hold it down in some instances.


Again, below average but probably not completely awful.


He's had a lot of unfriendly starts in his career, so there could be some easier explanations for his iffy numbers. When he's given friendlier starts, he does just ok. Again, he's a fourth line center at best.


Last chart. This is everyone's trends together. Why not. Looks pretty.


Really, only Silfverberg and Maroon (in a small sample) show real quality as depth players to this point. As long as guys like Thompson or Heatley aren't blocking them, there's really no issue in my mind. Thompson provides flexibility at center and Heatley has several 50-goal seasons that none of the Ducks youth have. Is he old? Yup. But 50 NHL goals in one season shows really friggin' good hands, and he's done it more than once. On the cheap, he's a risk worth taking and a slot worth giving up, if he can keep up.

A player like Palmieri has certainly earned a right to play a full season but hasn't showed anything that demands a larger role yet. I feel that, of all the young wingers, he's the closest to turning the corner. If he does, there's no reason he can't slide around third or even top out at second line minutes.

Someone like Beleskey may be nothing better than a utility fourth liner who can jump up and down the roster, especially if he can't carry any offense. He does have pretty respectable points totals for that role. Can he improve? Possibly.

Smith-Pelly and Etem should probably be competing against each other for ice time at this stage. If they both get in, all the better. But nothing suggests that either one or the other sitting out is killing the team, depth-wise. If the playoffs Smith-Pelly is the real deal, he's got the edge in my mind for earning a slot on this roster. If he ends up upgrading what Beleskey is, I'm ok with it.

So all in all, I'm tired of this already, goodbye now!