A while back, I started working on a project for the MST3K Wiki which I personally call Six Degrees of MST3K. This involves creating a listing in each movie article of the cast and crew that had worked in other MST3K flicks in some capacity and which ones. Currently, my method is to check out the This Day in MSTory feature at Satellite News, note which persons have more than one credit, and add them to a Notepad file, editing updates to the wiki at the end of the week.
The key problem with this method is that it leaves out persons who don’t have birthdays and/or deathdays listed on IMDB. Now there’s a feature in the IMDB Advanced Search which lets you compare two movies and see what cast and crew they share. But with over two hundred having been riffed by MST3K, doing each one at a time would be tedious. Is there some method where such a search could compare one movie against a host of others?
If the IMDb compare thingie has an API, you could write a program with a list of all the riffed movies and just have it iterate through them (for every item on the list, run the compare with every item on the list, skipping itself for obvious reasons).
If it doesn’t have that, I think your best bet would be to download the cast and crew for every movie and put them into a spreadsheet (first column is movie, second column is person, third column is role or position). Then you can iterate over those in a similar fashion.
Failing scripty automation, you can just use that spreadsheet and sort by the second column (person’s name) and see all the people who come up multiple times. Or maybe make duplicates of that list on two different spreadsheet tabs, and do a VLOOKUP something something based on the person’s name? That I’m a little uncertain of. If you use Google Sheets, I know you could write a SQL-type query to at least get a list of movies with overlap.
These solutions still have a one time cost of getting all the info for all the movies. Before you go to the trouble of doing that, I would suggest getting the info for just, say, five movies, some of which you know have overlap, and do a test run as if only five movies have ever been riffed. Whatever method you use to find the links amongst that small set will be the same method you use over the 200+, so you can get a feel if it’s worth the initial effort.
So apparently SandyFrank was using this Python package called Cinemagoer which sounds like what I’m looking for. However, my computer programming experience is what you might call minimal, and I’ve never used Python. Is it particularly difficult to use?
