The Oscars Defending Champion Rarely Repeats

Since 2000, the defending Oscar winner has come back and won the same category again in 5 of 592 category-years: under 1%. In Best Actor and Best Adapted Screenplay, the number is exactly zero: 0 of 25 in each, across 2000 through 2025. In Best Picture it is zero over the entire history of the award, 0 of 96 chances since 1929, which is what you would expect from a category no film has ever won twice. The trophy you won last year is not a down payment on this year's.

We went looking for a dynasty at the Oscars and found a revolving door with one very stubborn hinge.

Here is the shape of it. Take every category-year we hold for the Oscars from 2000 through 2025, all 25 categories, and the 592 category-years in them where somebody was actually defending a win, and count how many times the reigning winner walked back up and won that same category again the following year. The answer is five. Five, out of five hundred and ninety-two.

That is not "rare." That is a statistical rounding error wearing a tuxedo.

Wait, does the defending champ basically never win again?

Basically never, yes. And it gets funnier when you zoom into the marquee categories, because you'd expect those to be the sticky ones: the places where a beloved name or a runaway film builds momentum and just keeps rolling. Instead they're the cleanest zeroes on the board.

Look at three of the big ones. Best Picture we hold all the way back to the first ceremony in 1929; the two writing and acting rows below cover 2000 through 2025:

Three categories. One hundred and forty-six chances for a repeat, because the earliest year on record in each has nobody to defend. Zero repeats. If you had spent every one of those years writing down last year's winner as this year's pick in those three races, you'd have gone a perfect 0-for-146.

We want to sit with the Best Picture line for a second, because it's the one people feel the most. Best Picture doesn't even have a mechanism to repeat in the way it feels like it should: a film wins, the field moves on, next year is a brand-new slate. But your gut doesn't file it that way. Your gut files "last year's winner" as "the strong one," and it carries that feeling into a race the movie isn't even eligible for.

So where are the five?

Across the eight marquee categories, the defending winner repeated exactly once in 200 tries. One. And it has a name: Alejandro G. Iñárritu, Best Director for Birdman, then Best Director again for The Revenant the very next February.

That is the repeat everybody knows about. It is not the biggest one.

Widen the lens to all 25 categories and four more show up, every one of them below the line, in the craft races nobody live-tweets:

Read that top line twice, because it is the reason widening the lens was worth doing. Emmanuel Lubezki won Best Cinematography three years running: Gravity, then Birdman, then The Revenant. Three in a row. That is the longest winning streak anywhere at the Oscars this century, and it does not belong to Iñárritu.

And look at the titles. Lubezki's second and third wins were for Birdman and The Revenant: the exact two films that gave Iñárritu his back-to-back. The man everyone points to as the Oscars' one great dynasty had a cinematographer standing three feet away doing it one better, on the same two movies, in a category the broadcast treats as a bathroom break.

Does that break the rule?

No. It sharpens it.

Split the record and the two halves say the same thing at slightly different volumes. In the eight marquee categories, the defending winner repeated 1 time in 200. In the other seventeen, it happened 4 times in 392. Call that roughly twice the rate if you like, but both of them round to nothing, and nobody gets to call four-in-four-hundred "sticky."

What changes is where the exception lives, and that turns out to be useful. Repeats at the Oscars are a craft phenomenon, which makes sense the moment you say it out loud. A cinematographer, a composer, a visual effects supervisor: those are people with a technique, and a technique does not reset in February. A Best Picture race resets completely, every year, by design. So if you are ever going to bet on a repeat, bet on a person with a craft, not on a film with momentum.

Now compare the exact same question asked at a different ceremony (the Emmys) and a completely separate dataset:

That card is Emmy data, not Oscar data, and it covers 200 category-years the same way the marquee Oscar slice does. Same denominator, wildly different answer: 1 repeat at the Oscars, 42 at the Emmys. Television rewards the returning champion about one time in five. Film essentially never does. Different ceremony, same lesson about your gut: the crown is heavier than it looks, and at the Oscars it slips almost every single time.

So what do you actually do with this?

The move is not "never pick a former winner." A name can win, lose, and win again down the road. Plenty of careers do exactly that. The move is to stop treating "won it last year" as evidence for "wins it this year." Above the line, since 2000, it has been evidence of precisely nothing.

Here's a clean way to run it in a pool:

1. Find the defending winner in each category. That's your control group: the pick everyone else is about to make on autopilot.
2. Ask what's actually in front of the voters this year. New field, new performances, new momentum. The last ceremony is history; this one is a fresh slate.
3. Fade the crowd hardest in the big races. In Picture, Actor and the screenplays, the defending name is where lazy ballots pile up, and the record there is a wall of zeroes. That is free separation.
4. Give the craft categories one extra beat. Four of the five repeats live there. If a cinematographer or a composer is defending and the work is loud, that is the one corner of the ballot where the record says "maybe."
5. Don't overthink the genuine toss-ups. Faded chalk isn't a magic wand. Some races are coin flips no matter what you do. Take the coin flip and move on.

If you want to see this instinct trained on the ceremony with the deepest record, our piece on how often the favorite actually wins is the companion read, and the dynasties-and-upsets breakdown shows you what it actually looks like when a long reign finally cracks. The pattern rhymes across ceremonies even when the exact numbers don't travel.

Isn't five in 592 basically noise, though?

Fair objection. And yes, five repeats in 592 is close enough to zero that you shouldn't build a religion on the exact figure. But the direction is the whole point, and the direction is not subtle.

If defending champs repeated at anything like the rate people feel they do, that table up there would be dotted with hits. It isn't. It's a wall of zeroes in the three categories most likely to fool you, one lonely repeat across the rest of the marquee, and a small huddle of technicians who are very good at their jobs. When your gut says "this is the safe one" and the record from 2000 onward says 5-in-592, the record isn't the one that's confused.

The honest read: the ballot filled out from memory is the ballot that loses to the ballot filled out from attention. Not because attention is clever. Because memory keeps reaching for last year's answer to this year's question.

Ready to put the theory on a scoreboard?

Grab a category, find the defending winner, and dare your group to write anything else down. That's the whole game. And it's a lot more fun than it sounds when the confident friend realizes his "lock" is the pick the data has been quietly fading since 2000.

Start a pool and set it loose on your friends. If you want the picks half-built before the arguing starts, the ballot generator gives you a running start, and the scoring guide sorts out how the points land once the envelopes open. What you all play for is your business. We just keep the score honest.

The champ walks in wearing last year's crown, everyone nods, and the envelope disagrees. It usually does. It's been doing it, quietly, since 2000, and the room still hasn't looked up from the guacamole.