Hacker News new | past | comments | ask | show | jobs | submit login
Ask HN: What's so bad about 3D graphs?
2 points by cbjj on Oct 6, 2016 | hide | past | favorite | 6 comments
With the availability of different 3D technologies like OpenGL, WebGL, CSS-3D etc., one would think that 3-dimensional graphs would be more popular visualisation technique today. But it is extremely rare to come across any, except for certain special visualisations.

I would like to know why you think that is:

1. Is it due to the unavailability of popular 3D-graphing libraries/products?

2. Is it due to the inconvenience of interacting with 3D-graphs, like awkward to read the labels or zooming in etc.?

3. Is there something inherently better in 2D graphs?

4. Any other reason?

Would love to get an opinion on this (or any pointers to relevant resources).




There are some good resources for designing two dimensional presentation of data, e.g. Tufte. That quality of resource is rare when it comes to designing 3d presentations, in part because the field is new and perhaps because presenting 3d data on a screen or print collapses into a series of 2d presentations plus designing a UX for an interface.

Presenting in two dimensions draws on centuries of knowledge and practice. That's probably where to start.

Good luck.


If the data is 3d, no objection. If the data is 2d then a 3d plot is redundant, usually ugly and prone to mislead (see "How to lie with statistics" and anything by Tufte).


Could you explain "if the data is 3D"? Isn't it about how a presenter carves out the data for the presentation?

To explain my pov, I see that even if a multi-dimensional dataset is available, people still prefer to present a 2-Dimensional analysis of it.

And this is what I don't get. Is it due to the the lack of tooling or is it the fact that 2D graphs are just easier to grasp?


I mean: the independent variable is 2d and the dependent is 1d, like f(x, y), that's 3d data. By 2d I mean a pie-chart, which you _can_ display in 3d (by making a "thick pie"), but it is superfluous and detracts from the data.

I think 2d graphs are easier to graphs _are_ easier to grasp, and easier to make. Not sure tooling is the main issue, there's plenty of tools around to make 3d plots.


Perspective issues: pie charts are a problem. 3D elements are more of a problem.

I guess.


Could you elaborate on "3D elements are more of a problem". Do you have an example in mind, where you see a problem?

For example: The NYTimes ran this article (linked below) with a 3D chart, built using WebGL. I wonder why such visualisations are not more common?

Article: http://www.nytimes.com/interactive/2015/03/19/upshot/3d-yiel...




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: