Dating and InfoSec
So if you don’t follow [link to http://blog.okcupid.com/ no longer works] the folks over at OKCupid, you are missing out on some hot data. In case you’re not aware of it, OKCupid is:
the best dating site on earth. Compiling our observations and statistics from the hundreds of millions of user interactions we’ve logged, we use this outlet to explore the data side of the online dating world.
And in their latest post, they explore what brand of camera makes you look good. You should go read “Don’t be ugly by accident.” [link to http://blog.okcupid.com/index.php/dont-be-ugly-by-accident/ no longer works] I’ll wait.
You’re back? Ok. So here, let me lay this out for you. These folks are applying science, not to dating, but to online dating profiles. They’re not slinging some best practice shtick, or re-writing profiles at $50 a pop, they’re telling you exactly what photos work and which ones don’t. How are they doing this? Data. Experiment. Analysis.
I don’t want to understate the importance of finding a good partner, but I will say how sad it is that they have all this data on this highly intimate activity, and we have 2,000 entries in DatalossDB.
Yes, they’ve clearly shown how easy it is to talk about the effects without touching on the causes.
Their analysis is somewhat OK when they stick to purely technical bits, although putting all the 4/3rds cameras under Panasonic heading was a red flag for anyone with more than passing knowledge of the photography world. Where they completely failed is the synthesis of the disparate points that they found.
All in all, rather good example of how irrelevant the amount of data at your disposal is: if your analysis and synthesis can’t produce meaningful results that will stand up to scrutiny you’re just making people laugh.