During week 10 we used SocialBlade‘s statistics on the Top 500 YouTube Gaming Channels to improve on their simple table and create vizzes that would convey the information better than the original.

There were a number of really good submissions this week and I noticed a positive trend of a number of improvements, which I found very encouraging and that showed how this community of Makeover Monday participants is learning as a collective and increasing their analysis and dataviz skills.

What stood out for me this week:

  • many submissions included information about the measures, explaining what they meant and outlining the challenges and ‘unknowns’ in the dataset (e.g. how SocialBlade calculate their scores, etc.). This is great information to include for the viewers of your vizzes who don’t have any background on the topic and need some guidance to interpret the data. I really liked how many people included the information and that many clearly gave a lot of thought on how to embed the text as part of their overall design
  • submissions increasingly contain hyperlinks to external sources of information which is helpful and especially relevant for a topic like YouTube. This and URL actions to make it as easy as possible for your audience to consume the information and understand your viz are great tools to complement the content and values shown in your charts
  • like last week, quite a few of you made a submission, asked for feedback, iterated and finally produced a much improved viz, while showing their workings, taking suggestions on board and learning from others in the community. I enjoy giving feedback (when I have the time) but also watching the process unfold on Twitter whereby participants help each other. You know, Andy and I don’t have all the answers and don’t see every Tweet, so the more you guys can help each other, the better.
  • nice simple vizzes with simplified colour schemes, focusing on the key message

 

Some of the challenges I saw were:

  • Submissions not including pictures. No picture, no pin, no tracking. If you’re unsure, check out this article I wrote on the topic. We very much appreciate you making our lives as easy as possible when it comes to the admin side of this project 😉
  • Some size/volume comparisons that were made treated the views as videos and suggested that billions of views translate into every person on the planet having watched 55 videos (or so). Nope. Views are views. Videos are videos. Videos can be viewed multiple times by a viewer, so if I contribute 100 views to a channel that doesn’t mean I watched 100 videos. Maybe just 10 really good ones and I watched each one 10 times…
  • Colours and icons/logos/images
    • People used the YouTube black and red colour scheme and it helped many to keep it simple with regard to colours. However, black and white are high contrast and bringing red into the mix, especially when used extensively, can be very strenuous to look at.
    • My recommendation would be to use large blocks of black sparingly with grey as a good option to balance out the contrast and to opt for red as a highlight or to draw the eye to specific points
    • Using the YouTube logo as a shape for dozens of data points in a single chart looks busy and messy very quickly and distracts from a potentially very informative viz. Ask yourself: does the chart need to have a shape instead of a simple circle for each data point? Does the shape add any value for the point you’re trying to make? Look at your chart with the shape and duplicate it, removing shapes and just having a simple circle. Does the point still come across? Then stick with the simple version.
      If you’re highlighting a couple of top channels among the top 500, then you could use their logos (please consider copyright etc. here and check whether there is a version available for free use) to make them stand out from the rest. But using 100 different logos in a single chart is not going to make your viz a strong and visually compelling way of conveying information
    • Background colours: we questioned a few people this week on their use of extremely bright background colours. There is nothing wrong with sticking to white, grey, beige or black. A fire-engine red background with black fonts and busy charts is going to distract from the core message you’re trying to send to your audience. Don’t let design get in the way, it should only ever enhance your viz, not diminish its value.
    • Less is more, keep it simple and easy on the eye. Be creative as you go along but before you hit publish, put your viz to the test. Be critical with yourself and your work and if you can, get a friend, colleague or the online community to provide you with some feedback. It will help you get better and gain a new perspective on your work
  • Packed bubble charts and treemaps
    • Last week Andy called this point out and there have been a number of articles (e.g. this one by Jeff Shaffer) written on the use of treemaps as well as packed bubble charts. Unfortunately we still see these charts come through every week and while it’s fun to try things in the ‘Show Me’ menu, that doesn’t mean they HAVE to be used.
    • What information do you want me, as the viewer of your viz, to gain from a packed bubble chart or treemap? See if you yourself could gain those insights from your charts within a couple of seconds. If not, then the chart is probably not the right choice. Try a bar chart, it might just do the trick (it probably will. Very likely.)

 

We try to provide feedback on vizzes as they are published. We have day jobs, so we cannot answer all the questions that come up but we appreciate that there are many other members in the Twitter community who have stepped up and are helping participants to improve their vizzes. Thank you!
Our answers may often be short and not packed with fluff and compliments. That’s because we have 140 characters – at most. We’re not rude, just trying to be concise and fast. We provide the feedback to help you improve, not to criticize the work you have done or you as a person.

We enjoy having you all as part of this project and look forward to the submissions every week, so please keep up the great work and keep surprising us with your ideas and creativity!

 

Now let’s get into this week’s favourites!

Guillermo Mendez-Kestler

Author: Guillermo Mendez-Kestler
Link: Tableau Public

What I like:

  • A simple, fun, engaging viz
  • Two simple bar charts and one scatterplot were all that was needed
  • Images are used effectively and put a smile on my face as I’m looking at the viz
  • The colour scheme is simple and the images and their placement act as a colour legend to then guide me as I look at the scatterplot
  • using a subset of the data in the bar charts helps the story and keeps it focused, while all the data points are included in the scatterplot for perspective
  • Guillermo wrote that his kids helped him and got really into it. I find that very cool and I can imagine they were able to pick just the right images. Plus, having them involved means #MakeoverMonday didn’t take dad away from family time 😉
Joe Radburn

Author: Joe Radburn
Link: Tableau Public

What I like:

  • Joe created a viz that gave me a completely unique perspective on the data. I loved reading his verdicts and they made me check out the channels and watch some videos. And I agree, DanTDM is pretty funny!
  • I also appreciate that Joe shared more information about himself and how he fits into the bigger picture of viewers and subscribers as he told us about his favourite YouTubers
  • A simple colour scheme with highlights
  • His bar charts are neat I like the overall page layout
  • Joe is 8 years old and it’s really cool to have him as part of the project!
Marc Soares

Author: Marc Soares
Link: Tableau Public

What I like:

  • The simplicity!
  • A single scatterplot with great use of colours which are highlighted in the description above it
  • I like the axis labels being quite large to ensure the audience understands what they’re looking at without needing lengthy descriptions
  • Usually I’m not a fan of gridlines, but they work really well here
  • The grey font blends nicely into the background and lets the data stand out
Miguel Cisneros

Author: Miguel Cisneros
Link: Tableau Public

What I like:

  • Mike uses humor to draw in the audience
  • The viz is very creative and I love the static image which is replaced by an embedded video when you click on a circle in the scatterplot.
  • This viz not only showcases Mike’s skills but also the capabilities of Tableau.
  • I’m a huge fan of making everything easy for the viewer and giving them all the information in a single place, which is the case here with the embedded videos
  • I love the silhouette of the child in front of the screen which connects the viz back to its title
  • The filter in the bottom left corner is a nice addition and takes into account the sometimes odd channel names and use of symbols in the names
  • The filter doesn’t just give me a list of channels colour-coded by their grade, but also highlights the chosen channels in the scatterplot