4 ms·
Using square areas as a bar graph feels kind of frustrating. Doesn't feel intuitive? I know I'm being super nit picky.
by Waterluvian 8y ago
Using square areas as a bar graph feels kind of frustrating. Doesn't feel intuitive? I know I'm being super nit picky.
- projectramo 8y agoI think they used it because, in that little space in the middle of the article, vertical space is more precious than horizontal space. So they traded off a little of one for a little of the other. Why does it feel less intuitive than regular bar graphs? I thought the criticism of bar graphs was exactly that people substitute area for height so the visuals are misleading.
- Waterluvian 8y agoI think I have a harder time visualizing relative sizes on two axes than one. I never thought of your point before though.
- tomp 8y agoIt could be a horizontal bar graph, no?
- projectramo 8y agoyes, good point.
- TeMPOraL 8y ago> I thought the criticism of bar graphs was exactly that people substitute area for height so the visuals are misleading. The usual criticism is that it's harder to compare areas than lengths, so one should avoid mapping data to areas.
- probably_wrong 8y agoI think it's not just unintuitive but also slightly misleading, because you have to be aware that surface increases exponentially. The grey square, for instance, has 1.4 times the surface of the red one, but the difference looks smaller than that.
- todd8 8y agoGraphics are so often intentionally manipulated to convey the authors intent. I don’t like graphs that show large differences until I realize that the origin of the graph is at say (0,1000) instead of (0,0); this can make a change from 1001 to 1010 seem huge. (One little edit to your comment, I think you meant quadratically instead of exponentially.)
- aoeusnth1 8y agos/exponentially/quadratically/g
- bertil 8y agoI like it when you express a different concept in each dimension (typically, likelihood and amount of a payoff) but in this case, it’s odd. You can have horizontal bars if you are tight for space.