The good, the bad, and the ugly of data visualization

the good, the bad and the ugly

There are 4 basic categories of data visualization geeks in my mind: the statistician, the geographer, the designer, the computer scientist. Easy.

(Ok I know: the good, the bad and the ugly are three but I just needed an attractive title.)

There are things I hate and love of each of you. I’ll tell it right away: I am a CS and I love it.

The Statistician

Respect. Almost all the basics of data visualization come from statistics. We people from information visualization with CS background and HCIish tendencies love to think we invented the field. No, statisticians did. John Tukey, Edward Tufte, William Cleveland.

Guys, I hope you are not offended, but why in the world the vast majority of the stuff you do looks so ugly and complex? There is a whole world of people out there seeking to visualize the number of calories they ate in the last month to monitor their diet.

The world has changed, people want easy tools with nice interfaces. I know, it’s glossy stuff … but do you want to keep developing stuff for a pin-sized niche?

The Geographer

Jacques Bertin invented information visualization. And, even if most people don’t know it, geovisualizers run hundreds of useful experiments we still have to put into practice in our field.

Also, people are so much attached to maps, regions, continents, they understand it in a second, that I am a bit envious of you. C’mon maps are the most powerful visual tool humans invented in history. Geography rocks!

But you know what guys … Pfff … aren’t you tired of these maps? Isn’t it too limited? Do you really think there is still space for huge innovation here? I hope so.

The Designer

You know what? I hate you guys. Really. I am so jealous of how beautiful your stuff looks like. Neat and clean, attractive, artistic at times. Cool.

First, I hate you because I would like to be a little bit like you and I can’t because I am not a designer and I will never be.

Second, I hate you because you have a lot of success! Really, I spent 10 years of my life to get a master degree and a phd and since then I sweat my shirt every day writing papers waiting to be sent to the reviewer’s claws. Then I show what I do to my wife and she says: “Mmm … so this is your job?” And you? You have a lot of success because you are irresistibly attractive. You win. Well done.

But guys … can you make it a little bit more useful from time to time? Really I think you can do better than that, and then you will be the kings.

The Computer Scientist

These are my fellow colleagues. We coined the term “Information Visualization”, we have been pushing forward this stuff for years. Don’t we merit some respect too? In our team we have the inventor of treemaps and parallel coordinates, we dare for interaction more than any other and we are rocking this stuff every single day. Not bad right?

We also have the best modern books around (too few anyway) and the best conferences and journals in the field. And we have funded succesful companies like Spotfire (now Tibco) and Tableau. We are making this thing real. Sure … now I know … we deserve some respect.

And our faults? Hard to tell … maybe you can help me with this?

If I have to mention one: we still lack some solid theory, a science of information visualization. But for this you can ask to Robert and his fellows who are organizing a workshop on theory at infovis.

Conclusion

Ok guys, no really … I love you all. I know this is a bit rough but I needed to have some fun. I wish to thank all of you guys who shape and have been shaping this field from different angles for years. I love it. I’ve learned so much from you and I am still learning every single day. This is a nice place to be in 2010. Spread the word!

Any harsh comments? :-) You can contact me on twitter or leave a comment here below.

9 thoughts on “The good, the bad, and the ugly of data visualization

  1. Tom L

    Interesting. You’ve listed the creators and suppliers, but what of the market? I’d add those commissioning the use of dataviz, who might not have the skills but have the imagination to see its value; and, those consuming it for whatever reason. And what of human rights, environmental and social justice activists ? We are beginning to make extensive use of information design and visualisation as well: http://tacticaltech.org/reveal

    Reply
    1. Enrico Post author

      Good point Tom. The point is that while the spectrum of visualization providers is somewhat defined, that of visualization consumers is totally blurred. And of course this is a big issue because we need to know better our market. Well … thank you! This could be the idea for another post of mine. I gave a look to what you are doing and it looks amazing to me. The whole idea of using visualization to help activists or rights advocates is brilliant. I’d love to know more about it from you.

      @Carlos: thanks for your message!

      Reply
  2. Bryan Connor

    I definitely fall into the Designer camp but I swear I’m trying to figure out how to make my data visualization more useful and take some hints from the Computer Scientists and Statisticians of the data viz world. Maybe the ideal is some kind of combination of all 3 (or 4) of these camps and that’s what I am to do!

    Great blog, I’ll definitely be reading more.

    Reply
    1. Enrico Post author

      Ah ah … a designer! Please teach me your art! :-)

      The post started more like a kind of joke where I tried to stigmatize the main figures you find in visualization. But I agree … the best is to take the best of all of them. Plus, I know of many designers who’s visualizations are not only brilliant but also useful and informative.

      Thanks a lot dear Bryan, I am really grateful to readers writing comments to my posts. I am glad you are enjoying it and plan to come back.

      Reply
    2. Andy

      Ahh – the combo. That would be a “Data Scientist”, and I have two on my team. Data Scientists require knowledge and skill in all of these areas, understanding the data at the lowest level, performing statistical analysis at the highest professional level, creating visualisations to present their insight that would put a lot of pro-designers to shame and more often than not, the visualisation process will require them to continually learn new programming languages.

      They’re expensive though :-)

      Reply
      1. Enrico Post author

        Right. People might call this a Data Scientist … though you might also call it Visual Analytics Expert too.

        By the way, cool! You have two on your team?! I’d love to know more how good they are in vis and how vis is used in your area.

        Yes, they are expensive … ooops … WE are expensive :-)

        Reply
  3. Sarah

    Hello!

    Thanks for the article :) I’m an artist/designer who is fascinated by info visualization, but my work has been in animation/painting/musical graphic scores (it’s own type of info vis, I guess)~ definitely do not know how to program. What programs would be best to know? Is this impossible without being able to program yourself? I’m so intimidated by it :-x ((I thought it may be possible on some kind of team.)) Thanks!

    Reply
    1. Enrico Post author

      Hi Sarah, nice to see you here. Being able to program could definitely be useful. Processing (http://processing.org/) is a simple programming environment designed to be used by artists and designers and it might well be worth it spend some time to learn it. Anyway, just yesterday I discovered this new tool called Impure (http://www.impure.com/) which allows people to build visualizations without programming. Frankly, I don’t know if it is easier in the end but it looks powerful. Unfortunately, the documentation is really poor but I know they are planning to make it a lot better. Good luck!

      Reply

Leave a Reply

Your email address will not be published. Required fields are marked *