Fear and Moaning for Las Vegas
My thoughts on IronViz following the first qualifier
As a Tableau Zen Master I feel a certain responsibility to perform a few essential duties. Several aspects revolve around acting as a general cheerleader for newcomers to our community, helping people and continuing to innovate and share ideas and tutorials. Another important aspect is to share concerns or point out where I think there are issues with the way we as a community work, share or play.
This aspect is particularly hard when those concerns might go against popular opinion, or be seen as attacks on people, or their work, or dissent against Tableau itself. Nevertheless, I take my responsibility seriously and so I write this blog post to start a discussion around the latest Iron Viz qualifier. My aim is to prompt discussion and perhaps see if we can improve the competition/entries; I have nothing but respect for the participants, judges and Tableau staff who help make Iron Viz what it is.
A Deserving Winner
First of all, I want to heartily congratulate Joshua Smith on his excellent entry and well-deserved win, Josh put an incredible amount of effort into his visualisation and it showed. His design and analysis were unparalleled across the entries and his win was the least surprising in Iron Viz qualifier history in my opinion. Likewise congratulations to everyone who entered, and especially those who made the Top 10.
The competition this time was very different with not just a theme but also a dataset provided, making the competition fairer and giving a simple way to compare the entries. It was tremendous fun to enter and while lots of work, and not necessarily my first choice of data or theme, it’s always a pleasure to compete.
Rise of the Long Form Dashboard
What disappointed me as I looked through the entries was the sheer number of “long-form” dashboards, including my own.
The rise of the long-form is a trend that has grown in Iron Viz over the last couple of years (my colleague @Scribblr_42 wrote about it back in 2017) and this time we saw it in full splendor — all the top 10 winning dashboards were long form as were the vast majority of entries.
This popularity of the long-form in the Tableau community versus the wider data visualisation community seems disconnected if I’m honest. As I look through the literature by my desk I see Stephen Few, Alberto Cairo and Andy Kirk fail to mention the term, and there are very few examples as I flick through the pages of their books. With that in mind for 90% of the Iron Viz entries to use the same style and format it feels like something is broken.
Likewise comparatively few of the entries contained overt interaction, which, given the recent fuss around Set Actions seems incredible.
To see Iron Viz so out of step with real-life/everyday dashboarding, and yet held as the pinnacle of our community seems not only wrong, but also borders on arrogance on our part.
“Gaming” Iron Viz
Reviewing the entries and scores makes me concerned. Do people think they’ve “solved” Iron Viz qualifiers?
The score consists of three elements: Storytelling, Design, and Analysis:
Design: How appropriate is the design for the story that is being told?
Storytelling: Is there a clear story being told or question being explored, and is there appropriate context?
Analysis: How sophisticated and appropriate is the data that is being used for the story?
The long-form dashboard certainly makes it easier to demonstrate you’re telling a story, helping the judges award high marks.
Likewise, Josh’s entry, with trend lines, box plots and clusters, helped the judges see, quite clearly, that he’d done the analysis.
Is this the way to win Iron Viz? At the moment I think my honest answer is: Yes.
Is this the best we have to offer?
And so to my concern, when I review Iron Viz, and even my own entry, I am compelled to ask this question — is this really the best we as a community have to give?
My own long-form dashboard was not a conscious decision, it's a design that evolved from the story I was trying to tell. I didn’t intend to follow the crowd — but that’s clearly where I ended up. It’s only my 3rd ever long-form dashboard — the fact I built it for Iron Viz can’t be a coincidence.
As newcomers to our community look at Iron Viz — do they see visualizations that inspire them? Of course. But they also see visualizations a million miles away from visualizations they use day to day. Is what we are demonstrating through Iron Viz a fair representation of our craft? Is it our best? or is it a reaction to what we think is expected from us.
Storytelling should be so much much more than adding a narrative to a set of visualisations. Even small visualisations can tell amazing stories without the need for narrative.
Likewise, more complex interactive pieces can still tell stories through a single chart.
In a similar way, analysis can and should be SO much more than using trend lines and clusters to demonstrate relationships. Analysis to me means teasing out the complexity in a data set and presenting it in the simplest way, helping the audience see the detail.
Design too needs to rely on more than simply pictures and fonts. It is about choosing the most appropriate composition and layout for your visualisation. It is about simple things such as colour choice and the use of white space. It’s about letting visualisations breathe, not choking them with images.
The lack of diversity in the visualisations hurts us as a community, and does little to move us forward. We are more diverse than these 75 entries show — and I think we need to let a range of visualisations shine out if we’re to see Iron Viz grow as a competition.
Where do we go from here?
I’ve love Tableau to clarify more about what great storytelling and analysis means — to share with the judges and community some great examples of both from the wider data viz world. Demonstrate how diverse our field is. If the scoring continues to drive a long-form approach then I’d love to see a change to how the competition is scored to ensure we don’t put off potential entrants who might feel put off by the lack of diversity in the vizzes.
Rob Radburn put it best in a conversation recently, he said Tableau need to look at the scoring/judging, and ask would the viz below win IronViz. If the honest answer is that it wouldn’t (and I doubt it would) then I think we need to look again at the scoring.
To the community judges, I ask you to continue doing the amazing job you are doing. Judging is a hard and thankless task. Continue to challenge each other on your scores and continue to look outside the Tableau world to define good — if we trend towards being insular in what we define as good I fear we’ll lose our way.
75 entries are still far too few for a competition that is the highlight of the Tableau calendar. Makeover Monday gets at least as many entries every week. It’s disappointing to see so many shying away from entering.
For one I’d love to see more of the leaders in our community enter the competition — I fear many don’t enter because the competition has become so out of line with real-world dashboarding, and long-form is not a form they will consider is their specialty. Some already do enter with their own style and personally I think they produce some of the best visualisations in the competition, but I’d love others to help demonstrate a diverse set of visualisations and a range of styles. I’d like them to put themselves on the line and take a risk they might not do well in the scoring— it’s hard to say “this is my best shot” but we need people, especially leaders, to do that.
Once again, I offer my congratulations to everyone who entered, especially Josh. To future entrants, all I ask is that you don’t avoid long form because a Zen Master told you to. Judge for yourself what works best.