Maxima and minima
Once more, superimposing time series creates silly theories

Dampened by Google

Robert Kosara has a great summary of the "banking to 45 degrees" practice first proposed by Bill Cleveland (link). Roughly speaking, the idea is that the slope of a line chart should be close to 45 degrees for the best perception. It's not a rule that you see much on Junk Charts because it's one of those rules about which I don't hold a strong opinion.

Here are the examples given by Kosara:

The same data is presented three ways. The slope is a reflection of the scales used on the two axes.

Well, I lied when I said I didn't care. Look at this particular chart below:

Some of you may recognize this style... I'm imitating Google Analytics charts. Several of the other Web charting tools also seem to come up with gems like this. Pretty much every chart you see in the Google Analytics interface looks like a flat line. The chart above looks like nothing more than noisy data from week to week.

But then look at the scale! The leftmost part of the line is a rise over two weeks. The actual rise was 50% or 300,000, i.e. an earth-shattering change.

If you use Google Analytics, you are better off downloading the data to Excel and drawing your own charts.


Feed You can follow this conversation by subscribing to the comment feed for this post.


Can you truly look at this graph in isolation and state that the Y axis is inappropriate? Without context, a 50% rise in a value sounds like an earth-shattering change, but do we know that it's not normal variation for whatever value we're tracking? There are other changes of roughly the same magnitude in the same two-month period, including a steeper one that starts on 6/19.

The "bank to 45 degrees" recommendation always strikes me as a good starting point for analyzing unfamiliar data but a little tone-deaf for charts that will be updated and consulted frequently (such as a web traffic analytics dashboard). For familiar data sets, where you're aware of the typical magnitude and variation in the data and just want the current state at a glance, there would seem to be an argument for incorporating these norms into the design (for instance, by using a Y axis scale that includes more values than needed for the current view) in a way that depicts variation with the right level of drama.


Jeff: I agree with your general point that the scale should be set with some knowledge of the normal variability. However, that is not how Google or any of these online graphing tools do. I use Google Analytics every day so I can tell you most of their charts look like flat lines, and if I rely on those, I'll miss most of the signal.

The comments to this entry are closed.