2016-17 Midseason Rainbow Charts

As the title says, these are the midseason “Rainbow Charts” I released on twitter back in mid-January. Yes, I am putting them up on the blog just now, mid-February. I’ve been busy, yo.

Anyway, much like the original rainbow charts, these charts compare players against the results of their peers who have similar TOI. So a second pairing defender is being compared against the historical average of all second pairing defenders, not against his teammates.

By using standard deviations, the idea is to provide some context around usage, as often we get hung up on decimal places. Especially when looking at relative zone starts, the deviations are quite large, so it takes a lot to move a player outside of the “average” zone (the middle square).

Weighted Corsi is a combination of CF60, CA60 and Relative CF% standard deviations, weighted.

Click “Read More” to get to the charts. Ctrl+F will get you directly to your team if you don’t want to scroll through everything.

Anaheim Ducks

1.18.17 5v5 Usage by Line - ANA.png1.18.17 5v5 Usage by Pair - ANA.png

Arizona Coyotes

1.18.17 5v5 Usage by Line - ARI.png1.18.17 5v5 Usage by Pair - ARI.png

Boston Bruins

1.18.17 5v5 Usage by Line - BOS.png1.18.17 5v5 Usage by Pair - BOS.png

Buffalo Sabres

1.18.17 5v5 Usage by Line - BUF.png1.18.17 5v5 Usage by Pair - BUF.png

Calgary Flames

1.18.17 5v5 Usage by Line - CGY.png1.18.17 5v5 Usage by Pair - CGY.png

Carolina Hurricanes

1.18.17 5v5 Usage by Line - CAR.png1.18.17 5v5 Usage by Pair - CAR.png

Chicago Blackhawks

1.18.17 5v5 Usage by Line - CHI.png1.18.17 5v5 Usage by Pair - CHI.png

Colorado Avalanche

1.18.17 5v5 Usage by Line - COL.png1.18.17 5v5 Usage by Pair - COL.png

Columbus Blue Jackets

1.18.17 5v5 Usage by Line - CBJ.png1.18.17 5v5 Usage by Pair - CBJ.png

Dallas Stars

1.18.17 5v5 Usage by Line - DAL.png1.18.17 5v5 Usage by Pair - DAL.png

Detroit Red Wings

1.18.17 5v5 Usage by Line - DET.png1.18.17 5v5 Usage by Pair - DET.png

Edmonton Oilers

1.18.17 5v5 Usage by Line - EDM.png1.18.17 5v5 Usage by Pair - EDM.png

Florida Panthers

1.18.17 5v5 Usage by Line - FLA.png1.18.17 5v5 Usage by Pair - FLA.png

Los Angeles Kings

1.18.17 5v5 Usage by Line - LA.png1.18.17 5v5 Usage by Pair - LA.png

Minnesota Wild

1.18.17 5v5 Usage by Line - MIN.png1.18.17 5v5 Usage by Pair - MIN.png

Montreal Canadiens

1.18.17 5v5 Usage by Line - MTL.png1.18.17 5v5 Usage by Pair - MTL.png

Nashville Predators

1.18.17 5v5 Usage by Line - NSH.png1.18.17 5v5 Usage by Pair - NSH.png

New Jersey Devils

1.18.17 5v5 Usage by Line - NJ.png1.18.17 5v5 Usage by Pair - NJ.png

New York Islanders

1.18.17 5v5 Usage by Line - NYI.png1.18.17 5v5 Usage by Pair - NYI.png

New York Rangers

1.18.17 5v5 Usage by Line - NYR.png1.18.17 5v5 Usage by Pair - NYR.png

Ottawa Senators

1.18.17 5v5 Usage by Line - OTT.png1.18.17 5v5 Usage by Pair - OTT.png

Philadelphia Flyers

1.18.17 5v5 Usage by Line - PHI.png1.18.17 5v5 Usage by Pair - PHI.png

Pittsburgh Penguins

1.18.17 5v5 Usage by Line - PIT.png1.18.17 5v5 Usage by Pair - PIT.png

San Jose Sharks

1.18.17 5v5 Usage by Line - SJ.png1.18.17 5v5 Usage by Pair - SJ.png

St. Louis Blues

1.18.17 5v5 Usage by Line - STL.png1.18.17 5v5 Usage by Pair - STL.png

Tampa Bay Lightning

1.18.17 5v5 Usage by Line - TB.png1.18.17 5v5 Usage by Pair - TB.png

Toronto Maple Leafs

1.18.17 5v5 Usage by Line - TOR.png1.18.17 5v5 Usage by Pair - TOR.png

Vancouver Canucks

1.18.17 5v5 Usage by Line - VAN.png1.18.17 5v5 Usage by Pair - VAN.png

Washington Capitals

1.18.17 5v5 Usage by Line - WSH.png1.18.17 5v5 Usage by Pair - WSH.png

Winnipeg Jets

1.18.17 5v5 Usage by Line - WPG.png1.18.17 5v5 Usage by Pair - WPG.png





4 thoughts on “2016-17 Midseason Rainbow Charts

  1. I love your stuff. if you would only write more about the Predators (I grew up in Dallas, but live in Nashville now). I’m struggling a bit to interpret the charts. Looking at the Pred’s, it looks like we have four defensemen who are being used like first pairing defensemen, and who are performing at the same level as first pairing defensemen league-wide. Looking at the forwards, what does it mean that the fourth line guys are in the bottom half of the chart? Does it say they are being used against tougher competition than average, or that they are performing worse than average?


    • You’re actually correct – the Preds use two pairs of Dmen so much that their last pair barely gets any minutes. The Winnipeg Jets are the same way, and it indicates that you’re a bit more insured against injuries that way (providing, like they do in Nashville, they play up to that level).

      Yes, the y-axis is relative zone starts, which can be a proxy for “defensive” or “offensive” usage in the extreme. Nashville has a very clear structure where they prefer certain players in defensive roles and others in offensive ones. Many teams, like San Jose or Winnipeg, don’t have these very specific roles.

      Liked by 1 person

  2. I love data, and the most important findings are typically the counterfactuals, situations where the data shows your eyes are deceiving you. So, according to the graph, Fisher and Jarnkrok are getting more defensive zone starts, and performing at about the league-wide average for fourth line forwards; that is consistent with what we thought before looking at your graphs. Riebiero was performing well, but getting mostly offensive zone starts, suggesting that his performance might have been inflated. Your graph shows Arvidsson is performing way above other second-line forwards, which is aligned with what we thought by watching. The Johansson rating–well above his peers–seems surprising. Based on points and assists he is below expectations. Does the dot location suggest that he’s been dominant but unlucky, considering that the other metrics show him as excelling? Or is the positive ranking in performance an artifact of his classification as a second line center? I think of Johansson as our first line center. Would his dot be in a different location if you had compared him with other first line centers? How far would that have moved his dot?


    • Data serves many purposes. Yes, it is important to suss out the things we don’t know, but it’s just as important to have what we know through subjective means (eye test) confirmed through objective means.

      No, Johansson’s Wt. Corsi likely wouldn’t change a lot if you bumped him up to a 1st liner – would be closer to the center, but still probably outside of average.

      Clearly these graphs don’t take points into account. This is merely a single view.

      Liked by 1 person

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 )

Google+ photo

You are commenting using your Google+ 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 )


Connecting to %s