As some of you may have seen, I have a feature in this week's New York Magazine in which I use a database of the last thirty years of Oscar history to predict the recipient's of this Sunday's Academy Awards.
This is fundamentally not all that difficult to do, since the winners of other awards such as the BAFTAs and the Golden Globes are quite strongly predictive of success in the Oscars. When the other major awards are split between two or more contenders, we can look at other sorts of tiebreakers: The Academy really does not take kindly to comedies or action films, for instance. And there is such a thing as "sympathy points": if an actress or actor has been nominated for an award several times without winning (such as Kate Winslet for Best Actress), she becomes more likely to collect the hardware. (From a technical standpoint, the challenge is really just to build a reasonably reliable model without overfitting).
Spoilers follow below the fold. One of these, by the way, I'm almost certain that I'm going to get wrong, although I have a pretty good excuse. For the supporting detail, please see the original copy.
Best Supporting Actor: Heath Ledger, The Dark Knight (86% chance of victory)
Best Supporting Actress: Taraji P. Henson, The Curious Case of Benjamin Button (51% chance of victory)
Best Actor: Mickey Rourke, The Wrestler (71% chance of victory)
Best Actress: Kate Winslet, The Reader (68% chance of victory)
Best Director: Danny Boyle, Slumdog Millionaire (99.7% chance of victory)
Best Picture: Slumdog Millionaire (99.0% chance of victory)
2.18.2009
For Entertainment Purposes Only
by Nate Silver @ 7:40 PM...see also forecasting, off-topic
Subscribe to:
Post Comments (Atom)

79 comments
99%? Are you high?
Nothing for Milk? Seriously? That's really disappointing. The LGBT community was shut out by Crash in 2005. I had hoped things would be different this time.
I hope you're wrong about actress. I'm pulling for a Meryl Streep win.
well, you're clearly wrong about Supporting Actress.
Whats='s your excuse for the one you've got wrong? One commentator said maybe the input data were wrong. Is that it?
Hey wait a minute. You're about to appear on Olbermann? In a few minutes? Damn.
*
NATE CONTROLS THE UNIVERSE!!
Sirota must be shitting a brick.
hahhaahhahahhahahahha
My daughter will be very pleased with your 68 percent chance for Kate Winslet. Not only that, she grew tired of fivethirtyeight.com when I made it the home page. This post may change her mind.
Chance of not having to look at Ronald Reagan's pinched, sour face every time I open up 538.com: 0.0008%
You guys are leaving that up just to punish me, but I can't figure out what I did wrong...
So, Nate, which one do you think you're wrong on? Best Picture? Could be Milk if Slumdog Millionaire takes Best Director - and if the Academy decides to spread the gold around.
Sirota just denounced Nate's projections as "elitist"
wow.
Crash didn't “shut out” anyone — it was simply a better movie. Slumdog Millionaire is a shoo-in on the merits by crazy-wide consensus; if it wins, it won't be a slap in the face either. It's about the best movie.
@Jenny: Tell me you're not serious. Please, for the love of God, tell me you're not serious.
Jyrinx-
I wouldn't put it past the wacko.
Nate, I hate to argue with your numbers, but I'm not thinking this is Winslet's year. Maybe if she'd been nominated for RR, but I think the backlash against The Reader will give the award to Streep. (which is great since I liked Doubt better than either film anyway). And I still believe that Cruz will pull off BSA.
I also think that Penn might be more competitive than you figure, although I personally prefer Rourke's performance by a smidgen.
Siobhan-
This is Kate's year!
http://adamsmith.files.wordpress.com/2008/11/winsletvf.jpg
Check her out!
Whew!!!
So this blog is LGBT all the time.....way to connect with country....
What a day.....the attorney general calls Americans cowards because they don't racebait and don't engage in racial introspection.....what a crock...nobody has a sense of humor anymore....
NY Post shows a dead monkey in a cartoon and suddenly it is a metaphor for the assassination of the President.
Does anyone remember the Seinfeld episode where Elaine confronted the New Yorker publisher to explain the cartoon and he couldn't? Same thing here with interpretation in the eye of the beholder.
Definitely Luney Tunes time. Michele Bachmann is the most uneffingbelievably stupid Congressperson of our time.
Best Supporting Actor: Heath Ledger (Totally undeserved, Aaron Eckhart was better, but the Academy are nothing if not a bunch of saps)
Best Supporting Actress: Kelly LeBrock, Weird Science
Best Actor: Mickey Rourke (Wanna see him win, but somehow I don't think the Academy are gonna reward a human punchbag who has shat on so many people in Hollywood)
Best Actress: Kate Winslet (Gotten her tits out so many times she'll surely overtake Helen Mirren and Greta Scacchi)
Best Director: David Fincher (Gump pandering always wins. No way they're giving it to a little gobshite from Manchester)
Best Picture: Slumdog Millionaire (Still can't believe Heat wasn't nominated in 1996. Bunch of fuckin' clowns!)
Okay. Nate on Countdown, momentarily.
PorridgeGun-
"Heat" was 45 minutes too long, toooooooooo many characters, with a predictable ending, after ridiculous diner meeting.
I think Penn will win for Milk, and Amy Adams for BSA.
Put in a good word for Sam Seder at the 10pm slot, Nate!
Jenny said...
PorridgeGun-
"Heat" was 45 minutes too long, toooooooooo many characters, with a predictable ending, after ridiculous diner meeting.
Bollocks, Jen. If anything it's 45 minutes too short. Heat is in the pantheon with the likes of The Godfather Part II, The Deer Hunter, Apocalypse Now and The Right Stuff... Great movies that totally justify their epic running time.
PorridgeGun-
Pacino's taupe in "Heat" - The Horror! The Horror!
Greetings all. Sean Penn was good as Milk, but my best actor vote goes to Frank Langella.... he opens a line to the afterworld and channels Richard Nixon in "Frost/Nixon." That should be Best Picture also, but of course it won't be - too much noise for Slumdog Millionaire. (which, to be fair, I didn't see.) But Langella as Nixon is a once-in-a-lifetime role, and he plays it scary well.
that's toupe, not taupe.
If I had my way, Justin Long would have been nominated for Zack and Miri Make a Porno.
http://www.youtube.com/watch?gl=GB&hl=en-GB&v=mTCOXOD-S_M
Intrade Odds:
Movie: Slumdog +88/ Button +8
Actor: Rourke +65/ Penn +30
Actress: Kate +69/ Streep +20
S Actor: Joker +94/ brolin +5
S Actress: Cruz +60/ Davis +17
Good luck Nate,
I've got one objection to Mickey Rourke though. The academy doesn't really like obvious one-hit wonders, those positive deviations from an actor's average career, like Eddie Murphy's role in "Dreamgirls". Also, Mickey Rourke probably didn't have to "act" anyway, because he "is" the guy he portrayed. That's why I'd put my money on Langella
The Supporting Actress pick is a true longshot. The others are likely (Rourke and Winslet) to certain (Slumdog, Boyle and Ledger.)
Just for Jack:
With regard to 2005, Ang Lee won for directing and Larry McMurtry & Dianna Ossana won for adapting Brokeback Mountain. This year, the Original Screenplay award might go to Dustin Lance Black for Milk though I won't be surprised (I'll be pleased) if it goes to Andrew Stanton, Jim Reardon & Pete Docter for WALL-E.
Jenny said...
PorridgeGun-
Pacino's taupe in "Heat" - The Horror! The Horror!
That was no toupe, dear. Pacino had a fine head of hair in Heat, Carlito's Way and Scent of a Woman.
http://www.imdb.com/title/tt0105323/board/nest/103283884
bald + pacino rakes in 184,000 hit on google.
http://www.moviestar.nu/articles/baldmoviestarsenglish.html
I love it when you do stats, Nate.
That's your core strength. Do more of that.
Jen, his hair started to recede around the time he did The Devil's Advocate (1997). Even I noticed Pacino's hair was a bit iffy in that movie.
Also, what's up with Affleck's hairline on that website?
Alex S, I'm not so sure it is true that the Academy shies away from rewarding the equivalent of "one-hit wonders." Remember Jamie Foxx for Ray? And arguably Forest Whitaker for Last King of Scotland?
Nate,
Someone has to ask. Which prediction do you think is inaccurate? (Best Actor?)
I wonder if you didn't just introduce a new variable into the system Nate. The need not to be predicable, and therefore vote against one of your predictions.
What is it about this post or Nate's piece in the mag that makes this his personal feelings on the matter?
To be rather simplistic, regression is about putting data into a stats program, finding some relationships in said data and then having Stata spit out some predictions based on the final model. So when he says X has a 67% chance of Y happening, it's the data saying this.
@ Bryan,
Whittaker had a number of parts in A-movies before, Platoon, The Color of Money, Good Morning, Vietnam, Bird - he didn't have a lot of good roles in the years before his Oscar though.
You're on to something with Jamie Foxx, but in the year he got his Oscar he was also nominated for another movie, Collateral. The academy must have felt that a new star was on the rise.
Rourke had maybe 2 hits in the 80's (9 1/2 weeks was "nominated" for 3 razzies, Year of the dragon was nominated for 5 razzies) and then a row of bombs.
The best performance of the year came from mayor of Lansing, MI on Fox News:
http://www.youtube.com/watch?v=a-nLS6FJtSM&eurl
@ Juris: Definitely Luney Tunes time. Michele Bachmann is the most uneffingbelievably stupid Congressperson of our time.
Indeed. And Chris Hayes gets the WIN for describing her as "the demi-glace of wingnuttia."
@radicallyamerican: it's not as automatic or mindless a process as you imply. You do have to make decisions about which data to use, how to specify the relationships of interest (additive, interactive), and so forth.
However, this is probably largely a lark for someone like Nate to do. I would guess NY Magazine approached him, and he put in a few hours doing the data entry, analysis, and writeup. They probably pay him a modest sum, but it's a high visibility assignment.
@ Juris: However, this is probably largely a lark for someone like Nate to do.
As my girlfriend said, Nate's just showing off now :)
Yeah, there's some of that going on. But in an interview he gave shortly after the election he said that he was totally exhausted on November 4 but even so couldn't really come up for air til mid-February because of his other obligations including his baseball forecasts (some of that is just coming online today on Baseball Prospectus).
I have a suspicion that once he catches some new energy we're going to see some new and different stuff coming ro 538 in months to come.
With Nate's help you too can win the betting pool at your Oscar party!
Kate's role is a lead, not a supporting one. The academy got it right despite your and the other award show's best efforts. And apparently she will win, if your projections are accurate.
You missed two. Penn wins for Actor and Cruz wins for Supporting Actress.
@Juris
You're right. But I believe I stated a desire to purposely keep it simple, I'm in the same business as Mr Silver, just not at his level yet. One does have say in the functional form of the model. And I suppose one could even create data that says whatever one wants. But assuming one plays it straight, the percentages are coming from the stats, not Nate's desire to have so-and-so win an award.
I just wanted to tilt the conversation back towards this being an objective observation.
@ Juris
Heh. She likes to tease. Showing off is great, and (obviously) we think Nate rocks. He reminds us of my brother, who is also a something of a math savant, but who is far less verbally and socially adept.
As for me, I'm taking those predictions and making some money in my office Oscars pool :)
Saw this New York Magazine article earlier this week. The supporting actress category is clearly the misfire. No chance of Taraji Henson getting this one. The favorite is Cruz, followed closely by Viola Davis, who was a scene stealer in Doubt. My personal opinion is that Sean Penn's going to win Best Actor, despite Mickey Rourke being the favorite. Milk was a better film and Rourke has burned way too many bridges in Hollywood. Although if you ask me, Langella gave the best performance this year.
I think you are wrong on at least two of those Nate, if not more - sorry better stick to politics predicting :).
You say in your piece that Taraji has an edge in BSA because her film is nominated for best pic, yet that seems ignored in giving Rourke the upper hand to Penn. You also seem to discount Penn because he won 5 years ago, yet in 2004 Hilary Swank won, after winning only 5 years before, in the same category, and even though she was going up against 4 nominees that had never won, like Penn is this year.
This year, BActress and BSActress are too hard to predict by statistical models because all of the precursors are not as reliable given that everyone had KW in supporting for the reader and lead for revolutionary road. Taking off a half point from her for this is, as I'm sure you know, arbitrary. Finally, saying that many noms without a win creates a sympathy vote...well, just tell that to Peter O'Toole, nominated what, 8 times over the course of 45 years? Had to settle for a non-merit-based award.
Director, Pic, and Best Supporting Actor are locks as they've each won every relevant precursor. No %s are needed for those, they are no brainers.
Jack-be-nimble said...
So this blog is LGBT all the time.....way to connect with country....
What a day.....the attorney general calls Americans cowards because they don't racebait and don't engage in racial introspection.....what a crock...nobody has a sense of humor anymore....
NY Post shows a dead monkey in a cartoon and suddenly it is a metaphor for the assassination of the President
============
Actually Holder's comment, although harsh, was spot on. So many people are patting themselves on the back because they voted for a black man. That was a good first step, but we have a long way to go. His point was that we still as a society tend to self-segregate. Honestly, how many black , Asian, or LGBT people does the average American socialize with on a daily basis outside of school or work?
About the "cartoon" , my understanding is that the cartoonist was combining the current topics of the chimpanzee attack with the Economic Stimulus Plan. BALDERDASH, to put it nicely. It was a pure racist jab at Obama with a decidedly threatening image of assasination. If he really wanted to parodize, as he said, it might have been more appropriate to draw a PIG. Instead he selected an historically inflammatory image that insults black Americans. But then, what can we expect from a Murdoch publication.
I suspect Nate's not comfortable with his prediction for Best Supporting Actress. Since Winslet won most of the precursor awards, her absence in the vategorey combined with the lack of prior winners throws off the projection model. There's less data to use.
I'm an Ang Lee fan. However, I saw Brokeback Mountain and commented after the movie that except for it being two guys, it was really just a pretty pedestrian Hollywood love story. The story arc was plodding and predictable, and really, drunken molestation leads to years of love? Not really a strong premise.
I actually thought Milk did a much better job of describing a touching gay love story, even if it was as a sidebar. And, to me at least, it looks and feels more 'organic' than Brokeback Mtn. ever did.
@ Jericho: Nate said on Olbermann's show that he might have misdiagnosed that one. Put your money where your gut is.
@ Mikeybackwards: I agree that Milk was a much better movie than Brokeback Mountain. Numbers notwithstanding, I think it has a decent shot at Best Picture.
You are wrong about about Best Supporting Actress and best Actress.
First of all, all the buzz says it is a close race between Penelope Cruz and Viola Davis. And you're prediction was very weak.
The Academy also just doesn't like Kate Winslet to win. Meryl has been nominated again and again--and gosh it has been about twenty years since she won. Her performance, as always, was spotless. Some people may be confused as to why Kate is nominated for the Reader rather than Revolutionary Road. Meryl's win at the BAFTA's also gives her momentum.
My predictions are
Best Supporting Actor: Heath Ledger
Best Supporting Actress: Viola Davis
Best Actor: Sean Penn
Best Actress: Meryl Streep
Best Picture: Slumdog Millionaire
Best Director: Danny Boyle
Cruz wins BSA easily. Otherwise the model is probably right. Goes to show that there are not enough good variables out there to support a model across all awards under all circumstances, such as the Winslet category change. Nate probably should also add info about what various big-name writers, Ebert etc., are predicting to the model. Stay with me on this as I modeled the Franken election correctly in these comments.
@ Edwin - Fair enough, but the the "population" in Minnesota in no way resembles the "population" of the Academy. Just something to keep in mind.
@ Edwin - to clarify: the "voting population" of Minnesota vs. the "voting population" of the Academy.
Had the Academy already made its decisions and sealed all the envelopes before Nate made his predictions? The Academy must have some, if not a considerable, interest in not being highly predictable. I think they will not respond well to having Nate and his computer models spoiling there big night.
Haven't seen a single one of the contenders- Babysitters are just too expensive! I have to wait for Netflix- Do any of them require a big screen.
I just had an October flash back- I WAS so thankful to have access to meaningful data analysis during that election- our MSM does not even seem to have a clear sense of how to accurately present information when it has been correctly used as data-
Yet again I am so thankful for 538- but it is going to be hard to follow this site when an important race is a toss up. It was a pleasure in the Obama McCain race. But it would have been painful in 2000 or 2004-
Ok here's an question that may be unanswerable:
If Nate's analysis were available in 2000 would we, his readers, been able to use the information and analysis he was providing to change the results of the presidential race- For example would we have been able to identify Florida as a serious tipping state in real time and focused our efforts to make sure that it went to Gore in a BIG way that night-
Nate- keep up the honest data analysis it really is a public (and global) service.
Who is Sirota anyway? I guess I could (not should) go to his website and figure it out- Nate's reaction to him reminds me of the whole John Ziegler Affair-so I am not sure following up would be useful.
Nate = modern day nostradamus.
hey nate
any chance of providing the actual point estimates for your coefficients, and some simple model diagnostics? this model is really interesting; i'd love to know the relative salience of earlier awards and box office gross in determining success.
One important research design question: you definitely sampled more films than simply those nominated, right?
Nate: Enough with the fluff. Put your talents to work on the current status of the Franken Coleman contest.
Who wrote the copy to that article, Nate? I hope not you. I mean, Slumdog is "edgy"? Please.
At lunch today, I heard about the oscar predictions of Nate Silver on the radio! It was in the DC area on either 94.7 or 100.3 around 12:45pm. To fame and fortune good sir!
loved seeing u on countdown last night
w/all due respect nate, those are pretty safe choices even w/o any sort of data studies, algorhythms or what-not.
ledger is a sentimental fav, mainly 'cuz he died, same for rourke, mainly 'cuz his career died, winslett has all the buzz, and slumdog millionaire just happened to be the best film by a mile.
only surprise is taraji henson as best supporting actress, but i suppose they need to give forrest gump, i mean benjamin button, a nod.
Taraji P. Henson might deserve Best Supporting Actress but it will probably go to Amy Adams. Why? I'm guessing some voters in the Academy might like the idea of a movie that 'takes on' the Catholic Church. And this is sad to say but some people have the idea that if Jennifer Hudson won the award recently, that "they" already got the award recently so Taraji doesn't deserve it. Sometimes irrational behavior doesn't follow a model.
I have never paid much attention to the Oscars. But I'll admit it: if Nate decided to run some numbers for next winner of the Westminster Kennel Show, or the Pillsbury Bake-Off, or even top seller of Girl Scout Cookies 2009, I'd probably go have a look. I like what Nate does.
However, my adulation of Nate's work is not unlimited. In no case could he get me to pay attention to NASCAR.
Yeah, I saw the Sean Penn - Milk thing coming. It's one of those things that statistics can't cover -- there's no way Hollywood was ever going to give an Academy Award to a guy playing a professional wrestler over a guy playing a famous gay San Franciscan. Was NEVER, EVER going to happen. Sad to see, too -- Rourke's performance was clearly the better, IMO.
キャッシング
インプラント
クレジットカード 現金化
神奈川クリニック眼科 評価
育毛
Fx初心者が失敗しない
アップルマンゴー
視力アップ方法
シーランドダイビングサービス
fx初心者のため
サイパンへレッツGO
ニラ百貨店とは何
自動車保険 ランキング
低カロリー食品・レシピでダイエット
神奈川クリニック眼科のレーシック(福岡・大阪・名古屋・新宿)
武富士お申し込み・詳細はコチラ
スポーツメガネ コンタクト
マネーパートナーズFX
fx初心者の口座開設
にきび跡治療方法
勝田台食のこだわり会
銀座眼科 感染症
下腹を簡単ダイエット
レーシックの費用ってどのくらい
品川でレーシック手術するなら
ロータリークラブ
大阪 視力回復手術レーシック
視力回復方法
名古屋のレーシック眼科オススメランキング
正しい筋肉トレーニング方法
絶対痩せるダイエット方法とは
腰痛治療で快適に
神奈川クリニック眼科(ご予約はコチラ)
Independent Jazz Radio
日産EV(電気自動車)
大阪市 レーシック手術
ツヴァイの評判・口コミ(結婚相談所なら
レーシック手術 保険適用
京都 レーシック手術
中高年の仕事・転職・アルバイト求人探しお助けHP
レーシック口コミ体験談をご紹介
酒店經紀人,
菲梵酒店經紀,
酒店經紀,
禮服酒店上班,
酒店小姐兼職,
便服酒店經紀,
酒店打工經紀,
制服酒店工作,
專業酒店經紀,
合法酒店經紀,
酒店暑假打工,
酒店寒假打工,
酒店經紀人,
菲梵酒店經紀,
酒店經紀,
禮服酒店上班,
酒店經紀人,
菲梵酒店經紀,
酒店經紀,
禮服酒店上班,
酒店小姐兼職,
便服酒店工作,
酒店打工經紀,
制服酒店經紀,
專業酒店經紀,
合法酒店經紀,
酒店暑假打工,
酒店寒假打工,
酒店經紀人,
菲梵酒店經紀,
酒店經紀,
禮服酒店上班,
酒店小姐兼職,
便服酒店工作,
酒店打工經紀,
制服酒店經紀,
酒店經紀,
菲
梵,
Post a Comment