Blogalongabond Tableau

Tired and dehydrated from yesterday, I felt low all day today. I got home and decided to play around with Tableau, an analytical tool that helps you to build graphical representations of data.

A while ago I plotted execution rates per capita, to see where you were most likely to be put to death (assuming random acts of crime across the population, and other baseless assumptions) but this time I wanted to do something useful. I wanted to see which was the most popular James Bond film.

All year, I and a series of other nerds/loons/cultural commentators have been watching and then blogging about the Bond films, in sequence, from Dr No to Quantum of Solace, in preparation for Bond 23, which appears 23 months after we start. Or is it Bond 24, in 24 months? It's all a little hard to remember now, as Bond after Bond stalks across the screen and fires at us.

It's strange how much this has taught me, in a short time, about how rubbish some of the Bond films were, the pleasure that writing about a truly bad film can provide, and, possibly, how to write a bit better. (The test of that will come later, when I start crunching all the Google Analytics data to see if more, or less people are reading after two years of Bondage, compared to before.) But for now, it's also an opportunity to try to use Tableau and see if I can work out how to make it do what I want.¹

And so, all evening when I could have been reading books, or playing Scrabble, or enjoying staring out the window, I've been typing review scores into a spreadsheet and trying to make graphs. I now have a pretty good idea of how people perceive Sean Connery (we have only just started on the Moore-era films) and I was surprised how positive we've all been about George Lazenby. Soon, I'm going to try to show what the most polarising films have been, and if there's a point in this series where fatigue sets in and the army of reviewers begins to slacken their efforts and give up.

And most of it (oh lord, what have I become?) will be via the medium of the pie chart. Shame on me, thrice shame!

Actually, when that's done I'm going to write a screenscraping tool to collect together lies from estate agents in different countries, so you can see how often somebody claims you can rent a thousand square foot apartment in an Asian capital city for the price of a decent dinner, while not mentioning you're renting not the apartment, but a small corner of one closet therein.

Data is visible here.

¹Statistical gonks often get obsessed with one software package and claim it's better than all the rest. Some say this is because you prefer the things you have the most experience with. Other, more reflective geeks will mournfully remark that they prefer beautiful people, but have much more experience with ugly ones.