The Story of Film — Building my Iron Viz 2022
Iron Viz timing caught me by surprise this year. With a bit under a month to compete I simplified my goals. For 2022, I decided I wanted to:
- Submit a viz. Didn’t have to be my best work, or “better” than my submission last year, just submit something.
- Keep my containers under control
- Keep track of my progress over time, so that mayyybe I could turn it into a blog post.
So here we are! In this post, I’ll walk through how I approached Iron Viz this year. My hope is that someone can say “hey, that’s not that hard, I could do that”. I really enjoyed participating, and love seeing what everyone comes up with under such a broad topic.
Choosing a Topic and Collecting Data
The topic for this year’s Iron Viz was “art”. I knew I wouldn’t really have enough time to learn a new tool, and wanted to stay away from lots of colors. Last year I chose a palette that had 8 colors, and while it was a great learning experience, this year I wanted to keep it simpler. A friend and I have documentary nights, and the one we were currently making our way through was The Story of Film. It’s a 15 hour documentary by Mark Cousins about the history of movie making. Wikipedia has a list of all the films mentioned in order of appearance, so I got to work copying that information into a google sheet, and formatting it for my data source. I made sure to have an “id” field for each row, and added a column for what decade the film was released.
Iteration 1 — Building the Frame.
My first design choice was size. I didn’t think I’d be able to fill out a truly long form viz, so I opted to size my dashboard to the same aspect ratio as a movie poster. For my first iteration I put together some simple BANs for high level things I found interesting, and a heatmap that showed for each episode, how many films from each decade were mentioned.
What stuck out to me right away was although each episode progressed chronologically, there were still films referenced both before and after the time period each episode focused on. I hit save, and downloaded a picture of the viz. I decided that every time I saved a version to Tableau Public, I would download a picture so I could track my iterations.
Iteration 2 — Getting More Data on the Page
Now that I had a start, it was time to explore more data. I had a list of all the directors, and I wanted to see if there was an angle there. I started off with another heatmap, this time with the directors broken out by gender. There wasn’t anything that really jumped out except what was obvious from watching the documentary, men were overrepresented. There were also a lot of films with multiple directors, and I wasn’t sure the best way to parse that out.
Iteration 3 — Even MORE data
Still stuck on what to explore, I kept adding more data points and putting some charts on the page to see how things looked. I added a bar graph of where the directors were from, but nothing really stood out. It also felt like a complicated story to tell, one that I wasn’t quite ready to deal with. I didn’t think a map would suit, and would be a bit reductive. It didn’t seem fair to lump in many immigrants as all coming from North America. I didn’t feel comfortable categorizing folks with backgrounds from multiple countries. The long tail of folks that were simply the only person from their country picked for this installment of the documentary didn’t feel very compelling to me.
Iteration 4 —Feedback Time!
At this stage in the game, I felt like I was really hitting a wall. I had some charts, a few ideas of tricks I wanted to try, but overall I was stuck and not sure how to proceed. Last year I wished I had gotten more feedback before…the night before submission, so this year I signed up for the Iron Viz Feedback Initiative. I was hoping to get some fresh eyes and honest, constructive feedback about where I was at, and it was truly a game changer for me with about a week left to viz.
Below is a picture of my viz at the time of the feedback session; part of me wanted to have something more polished as I presented to folks I looked up to in the community, but I knew that they wouldn’t be judging me for what my viz was or wasn’t, and so I decided to just be very honest with where I was at. I had my existing charts as well as some placeholder blocks for items I hadn’t built out yet.
My feedback loop included Sarah Bartlett, Emily Kund, Chimdi Nwosu, and CJ Mayes. I’d highly recommend checking out their work and or giving them a follow, I’ve learned a lot following their work and their feedback was kind and direct. Exactly what I needed! Here are some items from the feedback session they asked me to consider:
- Swap the BANs with the Title, to help the overall flow of the viz
- Reduce clutter/noise by cleaning up things like episode labels, and give the heatmap plenty of room to breathe.
- For the items that don’t really add to the story (in this case breakdown of gender) it’s okay to just get rid of them.
- Try playing with the element of time and see if anything interesting happens — for example does director origin over time change?
- Keep accessibility in mind, especially making sure there is enough contrast in the colors to keep it readable.
- With the time left in the competition, it was okay to stop adding more data points if it didn’t move the story forward.
Talking through my viz with the group helped me narrow in on what I really thought was interesting, which was that each episode had films that spanned decades. It was really interesting to learn that Star Wars was inspired partly by old Japanese films, and as the documentary progressed I learned more and more directors built off and referenced each others work.
Participating in the feedback loop was one of the most fun parts of the entire process. After staring at my own work for hours, it was exciting to see it through fresh eyes and come away with some ideas I hadn’t considered before. I’d highly recommend anyone working on a major viz project to seek feedback from people you trust or look up to.
Iteration 5 — The Viz Takes its Final Form
Crunch time! With not a lot of time left, I left a lot of ideas on the cutting floor and decided to make the most with what I had. Here are some items I chose to focus on for the final sprint:
- Flipping the title and BANs. It worked! I liked the look of this much better.
- I love heatmaps, and adding some bar charts to make a marginal histogram was something I knew I could do quickly. I liked how this added a different layer of seeing the distribution of films across time and episodes.
- Picking a color for the main chart — I ditched the multi hue palette for a single color. I picked a blue from Blake Robert Mills’ MetBrewer project, which I highly recommend checking out.
- The scatterplot stayed! I felt like I needed something to fill the white space left by the marginal histogram. I chose to keep each dot for each film mention, and added more context about each film mention in the tooltips.
- For the text, I added some personal comments about my experience watching the documentary.
Overall, I’m very happy I chose to compete again in Iron Viz. The slogan is true, win or learn you can’t lose. One thing I found this year is that things that took me a very long time in the past (such as lining up the bars with the histogram) went a little quicker this time around. There were also other ‘behind the scenes’ learning that went into the process that didn’t make the final cut, but that’s okay too. There’s always more to learn!
You can check out the final interactive viz at this link.
Thank you to everyone who offered encouragement along the way. I can’t wait to cheer on the finalists for the competition!
