by Andrew Boyd
Today, what's best? The University of Houston's College of Engineering presents this series about the machines that make our civilization run, and the people whose ingenuity created them.
The votes were cast and counted, and The Hurt Locker was named best picture at the Eighty-Second Academy Awards ceremony. The film beat out nine other nominees, including the tech-nological and box office wonder, Avatar. But how does a group decide on a best picture?
We put a bunch of people in a room and ask them collectively, what do you, as a group, say the best picture is? If I think it's The Hurt Locker while you think it's Avatar, how can we speak with one voice about the best? Yet we do it all the time. We express our preferences, throw them into a voting system, and see what pops out on top.
In 2010, when the list of best picture nominees went from five to ten, the Academy changed its voting system. It used to be the most votes wins. Now each member of the committee rank orders the films from one to ten. If one movie gets first place votes from more than half of the committee, it's declared the winner. But if not, the movie that received the least first place votes is thrown out of the pile. The remaining films are then ranked from one to nine and the process is repeated. The committee doesn't put together new lists — each member's original list is used, but with the lowest vote-getter removed.
This system of voting goes under many names — instant runoff is one. It's easy to see why. The system works like a runoff election system where the lowest vote-getter is tossed out in each round. However, there's no need to call a new election after each step of the process. That saves time and money.
The Academy's not the only group using instant runoff voting. It's used to elect political candidates in Australia, Ireland, and even in parts of the United States.
It sounds like a pretty good method, and it has a lot going for it. But it has some hidden and very disturbing problems as well. For example, suppose the committee compared Avatar head to head with every other film — not just The Hurt Locker, but Up, The Blind Side, the whole list of nominees — and found that in every case, Avatar won the pair wise comparison. Then we'd hope that Avatar would emerge victorious from the instant runoff system. How could any other film possibly make a claim to being best? But that's a real possibility. Try to construct an example yourself, or look at the Engines web site. It's really not that hard. The problem can occur with as few as five committee members and three films.
Of course, we shouldn't be too hard on the Academy. Nobel laureate Kenneth Arrow proved that all voting systems have monumental failings. And perhaps that's just as well, reminding us that 'best' is, well, at best, an elusive aim.
I'm Andy Boyd at the University of Houston, where we're interested in the way inventive minds work.
For a related episode, see 2427, ARROW'S PARADOX.
Oscar's 'New Math': How Best Picture Will Be Picked. March 5, 2010.