Woe is me… I still didn’t get a very good prediction rate. Out of 23 predictions, I only got 11 correct, or 48% which is better than last year (by one) but worse than three other years when I was above 60%. After comparing my predictions to the results, I believe I can see what happened and, partly at least, explain why I didn’t do better.
First, let me talk about the individual categories. For the three shorts (Live Action, Documentary, and Animated), I admitted upfront that I wouldn’t do well – I don’t have time to view all fifteen nominees and so I know little about these categories and what makes them tick. I only got one of them right (The Boy, the Mole, the Fox and the Horse – Animated Short) and I have no explanation for that success.
I did much better in the three Special Categories (International Feature, Animated Feature, and Documentary Feature) getting all three of them correct. (If interested in the reasoning behind those guesses, see my Prediction essay.). Without seeing any of them, I feel my success in these three categories at least partially vindicates my algorithm.
In the Minor/Below-the-Line Categories, I actually predicted five of the nine correctly. All Quiet on the Western Front took Sound, Production Design, and Cinematography as I predicted. It also took the Original Score category, which I gave to The Banshees of Inisherin. At the time, I was listening to all five of the soundtracks, but I had already seen Banshees… and, so, I think I was biased in favor of the one movie that I had seen. This theme recurs most importantly in the major categories, and I’ll talk about it later.
On Visual Effects, I noted that, for the first time since I’ve been doing this, all of the nominees were also nominated for other things, and that made predictions much more difficult. So, yes, I got this one wrong, giving it to All Quiet… when it ended up going to Avatar: The Way of Water. I haven’t seen Avatar… yet, but based on what I’ve read, it sounds like a deserved win. So that’s just a bad call that my algorithm wouldn’t have predicted correctly.
Film Editing went to Everything Everywhere All at Once which I predicted based on the algorithm that assigns wins in waves, which Everything… was definitely riding this year. I also correctly predicted that Makeup & Hairstyling would go to The Whale although my algorithm wold have suggested All Quiet…. So this one is a case where overriding my algorithm was correct. But the opposite happened in Costume Design where Black Panther: Wakanda Forever won when I predicted Everything… as part of its winning streak. So The algorithm works more often than not, but not always. (Imagine that!)
Then there is that damn Original Song category. I thought it would be a toss-up between the two big-name singer/song-writers, Lady Gaga and Rihanna. But it ended up being Naatu, Naatu from RRR, something my algorithm would never have predicted. (And, I have to admit, I don’t really like that song!)
Overall, in the minor categories, had I gone with what my algorithm suggests, instead of biasing it, then my results would have been different, but the total number of correct predictions would have been the same. So my method helps but it doesn’t give me a slam dunk.
My algorithm seems to be more effective in the major categories (Acting, Directing, Writing and Best Picture) than in the minor categories. I only got three of the eight major categories correct, but, interestingly, that’s because I ignored my algorithm results instead of using them. I picked the Leading and Supporting Actress and Supporting Actor correctly because I went with choices from Everything.., as my algorithm suggested. But I got all the rest of the major categories wrong because, mostly, I overrode it.
I incorrectly guessed Banshees… for both Original Screenplay and Director, when the awards went to Everything…. Although there was a bit of contest between the two films – because both of them involved Oscar-nominated ensembles – I ended up picking the movie that I had seen. Had I applied my formula without prejudice, I probably would have decided in favor of Everything… which would have given me wins in these two categories.
I still would have lost both Leading Actor and Adapted Screenplay. I am looking forward to both Women Talking and The Whale because of these two awards – they are, apparently better movies than my OQI scores suggest. But I definitely would have been down two in these two categories.
And that brings us to Best Picture. I said in my original predictions essay that it was really a contest between Banshees… and Everything…. I picked Banshees… because I had already picked that movie to win the screenplay and director awards. But had I picked Everything…, for those earlier categories, then I would have also predicted it to win Best Picture. That would have brought my major category success rate up to 6 out of 8, instead of three.
So, when all is said and done, if I had gone with my normal algorithm instead of prejudicing the results with the one movie I had seen (Banshees…) then I would have scored an extra three correct predictions for a 61% prediction score, putting this year’s results more in line with my better success rates.
The take-away from all of this: Make my predictions, using the algorithm, without watching any of the candidates. Watching a movie introduces experiential biases and clouds the predictions the algorithm makes, which, overall, are fairly good. Next year, I won’t watch any movies until after my predictions are published!
As a final note, the Oscar ceremony this year was a great pleasure to watch. Although typically-long, it seemed to move fairly well, Jimmy Kimmel was a great host, the stage was modernized, and everybody behaved themselves. Lady Gaga wasn’t supposed to perform, but somehow they squeezed her in for a heartfelt and minimal performance, while Rihanna’s act was very glossy. So, maybe, the Oscars are back, just like people seem to be returning to the theaters for at least some of the movies – Top Gun: Maverick and Avatar: The Way of Water in particular. Maybe we shouldn’t declare either movies or the Oscars as dead quite yet!
