« Light entertainment: pie charts to the rescue | Main | Try a new way of learning dataviz; course announcement »


Michael Schettler

I like the version with lines because it makes explicit that the USPS prices are more driven by distance (within a weight segment) than UPS and Fedex.

The slope of the USPS segment lines varies directly with weight category and with distance within each weight category. The relationship between categories for UPS and Fedex doesn't seem as linear and within categories there doesn't seem to be much of a trend.


I think the lines, in *this* case, make sense since one expects distance to be a continuous variable, and they show trend over that variable.


Nate: Maybe I will one day convert some of you :) Be careful that the increments of distance are not uniform here.

Jon Peltier

People are altogether too concerned about drawing a line to connect points that may or may not represent a continuous variable. I'm more concerned that each data point has a marker. Then the lines are allowed to be less formal. The lines can be a Gestalt thing, and not purely an indicator of actual data.

Jörgen Abrahamsson

I like the original best here.
Nice touch with the lines as movement from old light blue to new darker blue dots.

There is also the possibility visualizing all the data. A scatter plot of weights and distances with price as the size of the dot.

Ben Harris-Roxas

I'm afraid I think the original chart illustrates the story best too.

The comments to this entry are closed.

Kaiser Fung. Business analytics and data visualization expert. Author and Speaker.
Visit my website. Follow my Twitter. See my articles at Daily Beast, 538, HBR.

See my Youtube and Flickr.

Book Blog

Link to junkcharts

Graphics design by Amanda Lee

The Read

Keep in Touch

follow me on Twitter