If you’re wondering where the idea for using data about the SEA Games came from, it was sparked by Tableau’s APAC team to support Sports Singapore. The data set, for me at least, was not the easiest to work with, which I’ll get into in a bit. Given the complexity of this data, I was quite impressed with how the Makeover Monday Community got down to business and created some wonderful visualisations.
This week I’m going to highlight three areas that I thought went particularly well and three areas for improvement. Let’s start on a positive note.
MAKEOVER MONDAY VIZ REVIEW
This week, Eva and I hosted the first live Makeover Monday Viz Review. VIZ REVIEW is our way to provide much more comprehensive feedback for everyone. Over the course of 60 minutes, we review visualisations people submit for feedback, talking about what we like and what could be improved. We both thought it went really, really well. The whole idea was Eva’s (of course).
You can watch the first review webinar here and join our next review on Tuesday 8 August at 12pm BST. Register here.
ITERATING ON FEEDBACK
We’ve talked about this many times. Seek feedback on your vizzes, make improvements, ask again, iterate, rinse, repeat. This week, Mina Ozgen did just that. Here’s the series of tweets that demonstrates Mina’s willingness to accept feedback, act on it, and make a better viz.
High level overview of SEA Games @VizWizBI @TriMyData #makeovermonday https://t.co/lr4QCBScpI pic.twitter.com/5XIQo6uDVK
— Mina Ozgen (@MinaOzgen) July 31, 2017
After #MMVizReview I have tried to implement some of the suggested adjustments (and this time I chose a more interesting country!) pic.twitter.com/IDaioinwb9
— Mina Ozgen (@MinaOzgen) July 31, 2017
Final iteration tried to blend a bit of each to both give a sense of geography and keep a semi clean label. pic.twitter.com/bk5oC8jIRi
— Mina Ozgen (@MinaOzgen) July 31, 2017
Mina is just starting her Tableau journey and one of her best traits is knowing that she doesn’t know it all. She’s humble enough to listen and continue to improve. This is a great example for us all.
DEVELOPING A CONCEPT
There are data sets that are often begging for symbolism or are naturally inclined to certain metaphors. Daniel Caroli demonstrated the development of a concept really well this week with this viz.
Since this data set was about winning medals, David designed his viz to look like an Olympic podium. Each medal is stood in its proper position and serves as a card for the stats for those medal winners. This is subtle, yet extremely well executed. Next time your data has a theme, do a bit of brainstorming or search Google images for ideas and inspiration. This can make the difference between a boring viz and one that stand out from the crowd.
Now, onto the lessons we hope everyone can learn from.
LESSON 1: OVERCOUNTING
There were many visualisations I saw this week where people simply counted the number of medals won. The problem here is that in team sports, each person is listed individually. So if a team has four members, should that count as four medals? I don’t think so. It should count as one in order to count the number of events won. This will help ensure that you are weighting medals equally. If you don’t, then you’re implying that a gold won by a team of five members should be given more credit than an individual winning a medal.
The lesson here is to look a bit more thoroughly at the data. When you see that Thailand has won 3,583 medals over five competitions, you should really be questioning the data.
LESSON 2: NORMALIZING THE DATA
Similar to overcounting, normalizing the data can be a great way to conduct data analysis and to add context to your data story. As an example, consider the metric Medals per athlete. Simply looking at the number of athletes, Thailand sends way more to the SEA Games than Laos (1556 vs. 325) and has won way more medals (3,037 vs. 561. Yet Laos has won more medals per athlete (0.434 vs. 0.501).
What does this mean? Well, it means Laos athletes are more likely to win a medal than a Thai athlete.
LESSON 3: MAPS FOR THE SAKE OF MAPS
It’s inevitable – include geographic data and someone will create a map. This week was no exception. For me, a map should only be included if it’s adding value to the visualisation. If a country is tiny and you create a filled map, you won’t be able to see it. Consider using sized dots on the map instead. Again though, the dot should add insight to the visualisation.
Next time you use a map in a dashboard, I want you to try something.
- Remove the map.
- Consider if the story or insight has been reduced by taking the map away.
- If the insight is the same, leave the map off. If the insight has changed, put it back.
Simple test. Give it a try. In fact, you can try this with each and every visualisation you include on a dashboard.
WEEK 31 FAVORITES
Author: Kathryn Ambrose
Link: Tableau Public
What I like:
- Explanation of the SEA Games
- Connected dot plot for the medal counts
- Great use of histograms
- Appropriate color choices
- Makes the complex look simple
Author: Sean Hughes
Link: Github
What I like:
- Nice analytical work
- Using grayscale colors
- Normalizing the data
- Effective use of highlighting
- Labeling the years only where they fit in the space
- Nice subtitle that summarizes the view
- Simple, effective layout
Author: Matt Francis
Link: Tableau Public
What I like:
- Contextualizing the data
- Making the data more about the reader leads to more engagement
- Highlighting only the most likely sport
- Labeling only the most and least likely sports
- Good example of making data viz fun
- Providing a summary of the events the reader has a chance of winning
- Good instructions
Author: Charlie Hutcheson
Link: Tableau Public
What I like:
- Beautiful design
- The font looks really sharp!
- Using colors in the title as the legend
- Great use of highlighting and labeling
- Ranking the countries within each year
- Simple, effective design
Author: Rachel Phang
Link: Tableau Public
What I like:
- Good story!
- Rachel learned something and shared it with all of us
- Using font colors in the subtitle as a legend
- Interesting categorization of the data by former French colonies
- Good use of a map (see my notes above)
- Consistent, limited use of color
- Neat and tidy design
Lesson 3 is a fantastic point no matter what field or technology presentation you are making. Thanks for plainly articulating that!
“Include geographic data and someone will create a map” – this is so true 😉 A map always looks interesting because people immediately relate to it (everyone recognizes a map). But in many cases it doesn’t give useful context to the data story you want tell.