5 ms·

Isn't this what Tufte refers to as "chartjunk"?

jweir 12y ago
Tufte points out 1 + 1 is 3 or more visually. The two lines create two elements, then the space is another, then the space on either is potentially another. http://csarven.ca/one-plus-one-equals-three-or-more http://csarven.ca/one-plus-one-equals-three-or-more These textures are noisy. There can be occasions to use textures, but these examples would work just as well with shades of gray.
- ricsca 12y ago
  Interesting point of view. Choices of style apart, there are situations where textures are useful. Like for example the case faced by swayvil (see the comment), or when you need b/w printable images. Another good reference is "Semiology of Graphics" of Jacques Bertin, he has a good opinion about textures.
  - jweir 12y ago
    "Semiology of Graphics" is a great book, but this was published in the 1960s when those graphs where drafted by hand and color printing was probably not an option. I am ignorant as to what tools a draft-mans hand for shading then, but I am sure hatching would be easier than creating a fine range of halftone grays.
    - jacobolus 12y ago
      Here’s what Tufte says: > Can optical art effects ever produce a better graphic? Bertin exhorts: “It is the designer’s duty to make the most of this variation; to obtain the resonance [of moiré vibration] without provoking an uncomfortable sensation: to flirt with ambiguity without succumbing to it.” But can statistical graphics successfully “flirt with ambiguity”? It is a clever idea, but no good examples are to be found. The key difficulty remains: moiré vibration is an undisciplined ambiguity, with an illusive, eye-straining quality that contaminates the entire graphic. It has no place in data graphical design.
ncza 12y ago
Tufte is not that omnipotent, did he even ever consider colorblindness? Hatches are awesome substitution for color.
- jacobolus 12y ago
  Of course he did consider color blindness. Tufte recommends Imhof’s book Cartographic Relief Presentation on the subject of choosing colors for data graphics, and I’d second the recommendation. Or for something online, take a look at Cynthia Brewer’s work.
jacobolus 12y ago
Yes. See http://www.edwardtufte.com/bboard/q-and-a-fetch-msg?msg_id=00040Z http://www.edwardtufte.com/bboard/q-and-a-fetch-msg?msg_id=0...

jweir 12y ago

Tufte points out 1 + 1 is 3 or more visually. The two lines create two elements, then the space is another, then the space on either is potentially another. http://csarven.ca/one-plus-one-equals-three-or-more http://csarven.ca/one-plus-one-equals-three-or-more These textures are noisy. There can be occasions to use textures, but these examples would work just as well with shades of gray.

ricsca 12y ago
Interesting point of view. Choices of style apart, there are situations where textures are useful. Like for example the case faced by swayvil (see the comment), or when you need b/w printable images. Another good reference is "Semiology of Graphics" of Jacques Bertin, he has a good opinion about textures.
- jweir 12y ago
  "Semiology of Graphics" is a great book, but this was published in the 1960s when those graphs where drafted by hand and color printing was probably not an option. I am ignorant as to what tools a draft-mans hand for shading then, but I am sure hatching would be easier than creating a fine range of halftone grays.
  - jacobolus 12y ago
    Here’s what Tufte says: > Can optical art effects ever produce a better graphic? Bertin exhorts: “It is the designer’s duty to make the most of this variation; to obtain the resonance [of moiré vibration] without provoking an uncomfortable sensation: to flirt with ambiguity without succumbing to it.” But can statistical graphics successfully “flirt with ambiguity”? It is a clever idea, but no good examples are to be found. The key difficulty remains: moiré vibration is an undisciplined ambiguity, with an illusive, eye-straining quality that contaminates the entire graphic. It has no place in data graphical design.

ncza 12y ago

Tufte is not that omnipotent, did he even ever consider colorblindness? Hatches are awesome substitution for color.

jacobolus 12y ago
Of course he did consider color blindness. Tufte recommends Imhof’s book Cartographic Relief Presentation on the subject of choosing colors for data graphics, and I’d second the recommendation. Or for something online, take a look at Cynthia Brewer’s work.

jacobolus 12y ago

Yes. See http://www.edwardtufte.com/bboard/q-and-a-fetch-msg?msg_id=00040Z http://www.edwardtufte.com/bboard/q-and-a-fetch-msg?msg_id=0...