Every year, the producers behind the Emmys, Globes, and Oscars ask the same question: who can we put on stage that will lead to better ratings (than last year, at least)? That question has been asked for close to a century now, and it’s been answered with varying degrees of success. For every 1990 Billy Crystal, there’s a 2011 James Franco and Anne Hathaway fiasco. Sometimes it feels like a producer might be better off throwing a dart at a random assortment of headshots.

But what if we used some statistics to help us achieve a better turnout? Could systematically looking back at past awards shows provide any indication about who to choose for next year’s telecast? Maybe, maybe not. We won’t know if we don’t try it. So let’s crunch some numbers, shall we?

Before we get started, let’s get one thing straight about statistics. As any fantasy football player or bookie knows, anything can happen on any given Sunday. A team with an 0-8 record against a particular opponent could end up winning the game. The players, rules, referees, etc. all change over time. So with that in mind, it’s worth noting that the views and opinions of the general public have changed with time. Prior to 1994, some people might’ve been open to an OJ Simpson-hosted Oscars. After 1994, not so much. The point is, history can’t tell us everything about the future.

To keep this article from running 5000 words and also due to a lack of information about the Emmys and Globes, this post will look specifically at the Academy Awards. Wikipedia lists the number of viewers since 1974, so that’s how far back we’ll go. Here’s a simple graph to start, showing how the number of viewers has fluctuated each year:

There’s not much of a clear pattern, besides the overall downtrend in viewers over time. That’s not surprising, especially as “cord-cutting” becomes more and more popular with Amazon, Netflix, and of course illegal downloading. Still, there were some obvious good and bad years. Are there indications about why some were winners and some were losers? Let’s look at the first type of host: The Repeat.

From 1974 – 2015 there has been a total of 9 returning hosts: Bob Hope, Johnny Carson, Chevy Chase, Billy Crystal, Whoopi Goldberg, Steve Martin, Chris Rock (as of this upcoming year) Jon Stewart, and Ellen DeGeneres. Let’s see how it plays out if we plot a graph using the numbers of viewers during each of their years hosting.

First off, Billy Crystal has hosted the Academy Awards NINE times. That’s impressive, although Bob Hope did 19! Secondly, notice the downward trend as these people host more shows. Crystal had a spike in 1998, but that was also the year Titanic won 11 of its 14 nominations and set box office records. I think the moral of this graph is that after two years, a host isn’t going to bring in more viewers if they’re brought back for a 3rd or more. And if you’re Steve Martin or Jon Stewart, it’s downhill after just one year.

Next up, how do women hosts fare against their male counterparts? Rather than make a graph of every host since 1974, I’ve pitted the best male performers against the (few) female hosts. Let’s take a look:

Women aren’t too far behind men, but overall they do come up a bit shy. Liza Minelli is the only one to match the men’s numbers, but she also co-hosted with Dudley Moore, Richard Pryor, and Walter Matthau. If you disregard that year, then Whoopi is the closest with 45.6 million viewers compared to David Letterman’s 48.27 in 1995. Unfortunately, the results are pretty skewed by the disparity between the number of female hosts compared to the number of male hosts. Maybe a better comparison is the Golden Globes – where Amy Poehler and Tina Fey easily beat Ricky Gervais over the past six years. We can’t get a totally accurate picture of men vs. women until (hypothetically) women host an equal amount of Oscar nights as men.

Those are the biggest variations of Oscar hosts, but let’s take a look at one more quickly. How do telecasts featuring multiple hosts fare against those with a single host?

Not very well. The numbers in the key on the right indicate their rankings, where #1 is the best performers and #9 are the worst performers for both categories (single host vs. multi). The results are pretty clear: multiple hosts historically perform worse than their single host counterparts.

And lastly, you don’t need a chart to realize that comedian hosts do far better than those that aren’t (i.e. James Franco, Anne Hathaway, & Neil Patrick Harris).

So what’s the takeaway from all this? With a few exceptions, a single host that’s a male comedian who is new or returning for the 2nd time is the best bet. That might make the academy feel good about picking Chris Rock to return, but a look at the drop in viewers between Jon Stewart’s performances might mean the choice isn’t as safe as they’re hoping it’ll be. Will Rock drop 7 million viewers in his 2nd outing, as Stewart did, or will round two lead to higher numbers like Ellen DeGeneres?

What do you think? Leave a comment, and thanks for reading!