I’ve got Dangerous Fenwick WOWY visualizations!

A few days ago, I completed a project I’ve been working on for a while now – DangerFen WOWYs.  That’s “without you with you” data, a very  useful analytical tool.

I’m pretty sure this is the first ‘danger’ style metric that has WOWYs available, so that’s kinda cool.  We may have a sh*tty team, but we’ve got state of the art tools for deconstructing the exact ways in which it is sh*tty!  Exciting, right?

Now if you’ve taken a look at this data, you know that it’s an enormous bunch of numbers, and while that gives utility and flexibility for analysis, it’s not exactly easy or convenient for a layman to parse.

So I’ve put together a visualization to make this data more accessible.  I’m going to explain how to read the visualization, and deconstruct it on an incredibly good player, after which you’ll get an enormous pile of graphs for every member of the team to peruse at your leisure.

Before I do that, as an aside, if you do want to work with the data yourself (or even just see what that data looks like), you can download the most recent versions from these links (all files are in CSV format):

This version of the output now includes the teammate’s position (D, C, L, etc), so you can filter down to a specific player vs specific roles (e.g. D partners) really easily.

So let’s get back to the visualizations, and start by taking a look at the graph of that incredibly good player, name starts with T and ends with aylorHall. [click the chart to embiggen]

So how do we read this?

1. Reminder that the key underlying data here is Dangerous Fenwick, which is unblocked shots, but weighted by shot type and distance. So, a 20 ft slapshot is more dangerous than a 50 ft wrist shot, so it counts more (for or against). So read it like you would a Corsi%. The higher the better. Over 60% is elite, and below 40% is approaching AHL quality.

I cut the charts off at 70% and 30%, but a few player combos exceed the limits (usually because of small sample size), and I just put a symbol at the top and the bottom to show that. Setting the chart limits wide enough to accommodate those rare instances just lowers the display resolution for most players.

2. The WOWY chart shows you how a player does with respect to 5v5 Dangerous Fenwick with his 10 most common (by 5v5 TOI) non-goalie teammates.

3. The vertical position of the black dot (“P+T”) shows you how they did together.  Take a look at Taylor’s dots. Notice how he drags every single one of his teammates above 50%.  This is really incredible.

4. The vertical position of the green dot (“P-T”) shows you how Taylor did without the other guy.  Again, notice how Taylor is above 50% every single time.

Notice in some cases his green dot is below the black dot (meaning there’s some chemistry with that player, because they’re better together, like with Drai and Teddy), and in some cases the green dot is above the black dot (like with Sekera or Nuge), meaning Taylor’s pulling those guys up, he’s not doing as well with them as he does with others.

5. And finally, the vertical position of the brownish red dot (“T-P”) shows you how the other guy did without the player.  Notice how almost every single player save Brandon Davidson is below the green and black dots, and below 50%, without Taylor Hall.

I’ll say that again: notice how every single one of his teammates save Brandon Davidson is worse off, and below 50%, without Taylor Hall, and better off and above 50% with Taylor Hall.

We may have a lot of sh*t on our team, but Taylor Hall is a legit superstar folks. 1000% legit. Don’t let anyone tell you anything different.

6. So now put all this together.  By looking at the pattern across all 10 players, you should be able to get a pretty clear picture of which players are zooming their teammates, and which are being zoomed.  Who’s dragged up, who’s dragged down.  Who’s got chemistry and who doesn’t.

7. One note on the jumble of numbers at the bottom of each column. That tells you, in minutes, the ice time together (P+T), the ice time for the player away from his teammate (P-T), and the ice time for the teammate away from the player (T-P). It’s there for completeness, so you can get a sense of how much data is driving the numbers you’re seeing.

LOTS of really valuable information packed into this little chart.  I’m fully open to suggestions for improvement, but so far I’m delighted at this first cut, and I hope you find them useful as well (I’ll be updating regularly just as I do with my player dashboards).

Now on to the rest of the team, sorted alphabetically by first name for ease of search.  If you want to see the anti-Taylor Hall, check out Lauri Korpikoski’s chart.

Adam Clendening

Anders Nilsson

Andrej Sekera

Andrew Ference

Andrew Miller

Anton Lander

Anton Slepyshev

Benoit Pouliot

Brad Hunt

Brandon Davidson

Cam Talbot

Connor McDavid

HE’S JUST A TEENAGER AND HE’S ALREADY ACCELERATED TO LUDICROUS SPEED!

Darnell Nurse

Eric Gryba

Griffin Reinhart

Iiro Pakarinen

Jordan Eberle

Jujhar Khaira

Justin Schultz

Jultz is a mixed bag, as you should expect. His defensive gaffes enrage us, but it’s important to remember he does make certain things happen, and has some good chemistry with some of the top players on the team. I’m pretty sure that’s why he’s gotten so much rope from every one of his coaches.

Laurent Brossoit

Lauri Korpikoski

MY EYES! MY EYES! HOW DO I UNSEE THIS?!?

Leon Draisaitl

Luke Gazdic

Mark Fayne

Mark Letestu

Matt Hendricks

Nail Yakupov

Nikita Nikitin

Nikita remains the wrecker of fancystats.  I still have no idea how his numbers can look so good and he can look so bad on the ice.

Oscar Klefbom

Rob Klinkhammer

Ryan Nugent-Hopkins

Tough season for the Nuge, whose chart usually ought to look like Taylor Hall’s, and this year looks more like Korpikoski’s steaming pile.

Teddy Purcell

Zack Kassian

Advertisements

11 thoughts on “I’ve got Dangerous Fenwick WOWY visualizations!

  1. One suggestion for you, G: lose the seconds from the “jumble of numbers” at the bottom of each column, & round to the nearest minute. That’ll cut your characters pretty much in half, because you’ll lose a bunch of colons as well. Will barely make a flicker of difference on the precision of your data while being a lot cleaner for the user.

    Like

    1. Thanks G I have been patiently waiting for you to explain all your numbers. This is incredible. Thanks for the work. The only thing I can think of in your quest to find out why Nikitin looks so ugly on the ice versus his stats….Brain Farts….HUGE ONES ! Other than that I’ve got nothing. Be careful following Bruces advice above. You want to be very careful with your colon. Hang onto it as long as you can.

      Liked by 1 person

  2. Nice stuff G. One thought… Is it possible to have two rows for each player? Maybe top 9 forwards and top 6 D?

    I recognize it’s a balance between easy readability and more information, so I’m not sure my suggestion is better but I do think it helps with an “apples to apples” comparison. You might get a better idea of who’s zooming who if you can easily compare the WOWY for forwards and the WOWY for D separately for each player.

    Like

    1. He was actually doing his usual ‘saw off against the best magic trick’ until he got sick. He missed a day but was never the same. He fell off a cliff at that point. Check out my “Player Dashboard” article and you can see Nuge’s cliff clear as day. Mono maybe? Ebola? Anyway, I hope when he gets back, both his hand and his Nubola are cured!

      Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s