Week 35 – Player Arrests in the NFL Since 2000

Another fun week of Makeover Monday has passed and another timely topic. This week we looked at arrests in the NFL since 2000 as the NFL season begins next week. If you missed the Viz Review on Wednesday, head over to the WEBINARS page to watch it. Eva and I really enjoy doing these!

Let’s get right to it with the lesson learned followed by my favorites this week.

LESSON 1: USE VISUALS THAT ARE EASY TO UNDERSTAND

I’m super supportive of people trying new techniques, practicing charts types they hadn’t built before, exploring creativity. If you choose to do so, it’s also mighty important that the visuals you create are easy to understand. Keep in mind that 99% of the people who see your viz will see a static image on Twitter. That means the design needs to be intuitive and easy to understand at a glance.

Let’s look at an example from Staticum that we looked at during Viz Review…

The problem I had with this particular viz was that the visual was overly complicated due to the overlapping boxes. The sizing of the boxes makes them spread across multiple rows and columns. It’s hard to know which boxes belong to which dimensions. Our recommendation was to remove the overlapping.

Staticum implemented this simple fix and the viz is now much easier to understand.

LESSON 2: LABEL AXES WHEN IT’S NOT OBVIOUS

Often times we’ll hide an axis if we do something like labeling the bars on a bar chart or when we create sparklines. If it isn’t obvious what an axis represents, be sure to include an axis title. Neil Richards created this viz with several reference lines that are acting like an axis.

What confused me here was I didn’t know what 1 yr, 2 yr, etc. meant. Neil quickly iterated and added an axis title that made it much clearer that the years represented time since previous arrest.

LESSON 3: AVOID USING TOO MANY COLORS

It’s easy to fall into the trap of using too many colors when you are showing all 32 teams in the same chart. Many teams have colors that are very similar, making it very difficult to distinguish them. It’s ok to show all 32 teams and when you do, consider using highlighting to either (1) focus on a team you want your audience to focus on or (2) allow your audience to pick a team themselves.

For example, let’s say you want to look at cumulative arrests by team across the seasons. You assign each team’s official color to the team name and create a view like this:

See how all of the colors blend? If you want to go with option 1 and highlight a team you are interested in, you can create a simple boolean calculation that highlights your team. Put that field on the color shelf, stick a label on the line and BOOM!

If you want your audience to pick their team, create a parameter, make a boolean calc that check the parameter vs. the team field. Put this field on the color shelf and BOOM!

If you’re ever unsure if you’re using too much color, you probably are. If you need a second opinion, ask someone that doesn’t know anything about the data. Can they see what you want them to see?