File talk:Rational harm assessment of drugs radar plot.svg
Why isn't this a bar chart? This plot is very hard to read, about the only thing you can see is that heroin is the worst. But it's almost impossible to tell what drug is second most addictive; is it cocaine or is it tobacco?--Qvasi (留言) 12:47, 20 May 2017 (UTC)
- The polar format is more often used to graph cyclic time-series data, but I don't find it that hard to understand. For alternative graphings of the same data, see Category:Nutt et al's drug harm chart... -- AnonMoos (留言) 23:17, 24 May 2017 (UTC)
This is the worst visualization of this data imaginable.
[编辑]
This style of graph is completely inappropriate for displaying this type of data. The original scatterplot made illustrative sense. Why was this turned into a radar plot which makes comparison of the points extremely difficult? Why are the points connected? There is no meaning to interpolating lines into a discontinuous group of categories with no meaningful intermediates. Why display it in a circular fashion, which is harder for viewers to visually inspect and compare values? I could keep going but this is a textbook example of trying to make a graph look fancy but making a trainwreck out of it. There are reasons why there are rules for graph design, and reasons why the original Lancet article displayed the data the way it did.
- Absolutely agree -- I regenerated it as a bar plot, however, it looks pretty busy and I'm not sure it really any more efficacious in that form. I'll try uploading it now, as well as a python notebook (zip file, in case someone else cares to produce a better one). IMO, though, the data and methodology is a bigger problem, the whole thing is nonsense IMO. 2600:1012:A120:C72C:1974:81A6:2FB6:F199 21:22, 30 September 2023 (UTC)
