Wingspan, my IronViz ‘23 Qualifier: Part I
Behind the Build
What a wild ride! For the “Games”-themed Tableau IronViz 2023 qualifier I had 50+ ideas, spent about three weeks of the month on a completely different topic than I ended up submitting. I pivoted to this idea at the last minute, inspired by hosting a games night with friends the weekend after I’d decided not to submit this year after all. Sitting around my dining room table, explaining the game to a few new players and then all collectively losing badly to our friend Dash, I started to wonder… was there a data set out there? Had some other board game nerd already typed up the entire game? Friends, the answer to that was yes.
The best part of Iron Viz is the community that you can build around working on a hard thing at the same time. This group, formed after someone made a comment about “we should have a support group,” was sustaining and inspiring for me all month.
The other thing that made this build feel possible was that the concept was deep enough to allow for analysis but limited enough to build out completely. Important because the Thursday in the screenshot above was in fact 10/27, T-minus-four days before deadline.
Here’s my original outline: This game is fun! Here's how you play! Also, Birds!
And here’s my (entire) initial wireframe:
I typically wireframe old school. This one’s clearly quite detailed.
At this point I spent an entire research session trying to figure out how to embed a spotify player in the viz so that readers could have an immersive experience. The game is available on Steam and I find the soundtrack very relaxing. I almost always play games on mute but I turn on the noise for Wingspan. In fact, I left the game open but minimized while I built the viz. But you don’t have to take my word for it:
Thankfully I did not waste too much time on this before I got the heads-up that embeds (much like art you did not make yourself) was discouraged by the IronViz team. Back on task! Right?
Well, almost. The next thing I did was play one more game of Wingspan. I did not win, but I did get into a focus zone. I have the AI turned up to “hard” now and I’ve gotten a little rusty! That game is the last data point in the “how to win” section.
Determined not to repeat my earlier mistake of over-designing before you start building, I jumped straight into making charts. I was most curious about the distribution of nest types by habitat and the relationship between versatility and victory points per bird. I knew that if I had that proof of concept, I could scale back if need be and still have something to turn in.
This was originally the entire viz. It was fun, so I kept going!
I had already spent a lot of time over the course of the month thinking about design and layout options for longform vizzes. I find it enough of a hassle to change the font settings in Tableau that I rarely do it, so I wanted to push myself to do something different for IronViz. To that end I’d built a font sampler based on Brittany Rosenau’s tweet. I pulled it up and picked out a pair of fonts: Georgia and Verdana.
Then I grabbed a background color from the original board game art using a browser color picker extension and pulled out the hex code from tableau for "medium black.”
Changing the colors is one of the easiest ways to make a viz look more polished.
Then I took my hex colors over to accessible-colors.com to see if they passed. They didn’t. I played around with making the text darker and the background lighter until I was happy with the look and got a passing combo. I knew I wasn’t going to have time to create a custom gradient background. Picking a background color that’s not white or near-black helps a viz be memorable (at least that’s my theory).
The next (and last) design decision I made was layout. I knew I was going longform. I also knew I needed the flexibility to build out of order. And I didn’t know exactly how many sections I would need (or have time for). So, mostly guided by the size of my MacBook screen, I set a standard section size of 1100 x 688. My plan was to create a floating vertical container for each section… but what should the whole viz size be?
I have never made a 10,000 pixel high viz. Friends on twitter who were also jumping into last minute builds were joking about “activate 10K mode” but I knew I would get overwhelmed with that big of a blank screen. I also didn’t want to do the math on y coordinates myself when I could make a spreadsheet do it for me.
I renamed the containers in the Item Hierarchy for the sake of sanity. Dropped a header/section title into each container so I could find the dang thing. Sent the nests down to y = 4278 and got to work on my intro.
This floating container section layout approach worked really well. I will use it again. They did change size/location the one time I made the viz longer, but thankfully it was easy to go through and put them back since I had height/y coord written down.
The intro was hard to write and full of potential distractions. Among data sets I stopped myself from going hunting for:
A taxonomy of modern & classic board games, ranked by complexity/mechanism (to put Wingspan in context)
The 23 games rated higher than Wingspan on boardgame geek
The 29 strategy games rated higher on boargame geek
Sales and units sold data over time, including expansion release dates
Awards Wingspan had been nominated for and which it had won
I made the compromise of writing a data-forward intro that referenced the key insights from each data set. I told myself that if I had time and thought it would strengthen the story I could come back later but for now just keep building.
By Sunday afternoon I knew I’d have something!
This was about the time that I realized I was creating an incredibly text-heavy viz. The combination of font size and storytelling made the entire Tableau Public preview tile text… and this is a dataviz competition! I earmarked the intro section for future editing but decided to lean into the essay-with-charts vibe. Deciding how much to write yourself into a viz is hard, but for my previous IronViz entry I felt like I had erred on the side of not-enough-context. This time, I decided to push to the other extreme.
But not without some hesitation. Thankfully I got a nice boost from the #datafam.
Fonts & Encouragement - community is the best part of this project.
From this point on it was a build-a-thon. I skipped around. Took breaks. Added another section to the “Getting Started” bit. Worked on whatever seemed fun next. Tried to figure out how to shoehorn some data into the intro and turned the pieces list into a second data source. Decided to build a player mat in Tableau and built a little data source for it in the next tab. Spent a minute being proud of how much better I’ve gotten at creating well-structured data that behaves nicely when you drop it into Tableau.
I ended up building five “mini” data sources to supplement the bird/card data
This was also about the time that I started up the countdown timer. I am pretty terrible at time-related math and get anxiety flashes about AM vs PM, time remaining, all that. So I triple checked the timely tweet from team Tableau Public that spelled out the exact day, date, time and time zone that the submissions would close and put it into a countdown. Then, in an attempt to find anyone else still crazy enough to be working this hard on a completely optional passion project, I tweeted about it and located a few more co-conspirators.
Relative time calcs are very useful. Especially when working across time zones.
Could I have put my pencil down and walked away? Absolutely! A lot has changed for me since the last time I entered Iron Viz (was that really just in January?!). I’ve got a great job that I genuinely enjoy now. I get to viz for work every day. I’ve got datafam all over my DMs. In January, I felt like I needed to enter Iron Viz to keep building my momentum and, to some extent, my visibility in the Tableau Community. Those aren’t factors any more. So, why stay up until Midnight on Halloween? (Literally, my last edit was at 11:59.)
It was fun, and I felt like it.
In my opinion, that’s the best reason to enter Iron Viz!
At 9:29 PM, I took a break from building to fill out the submission form. I wasn’t done, but the link was live and I knew I’d get there somehow!
Stay tuned for Part II, where I’ll share more about the viz itself.
