A new simulation shows how the Oscar system is totally dependent on first-place votes
Last month, I used a recount of the Critics Choice Movie Awards ballots to show how the new Oscar voting system was likely to produce eight Best Picture nominees, not 10.
But it turns out that might be an overly generous number.
Because I didn't get to directly handle the CCMA ballots, I wanted to run another test to see just how the system plays out. So I took the more than 200 Top 10 lists collected on the Movie City News website, and tallied them as though they were Oscar ballots.
The result was that the process relied almost exclusively on voters' first choices — and it only resulted in six Best Picture nominees.
Whereas the old system could include 10 or more rounds of counting during which films could gain ground, the new process limited those extra rounds so dramatically that only one film was able to move up.
And while the Oscar ballot asks voters to rank their five favorite films in order of preference, the movie ranked first on the critics' lists turned out to be the only one that really counted on nearly 80 percent of the ballots.
Academy voters don't always like the same movies the critics did, so you can't use my results to say that "The Tree of Life" is in good shape or "The Help" is in trouble.
But you can use them to say that if you don't have enough people put you in the No. 1 spot on their ballot, you're in very big trouble.
Here's how the count worked:
First, 55 different movies received first-place votes on the 211 ballots. (I couldn't use any of the MCN ballots that listed favorites in alphabetical order, because you need to rank your favorites for the Oscar ballot to work.)
The magic number required to secure a nomination in the first round of counting was 20, and only two films got there. "The Artist" received 23 first-place votes, while "The Tree of Life" received 22.
"Drive" was in third place, with 15 votes, followed by "Melancholia" with 14, "The Descendants" with 12 and "Hugo" with nine. "Margaret" and "Moneyball" each had six, four movies had five, and the other 81 votes were spread out among 43 other films.
(In the weighted system that MCN uses to tally the results, "The Tree of Life" actually finished well ahead of "The Artist," because it was listed on more ballots. But "The Artist" got more first-place votes, which is what the Academy relies on.)
The next step was to determine if any film got 10 percent more votes than it needed to secure a nomination; if so, it triggers a "surplus rule" that redistributes part of each of its votes to the film ranked second on the ballot.
(In the past, the surplus rule was triggered when a film got 20 percent more votes than it needed – but with the move to 10 nominees, the threshold was reduced to 10 percent in the Best Picture category.)
NOTE: In the original version of this story, I worked from the understanding that the rule required more than a 10 percent surplus, a conclusion I had asked the Academy confirm. It took a week's time and a consultation between AMPAS COO Ric Robertson and PricewaterhouseCoopers partner Rick Rosas for me to get an answer to my question: in fact, a surplus of exactly 10 percent would also trigger the rule.
The rest of the story has been adjusted to reflect that new understanding.
"The Tree of Life" got 22 votes, exactly 10 percent more votes than it needed. "The Artist" got 23 votes, and triggered the rule to a slightly larger degree: On each of its ballots, 87 percent of a vote went to "The Artist" and 13 percent to the movie listed second.
That redistribution gave five partial votes to "The Descendants," three to "Melancholia," and two each to "Hugo," "Drive," "War Horse" and "Tinker, Tailor, Soldier, Spy." But because those partial votes were only worth .13 of a vote, they didn't do anything to change the standings.
The redistribution of one-tenth points from "The Artist" did even less to affect the standings.
In past years, this would be the beginning of numerous rounds of redistribution, in which the films with the least support were eliminated and their votes allocated to the voters' number two choices. But under the new system, only one round of redistribution takes place, and only with the films that receive less than one percent of the vote.
In this case, that meant redistributing ballots of the 30 films that had one or two first-place votes. On each of those 30 ballots, I crossed out the top-ranked film, and allocated the vote to the critic's second choice.
If that second choice was "The Artist" or "The Tree of Life," which had already secured nominations, or if it was something else that was no longer among the 23 films still in the running, then I looked to the number three choice, or to the highest-ranked film on each ballot still in the running.
Example: Mal Vincent listed "50/50" first – but he was one of only two critics who ranked that film first, so it fell below one percent. But his second choice, "The Artist," had already secured a nomination and didn't need his vote, which instead went to his number-three pick, "The Descendants."
In the redistribution round, "Drive," "Hugo" and "Poetry" picked up five new votes, "The Descendants" and "A Separation" and "Martha Marcy May Marlene" four, and 11 other films received new votes.
After adding those new votes to the totals, this was the new top 10:
"The Artist" – 23 votes
"The Tree of Life" – 23
"Drive" — 20.26
"Melancholia" — 15.39
"The Descendants" — 16.65
"Hugo" — 14.26
"Margaret" – 9.13
"A Separation" – 9
"Poetry" – 8.13
"Moneyball" — 8
At this point, any movie with more than 5 percent of the vote is a nominee; any move with less than that is not. With 211 voters, 5 percent is 11 votes. That means that everything from "Hugo" on up is a nominee, and everything from "Margaret" on down is not.
It's important to note that "The Artist," "The Tree of Life," "Drive," "Melancholia" and "The Descendants" all had more than five percent of the vote in the first count.
"Hugo" was the only film that started out below the line and ended up over it – the only one able to move into a nominee's slot during the brief redistribution.
Under the old system, films like "A Separation" and "Poetry," which made impressive moves, would have had many more rounds in which to continue climbing toward one of the 10 slots. But the new process limits those opportunities, and this simulation suggests that if you're not in the money on the basis of your first-place votes, you're probably not going to get there.
On 165 ballots, almost 80 percent of the total, the vote went to the critic's first choice. (45 of those ballots also allocated a very small percentage to the second choice.)
On 18 ballots, the voter's second choice came into play; on eight ballots, their third choice did. Three voters even wound up casting a vote for their seventh, ninth and 10th choices, although Oscar voters wouldn't be able to do that because they only had five slots this year.
Almost a third of the critics, 69, voted for films that got more than one percent of the vote but less than five percent, and thus did not have their ballots participate in any redistribution.
And two voters, a little less than one percent, had their ballots discarded because they didn't vote for anybody with enough support.
Obviously, the way it worked for critics' lists isn't the way it's going to work for the Academy. But the huge dependence on number-one votes is clearly a key factor in the system – and the results are prompting me to go back to my original prediction that we'll see seven Best Picture nominees, not eight.
And even that might be a bit optimistic.