January 16, 2007

Golden Globes still more reliable than Huffington Post

A few days ago, HuffPo's Robert Elisberg debunked the myth that the Golden Globes are a good predictor of the Oscars.

The Golden Globes as a "Precursor to the Oscars" is not only not close to true, it's worse than not close to true. Which is near-impossible. Keep in mind that six of the 13 Globe categories are split into drama and musical-comedy - which allows for twice as many chances to be "right." Some categories have had as many as nine nominees. People watching at home eating cheese dip probably get half the Oscar winners right by pure guessing.

So far, so... wait a second. Since the Oscars notoriously hate comedy, nobody really thinks that the Globe's musical/comedy nominations are a good precursor. It would seem more fair to just analyze the drama entries. So how does Elisberg conduct his analysis?

Pick a random year. Say, 2001. The Globe winner for Best Picture musical/comedy ("Almost Famous") wasn't even nominated for the Oscar. The two Golden Globe winners for Best Actor were Tom Hanks and George Clooney. Swell actors, but the Oscar went to Russell Crowe ("Gladiator") - and Globe-winner Clooney didn't even get an Oscar nomination.

Wait — pick a random year? What kind of test is that? And then he somehow neglects to mention that Russell Crowe did win the Globe for best actor in a drama, which would seem to undermine his "worse than not close to true" assertion.

"In fairness, that's only one year and hardly definitive," Elisberg admits — before adding, "Unfortunately, the results were as dismal the year before. In 2000, the Golden Globes gave their two Best Actor awards to Denzel Washington and Jim Carrey - but the Oscar winner was Kevin Spacey (and Carrey wasn't nominated)."

Well, yes, but the Golden Globe pick for best drama that year — American Beauty — did win the Oscar. So that's one more year at random and one more piece of missing information. Some fairness.

Certainly, other years may show better results. Or...okay, maybe not. But the bottom line is that the Golden Globes do not "predict" anything. Set that in granite and plant the gravestone, once and for all.

By this point, I was getting wary. I love a debunking of conventional wisdom as much as anyone, but why didn't Elisberg actually calculate how good the Globes are at predicting the Oscars? It's easy enough. I know, because I just did it.

Looking at the last 10 years of the Golden Globes and Academy Awards, it becomes clear that the Globes are close as a precursor of some things, not close as a precursor of others, and "worse than not close" at pretty much nothing. Looking only at the Golden Globe drama categories, here's what I found"

Of 52 films nominated for best picture, 29, or 56%, were also nominated for Oscars. It's also worth noting that every time the Academy Awards nominated a musical or comedy for best picture during those ten years, the Golden Globes also nominated those films and usually gave them the top prize for that category.

Of the 10 films that won best picture Globes, 6 went on to win best picture at the Oscars — and in the 2 years that the Oscar went to a musical or comedy, both of those films won the Golden Globe too, putting the Globes' predictive ability at a very respectable 80% it's worth noting that in the 2 years that the Oscar went to a musical or comedy, both of those films won the Golden Globe too.

Of the 10 actors who won Globes for dramatic films, only 3 went on to win the Oscar.

Of the 10 actresses who won Globes for drama, 5 went on to win the Oscar. And in the two years that the Oscar went to an actress in a comedy or musical, the winner was the same woman who had won the Globe in that category.

As for Elisberg's more defensible opinion — since it is only an opinion, not stats — that the Globes are a joke because they're selected by 86 junket whores, it should at least be noted that in both cases where there was a disagreement over best picture, the whores arguably picked the more deserving film: Brokeback Mountain over Crash and The Aviator over Million Dollar Baby.

Posted by Daniel Radosh


in the 2 years that the Oscar went to a musical or comedy, both of those films won the Golden Globe too

So what? This actually makes it harder to use the Globes as predictors, since we have no way of knowing whether the Oscars will go for a musical/comedy in any particular year. You're using your data very selectively.

Yeah, I guess if I really want to use comedy winners in my data it lowers the Globes' average from 60% to 40%. Fixing.

lowers the Globes' average from 60% to 40%

You aren't being fair to yourself here. If you are to use the stats this way the best the Globes can do is 50% - only one Globe winner can possibly be the Oscar winner. Good choice in dumping the percentages from the body of the post.

So, does anyone who actually knows statistics want to take a stab at this?

Here's another "I have a theory about the Oscars, but I'm not going to test that theory, even though it would take me less than a hour to do the necessary research to find out the truth, but I got a quote from some guy from Entertainment Weekly, so that should cover it" story.

I may not know much about statistics and percentages, but I do know that predicting the Oscars is a 20 billion dollar industry.

Post a comment

Powered by
Movable Type 3.2