top of page

The Perplexing Appeal of POTATOKRIEBEL

Potato Mountain by Flickr user Johnny Jet, used under CC 2.0 Generic Attribution license.

Sometimes, paradoxically, the best results can be born out of pure laziness. This was the case for my most recent contribution to the epic Makeover Monday project, now in the 9th week of its second year.

In Week 8 of 2017, the data set provided for remixing was “EU Potato Sector.” As inherently thrilling and intriguing as that data seemed, I was fully engaged with the Hacking Open Data project, which took up most of my vizzing time outside of work. After completing that, I had some real-life commitments pop up, and I just couldn’t find enough time and motivation to complete a Makeover that week.

Now, I use the term “complete” intentionally, because although I participated in MMs last year, I did so only on occasion. So far this year I had been 7-for-7, and was beginning to become habituated to producing something weekly. I thought that it might be possible to keep up with that pace for a substantial part of year. Even if I couldn’t go 52 straight, I still wanted to participate. I hoped not sporadically.

But here it was, Week 9 upon me, with Andy Kriebel bravely choosing to share his personal Amex spending habits with the world. I decided, well, maybe there’s a loophole. I know that last year, the Andys Two decried the trend of participants seeking out additional data sources to include in their Makeovers…but what if the “additional data source” was really just last week’s data source? Then it would be like doing weeks 8 and 9, but with the added bonus of only having to complete a single viz. Sweet!


I even had the overall design in mind (like a mental whiteboard) before starting, because the data sets were relatively sparse. I’d have three columns, with the third being slightly wider.

Column one: BIG ASS NUMBER OF ANDY’S SPENDING with clickable buttons to filter by spend category.

Column two: COST OF POTATOES IN SELECTED COUNTRY with useless map showing selected country.

Column three: PIC OF HAPPY ANDY AND POTATOES with amount of spuds his Amex spending would have funded.

Originally the idea was to show how many potatoes Andy could buy with each category of spending on his Amex card, but when I tried to link the data sources together by year, it just didn’t seem to work out…TRAGICALLY the potato data didn’t include 2016! FOILED! (Like a baked potato!)

So instead of taking a thin slice of potato data (I was already peeling away the production volume data, because I didn’t care, or see how it would fit into my intended viz), and instead of choosing a random year, I decided to allow the user to pick any valid year and country from 2004 through 2015.

Also originally, I was going to create custom shapes to show various amounts of potatoes (a single potato, some scattered around, a small pile, a larger pile, then a giant pile) based on the calculated result in column three, but when I built the viz I discovered that the shapes couldn’t get large enough to provide me the visual effect I wanted.

I also couldn’t link to the image through a URL box in the dashboard, host the various images on Dropbox, and get the calculated value to change the URL. It might be possible but I couldn’t figure it out so I changed tacks pretty quickly. Kill your darlings, they say.


Anyway, after a few hours of fiddling I got the two datasets to join properly, got the calculations to run the way I wanted to, and even achieved some of the effects I hoped for (like, getting the 8 categories of spending to render as four rows of two columns, with one smooth color ramp, as a filter). I did have to seek out a different data set, unfortunately: the potato data listed prices in EUR, while Andy’s spending was in USD, with each transaction tied to a specific day in 2016. I found a historical list of EUR-to-USD currency exchange rates for 2016, and added a column that determined how much each of Andy’s transactions would have cost in Euro on the specific day they were made. Then I could calculate how many delicious starchy taters could have been had instead.

Here’s the final result:

Let me reiterate: I had a few main goals here.

  1. Get a two-for-one deal in terms of finishing two weeks of MM projects in a single viz.

  2. Use big-ass-numbers, since Adam Crahen and I were just talking about that last week. (Tragically I could not come up with a way to include a 300-slice pie chart but I remain determined to do so in the future.)

  3. Include references to previous issues in this year’s MMs (the use of maps for no reason; the proper attribution of images; the validity of the data).

  4. Deliver a good-looking and patently absurd final product.

I do believe I succeeded in these goals, but what was most surprising to me, once I posted the viz to Twitter, was how popular it was compared to other MMs I’ve posted. The Trump’s Tweets one might have had slightly more engagement but this one was pretty popular, way beyond what I would have expected from something that seems really simple at first glance.


My takeaway is that people like one, some or all of the following:

  • Silliness.

  • Big-ass numbers.

  • Potatoes.

  • Pictures of Andy Kriebel.

I hope to provide the Tableau Public community with much more of these (well, at least the first two) in the future. In the meantime, I think it’s time to celebrate with some well-earned chips.


Recent Posts

See All
bottom of page