This is a copy of a post I made a couple months ago in the directors group. But It hought I might like to have it in my blog too.
For this next post I would like to reveal part of my freakish nature. This might seem like a complete waste of time to many of you but it is just part of my obsession for chronicaling and rating and making lists in regards to movies. I have been working on a personal algorithm to help me calculate who my favorite directors are.
If you are interested at all my pathological quirk of developing this algorithm, here is the way I calculated it. I have a list of all of the movies I have seen and most of them I have assigned a rating between 1 and 10. For each director of any movie I have seen I have extracted three points of data.
A = the average movie rating of all of the movies I have seen by that director
B = the number of movies I have seen by that director
C - the total number of movies that director has made
Just extracting this data has been difficult. For one I am not counting "shorts" which IMDB classifies as any movie that is less than 45 minutes. I just never rate these films. I find it more difficult to rate a short as compared to a feature length which has more to absorb. Also I do not include movies for which the director has done just one short segment. Then I have to decide if I am going to include movies in which the director has "co-directed" or in which there is more than one director. Also should I count movies in which the director is uncredited? Is there any point in counting movies that the director did that aren't even available anymore? What about movies that the director has rejected after it was made for whatever reason? OK, so these are difficult questions that I will have to decide as I revise my data. But this is the general data for my algorithm.
And now for the final formula. The final score I give to each director is calculated like this:
((((B*A)+((((A-5.5)*(B/C))+5.5)*(C-B)))/C)+(((B-1)*0.5)*A))/(1+((B-1)*0.5))
Ok if you can follow that you might just be as crazy as me. Basically what this means is that directors are penalized to some extent (penalized if they are good but actually get a bonus benefit of the doubt if they actually do suck) for having made movies that I have not seen yet. So the more movies they have made that I haven't seen, the more their score drops. The more movies of theirs I have seen and the higher the percentage both increase the strength of the average score.
So far I have only calculated data for directors for whom have at least one 9 or 10 star movie I have seen. I'll add more data for the other directors whose movies I have seen later, but their scores would not be high enough to break into the top 25 or so anyways.
Here is my list of top 25 directors of all time based on my algorithm. I am only including directors for whom I have seen more than one of their movies. (I have also removed a few directors that I just didn't consider relevent for whatever personal reasons)
1. Joel and Ethan Coen
2. Bruce Robinson
3. Hal Hartley
4. Jim Jarmusch
5. Stanley Kubrick
6. Paul Thomas Anderson
7. Terry Zwigoff
8. Quentin Tarantino
9. Terrence Malick
10. Ming-liang Tsai
11. Terry Gilliam
12. Sergio Leone
13. Whit Stillman
14. Alejandro González Iñárritu
15. Terry Jones
16. Errol Morris
17. Mel Brooks
18. Todd Solondz
19. Robert Zemeckis
20. Wes Anderson
21. Werner Herzog
22. Frank Darabont
23. John R. Cherry III
24. George Lucas
25. Akira Kurosawa