For this week’s makeover, Andy picked a dataset that looks at all the details around arrests of NFL players over the last 17 years.
The original viz is a selection of interactive charts and tables on http://nflarrest.com/, including an ‘Arrest-O-Meter’ at the top, which has a day counter since the most recent arrest:
What works well:
- The interactivity is engaging and I like being able to change the filters and explore the dataset further
- The top 5 lists below the bar chart are a familiar way to rank things when it comes to sporting data so the audience will quickly be able to understand it
- a lot of information is communicated in the fairly compact space of a single screen
- The Arrest-O-Meter is an impactful addition to communicate the message of crimes being committed by NFL players
What doesn’t work so well:
- The colours. There are too many colours and with the stacked bars they are hard to differentiate
- The team name abbreviation below each bar are meaningless for anyone who doesn’t know much about NFL
- The filtering options by team are hidden at the bottom and result in scarily colourful donut charts that are not very useful for comprehending the information provided
- Overall I’m not sure how bad these statistics are as I don’t have a lot of context or anything to compare it to, so other sports that could be used for reference would help, e.g. soccer, tennis, rugby, etc.
- As the website is about player arrests I would also appreciate a more player-centric view of the data
What I did:
- Having lived in New Zealand and Australia, football players’ behaviours were a constant topic in the media, so I was pretty interested in these statistics.
- Due to a string of events including flight cancellations and a 500km trip home on Sunday evening with strangers in a rental car, I opted to create a summary viz that I could
- build in the car (no super precise mouse actions required)
- create with the battery power left in my laptop
- get done quickly before it was my turn to drive
- I wanted a player-centric summary view and to work with a BAN as well as a bit of context, so I created a simple mobile layout calling out the number of players that had been arrested due to a crime (because crimes committed in the dataset are only those that led to arrest, so I felt ‘number of crimes’ wouldn’t capture the situation accurately).
- I also added a line chart with crimes over time to give context of how many crimes had actually led to arrests.
- For my font I used Montserrat Light, SemiBold and Medium. It won’t render properly on Public, but it looks good in the picture
- I chose a dark background for added drama
- And that was it 🙂
Click on the image for the interactive version