Support Analytics Blog

I borrowed the illustrations below from the latest copy of Information Management in an article by Aaron Hursman.

The first info graphic below shows how the stop light-type chart tends to distract by showing too much.  You can make out the red circle, but since there is so many other colors and circles, it distracts from being able to focus on a few important points.

On the other hand, the graphic below shows how efficient it can be to highlight only exceptions or radical numbers.  The downside is you lose the yellow colored circles that may denote a potential problem.  One solution for this would be to have it interactive where one could select Red, green or yellow.  I know this isn’t a real example, but sticking with the concept, this method can really help to key the audience in on specific areas you want them to focus on.

With the visualization below, you can see how the artificial glare in the pie chart distracts the reader and actually makes it look like it’s a segment or grouping.  Adding “flash” or other bling to charts is rarely a good idea.  One other quick question, where is the slice for “Closeout” in blue?

I would agree that the graph is an improvement over the table below.  However, I’m still not a huge fan of the bullet charts when you put many of them together.  I just find myself going back and forth from the chart to the legend.  Maybe it’s just me with a bad short term memory.  This chart is good for an analyst, but for someone at a higher level it may be too much.  Actually, I favor percentages when illustrating a comparison to prior year or plan.  The takeaway here is that the chart is better than the table, but know your audience and try to add value by pointing out the areas of excellence or concern for the audience.

[Source 1, Source 2]

A big thank you to Support Analytics’ newest advertising partner – TIBCO Spotfire.  If you are involved with data visualization and read most of the popular blogs and books on the subject, you have probably seen TIBCO Spotfire in action.

With TIBCO Silver Spotfire you can be viewing interactive dashboards in your browser or on an ipad in minutes.  Best of all, they are offering a free trial with no obligation.  Check out there offer by clicking here.  They also run a great blog on business intelligence and analytics called Trends and Outliers.

Over the last few years, it’s no secret that I’ve pulled some good and poorly designed data visualizations from the BusinessWeek magazine.  Since Bloomberg took over the magazine the chart/graphs and data/info visualizations have declined significantly.

Here’s my promise to the readers of this blog – this will be the last chart that I pull from BusinessWeek because it was poorly designed.  Take a look at the bar chart below that was shown in the October 18/24 issue that I just picked up.  Folks, this isn’t even charting 101, this is a pre-requisite.  A multi series 3-D bar chart, really?  I’m disappointed and they should be embarrassed because they need to be better than what’s shown below.  I probably won’t renew my subscription that I’ve had for at least 5 years, not because of the charts, but because I find less and less interesting articles with each new issue.  Then, throw in the awful charts and my choice is an easy one.

[source]

Tip: Four line graphs for each segment with time on the x-axis would be a much better choice than this even if this one wasn’t in 3-D.  Boring? Maybe.  Effective? Yes.

As I thumb through my latest edition of BusinessWeek, I see many illustrations where the data visualization looks like the one below.  Also, it seems that with each new week come worse charts and graphs.  Looking at the one below, the 3-D effect adds absolutely no value and makes this chart pretty much useless.

[BusinessWeek, Oct 4-10]

I am pulling for the new management to do a better job, but I’m not very hopeful.  Want to make this chart better?  Simply remove the third dimension and make it a simple, plain old line graph.  If I had to guess, I’d say this might just be a case where the functionality lets me create it in the software, so it must be good, right?  Wrong.  Maybe Stephen Few is right in that great data visualizations start with well designed software that do not let the creator choose an option like this.

What do you think?  Blame it on poorly designed software or uneducated designers?

P.S.

Here’s another gem from the most recent copy of BW.

[Source]

What better way to start the day then to check up on the Fantasy Football league.  As I checked my team and league standings, noting that I took down an undefeated team this week with the help of Tom Brady and Anquan Boldin, I noticed an article by Dave Richard, Senior Fantasy Writer that utilized Tableau Public.

The interactive data visualization below allows one to see how each player (filter by position) stacks up against the opposing team’s defense each week.

Here is the tracker.  I would love to hear your feedback in terms of if you think this is useful or not.

I can check all of the boxes to just filter by the running backs I have on my team as shown below.  This could help me plan for bye weeks and upcoming matchups. Also, I can quickly see that Hightower and William are both on a bye week at the same time. 

Next, you can see that hover option that gives a forecast of the points the person is expect to get along with the opponent.

Finally, this interactive chart allows a quick comparison between any number of players.  Below you can see my two quarterbacks and for week 6 I may want to go with Cutler because Tom is going against a top rated defense in the Ravens.

1. Pretty pictures will catch you eye, but won’t hold your attention
2. Almost anything is possible with the right programmer or tool
3. Less isn’t always more (ever try that with sweets on a child?) – learn how to effectively use appendices for data
4. My four year old daughter likes to play “work” by typing things into Excel spreadsheets. 
5. When I point at a pie chart that has fairly even slices, and ask which one is bigger, it takes a lot longer and accuracy diminishes compared to a column chart.
6. Helps the analyst answer a question like, “so what”?

What can you add?

I wanted to give the readers here a peek at the article Support Analytics wrote for the Spotfire blog.  Spotfire is a leader in the data visualization space.  If you are interested, the blog post discusses two critical parts to data visualization and analytics and can be found here.

I came across this chart and found it very interesting because of the stated benefits that some foods can have on your health (probably not verified by the FDA).  For instance, did you know that eating bananas can help quiet a cough or that apples improve lung capacity?  I came across this because for the last 8 months or so I’ve been on a diet kick to lose a few pounds and body fat while trying to maintain a healthier lifestyle. 

This grid shows how unorganized data can really cause a problem for the reader or viewer of the data.  This list is great if you want to check out the health benefits of eating a certain food item.  However, if you know that there is a history of heart disease in the family and you want to eat as much food that will protect your heart as possible, you need to scan each item to figure it out (as shown below to illustrate). 

[Publisher of chart]

The creator of the chart didn’t even put similar benefits in the same column as shown in my illustration above.  An interactive version would have been best so that the reader could sort or filter as needed to get what they want.  The next best route may have been to publish a few lists with the most prevalent benefits listed in descending order.  Another option would have been to list all of the health benefits in their own column and put check marks where that benefit is found in each food. 

Organizing or optimizing data that makes sense will go a long way with the readers.

It has been no secret that I have been a big fan of BusinessWeek (BW) over the last few years.  Many of the data visualizations have given me material for this blog whether it’s praise or constructive.  Ever since Bloomberg took over the reigns and redesigned the magazine, the data visualizations have been down right terrible.

As you can clearly see below, the charts and graphs went from mediocre to downright awful.  I never would have looked to BW for innovation on the data visualization front like I would The New York Times, but this material should be embarrassing. 

[Source]

I get it that this area isn’t the focus for BW, but at least bring it up to par.   Josh, call me and we’ll get this straightened out.

I have a vacation coming up where I’m going away with my family for an extended period.  On this vacation, I won’t be doing any work, analytics or any work that is data/chart related.  OK, I may end up seeing how many Corona lime wedges on average fit into a a whole lime [pie chart].  There is going to be fun, sun, family time, some tigers, lazy rivers and relaxation.  Scratch that; I forgot I’m going with my 4 and 1 year olds, so there’s bound to be fun and sun, but not so much on the relaxation front.

Now, the point of this post – I have three books related to data, statistics and analytics that are in my queue.  I read these books for work, but also because it’s my passion so I get a pass while on vacation.  What I want to know from you is what book would you like me to tackle first with a review to follow on this blog.  Here are the three books to choose from:

1. Creating More Effective Graphs by Naomi Robbins
2. Analytics at Work by Tom Davenport
3. The Flaw of Averages by Sam Savage

Post the book you want me to review first in the comments section below or you can send me a message!  Thank you!