Packed Bubbles Finally Make Sense
I couldn’t understand it when Tableau Software introduced its packed bubble chart in version 8. It’s a useless form of display, assuming you care about the data. Today, however, I discovered what Tableau must have had in mind when they added this chart. The following example appears in an article that was published today about Tableau’s annual customer event, which is currently taking place in Austin:
Clearly, my inability to recognize the value of packed bubbles was a failure of imagination.
Take care,
12 Comments on “Packed Bubbles Finally Make Sense”
Wait, so you’re saying that modern innovations in datavis are devised, foremost, to be used as marketing assets?
Outrageous!
Andrew,
To clarify, I am not claiming that Tableau created this image. I have no idea who created it. It’s possible that Tableau’s marketing department is behind it, but also possible that someone outside of Tableau created it. The article that I read did not cite the source of the image. The article was reporting on Tableau’s annual customer event, so it’s likely that image was presented there.
Hi Stephen:
Not sure if you’re aware of this project http://www.worldpotus.com/#/ by Google.
It heavily relies on bubbles to represent search behavior in different regions of the world.
Would like to read your thoughts about it.
Regards,
Hicham
Hicham,
I don’t see any packed bubble charts in the Google example. Packed bubble charts merely pack a set of bubbles on the screen, which vary in size and color to represent values, without any meaning associated with their positions.
@Stephen
Regardless of who created the image above (as I’m not sure it matters), the double-meaning in my comment was deliberate.
Hi Stephen:
After clicking on either of the candidates you should see bubbles with different sizes and colors. Each bubble represents a geographical area and its size is proportional to its Google trends.
try opening the site on an iPhone, Safari browser on a Mac or using IE. I had the same problem with Chrome where nothing appeared.
Regards,
Hicham
Hi Hicham,
You’ve misunderstood my point. I see the bubbles, but they don’t constitute a “packed bubble chart,” which has specific characteristics. In a packed bubble chart, neither the positions of the bubbles nor their position relative to one another has meaning. Positions are random. The algorithm merely places bubbles to optimally fill the space. In the Google example, the bubbles are not packed to fill the space and their relative positions appear to have meaning (e.g., with the largest in the middle).
Regardless, it is best to reserve the use of bubbles, which encode values by varying in size, for those occasions when neither 2-D position nor length can be used to encode values. In the Google example, bars would work much more effectively.
Steven,
Here is an example where someone from the Tableau community is doing a makeover of a packed bubble chart and removed the chart. http://www.vizwiz.com/2016/11/popping-bubbles.html
The original designer of the packed bubble viz responded that packed bubbles is a essentially a ‘point of view’.
In my opinion packed bubbles fail to ever communicate or support an actual point. I don’t understand a situation where packed bubbles tell me something other than ‘big and circular’. To me data visualization is more along the lines of industrial design. If I design a chair that is uncomfortable, easily breaks, or somehow catches on fire then it is not a point-of-view. It is a failure of design. Even in pure art, if the artist fails to communicate their point of view to a majority of the targeted audience, then in my opinion the artist has failed in their attempt with that piece of art.
Periodically someone finds an image of a holy figure in their morning toast, the lines in their rotting apple, etc. Inevitably you hear about them getting offered ludicrous sums of money for their “holy” artifact.
That said, I find the image above quite inspiring, though clearly a ‘fake’. I wonder how long it will be before a serendipitous packed bubble chart Virgin Mary sells for 10k on ebay, and people flock to find the hidden meaning in the data that underlies the image.
I find the mere discussion of bubble charts as more evidence that part of the world of data visualization and analysis has entered what I consider the Jackson Pollock stage. Throw it out there and some will rise to the results with huzzahs aplenty.
Hi Stephen,
I use bubbles chart for explore qualitative data from a survey.
I use size from answer text large for bubbles size, and use color depending user valuation.
for me, it’s a good way for explore a lot of answers from an open text field.
https://public.tableau.com/profile/karalata#!/vizhome/4Qdemo/Taskcompletionrate
what do yo think? it’s useful? how do I can improve it?
thanks a lot!!
Jaume,
Perhaps you could begin by explaining what your packed bubble charts reveal. Once you do that, I’m confident that we can suggest better ways to present this information.