Supporting Actress Statistics

At long last. StinkyLulu's finally done some of the long-neglected sifting, stacking & sorting of the information that's been accumulating since mid-August 2006, when StinkyLulu started to counting, tracking and calculating the vital statistics of those nominated Supporting Actresses.

Yet, 'tis but a start... Much like Supporting Actress Sundays, the compilations that appear below are an ongoing work in progress and, as such, are by no means comprehensive. But StinkyLu hopes this info 'twill prove an illuminating resource for considering precedent in those always contentious conversations about what's too little and too much, too long and too short, too big and too small when contemplating what makes a "Supporting" performance.

The "data" below has been arranged along three criteria: actual screentime; scene count, and the percentage of the film's total running time in which the actress appears. And even though StinkyLulu's no social scientist, calculating these numbers makes for a curious and idiosyncratic "science" at best... So, a few rambling words of clarification.

Click on bold text for link to list.
Actual Screentime: When calculating screentime, StinkyLulu tries to track the actual seconds that the actress's presence (her face, body or voice) occupies the screen, including narrative and non-narrative sequences (i.e. credit sequences). Voiceovers do count as screentime, as do moments when the actress is singing -- alone or en ensemble, so long as it's the "character" singing. The actress's disembodied presence (as when footage of Valentina Cortese is screened for other characters in Day for Night but the actress herself is not in the scene...or as when Adriana Barraza's voice can be heard on the other end of Brad Pitt's receiver in Babel) usually counts, but the random appearance of the actress's elbow in the bottom corner of the frame (Mary Badham in To Kill A Mockingbird) generally does not. Dialogue scenes can be the trickiest to tally, as tracking and timing every cutaway is too anal even for Lulu. For tight dialogue scenes, if the on-camera character is addressing the nominated supporting actress during the cutaway, it counts toward her screentime (Ingrid Bergman's interrogation in Murder On The Orient Express). If, on the other hand, the on-camera character is not particularly or directly focused on the nominee but she happens to be in the scene, it only counts when she's actually in the frame (Lily Tomlin during "I'm Easy"). Finally, certain POV moments count to the actress's screentime if the film clearly wishes for the device to put the viewer "in" the character's head or experience (as in Rinko Kikuchi's dull hum perspective scenes in Babel).
Scene Count: Basically, StinkyLulu uses a variation of the "french scene" principle that privileges the character of the actress under review: the scene begins when she enters and concludes when she exits. But even here, editing and non-linear narratives present some tricksy challenges. Montages count as a single scene, no matter how many set-ups are featured within it. Flashbacks count as two, tallying both the framing scene and the actual flashback (provided the supporting actress appears in both). Extended scenes -- like the courtroom scene in To Kill a Mockingbird or the revelation scene in Murder On The Orient Express -- typically count as a single scene. When determining whether or not a disembodied presence counts as a scene, it tends to depend on whether the appearance operates as an extended evocation of the actress's presence (the painting of Susan Peters in Random Harvest) or functions mostly as a prop (the newspaper photo of Cate Blanchett in the final scene of Notes on a Scandal).
Percentage of Running Time: Simple division. The number of seconds that the actress appears on screen divided by the number of seconds in the film's running time. The resulting percentage is rounded up or down. That's it...
Plus a special extra...
Winner's Stats:
Listing of screentimes & percentages for Best Supporting Actresses only.

And now...
On with the Stats!

1 comment:


this problem seems to some like you're bonkers. To me it's more on par with feeding the hungry or something.

I love it. I think it's worthwhile.

thank you for this very interesting project.