[RBerenguel]

As much as I love coffee, stupid development projects (if I were to raise my hand when asked "Have you written some stupid project in the last 3 days?" I'd stop writing, both hands up) and neat data hacks, I can't see the need for something as complex. A spreadsheet table (or even better, a plain text file and a little R, which already has plenty of word similarity libraries) would have been far, far more straightforward. But I tip my (metaphorical, I make shoes, not hats) hat for its geekiness ;)

[podopie]

Totally agree, especially because even I personally would have rather just done all the analysis in Python or R. I figured this to be a great opportunity to play with d3.js a bit, but also build some data analysis on the spot. I like that anyone could easily fork this, play with the code just a bit to fit their needs, and have the majority of their "analysis" done (I only hardcoded the "winner" markings). They can just focus on the fun of brewing. Thanks for the geekiness appreciation!

[RBerenguel]

I completely understand the play bit with d3, I'm always coming up with complex concoctions to use it... Then I realised one of the simplest ways is to post the csv file generated by Python or R and write a neat front-end to display it. But I'm always coming back to the stupid or crazy plans, too so... Don't tempt me again!!