Yes, yes, the Oscars are coming up and some people are getting excited. Red carpet, Brad Pitt and all that. Many respectable news outlets have been quite confident in La La Land will be taking home the ultimate price of best picture this year. However, apparently other films (potentially not starring as pretty people as La La Land) have been nominated and are competing for the gold. Given that the competition is not over, this post seeks to answer the question: How secure is a La La Land victory? Or are there other serious contenders?
Some might spend hours in the dark reviewing films to answer the question, I am – of course – looking at data instead. My aim is to identify which general characteristics of a nominated film will make it more likely to win using information about past Oscars. The question then is whether La La Land has these characteristics or whether there is a better match among the nominated films.
You can get lost in the options of characterising films. So, I choose to restrict myself and revisit the IMDB data I collected for my previous post that investigated whether films were getting longer. With help from my Italian friend (Stefano) I updated the database and downloaded data on nominated films all the way back to 1986. This gave me 30 years of data on nominated winners (30 films) and losers (148 films) to analyse.
It is written in the (IMDB) stars*
The obvious first thing to look at was the number of stars on IMDB and whether winners in the past have on average received more stars than the losers. Unsurprisingly, it turns out that Oscar winning films tend to have a higher star rating than the nominated films that lost. The margin is not massive, but winners’ average rating is 8 and losers’ is 7.7.
La La Land does indeed do quite well on this metric. With 8.5 stars, La La Land leads the nominated films by 0.5 stars. Therefore, based on that metric, it seems like La La Land might walk home with the price.
But there is a catch. Although on average the winning films tend to have a higher star score than other films, only in every second year has the winner also been the top rated film. What that means is that the best rated film is not always a winner. Given this observation, this year there are still five other films that score above average and are therefore still in the contest (see all films scoring 8 or more in the figure below).
The popular vote*
Second thing that seemed obvious to look at was how many people have made the effort to go and vote for the film on IMDB. My theory for why this should be a good indicator is that (1) many votes implies that many people saw the film; and (2) the kind of people who vote on IMDB tend to be more into films than the general public and therefore their preferences should align not too badly with that of the academy. **
The winners of the last 30 years have on average received around 76% more votes than the other nominated films. But what is interesting about this section is that La La Land is not the most voted for film of the bunch. Instead the film Arrival beats La La Land by more than 50,000 votes. In addition, the film Hacksaw Ridge (which had an above average star rating as you can see above) also has more votes than the average nominated films this year. So, there are two films emerging as serious contenders: Hacksaw Ridge and Arrival.
Less (money), more (Oscars)
“Buying” the Oscar does not seem to be a good strategy either. At least that is the case when looking at the budget data for nominated films. It turns out that since 1986, the winning film was made on a lower budget than the competing films on average (winner around $32m and losers around $40m).
Now, on votes and stars La La Land seems to be in a good place. It is however not alone and so far we can see that both Hacksaw Ridge and Arrival still pose a real threat. And the budget indicator does not look promising for La La Land at first glance. On average the films nominated this year received around $21m in budget and La La Land was estimated to cost around $30m, i.e. it is not cheaper than the competition. But a closer look shows that La La Land should not get too nervous about being relatively expensive. The other two contenders that I have identified based on IMDB stars and votes, Hacksaw Ridge and Arrival, are even more expensive than La La Land.
And the winner is…
Now, just to make it clear. I have not seen a single film of the ones nominated this year. That is for two reasons. The first being that I have been living without a TV for most of last year. The second reason is that I live in Germany, speak limited German and all the films in the Kino here are dubbed.
But based on my numerical analysis on films of the past, my assessment is that La La should be worried and that Arrival has a good chance of winning. But, in addition to the quantitative data from IMDB, I think it is wise to also take into account the qualitative stuff that is flowing around. And that almost everyone tells me everywhere I go that La La will be the winner.
However, this analysis has at least given me the confidence to look at what the betting shops are offering for Arrival. I normally do not bet much, but since the odds on Arrival are 100/1 (meaning that if I put £10 on it winning, I will be banking a £1,000 in profit) I will put some money down this time.
Thanks for reading,
*I understand that using star ratings and number of votes as a predictor is flawed – in particular because winning the Oscar might contribute to number of votes and high star rating on IMDB which would imply reverse causality.
**I acknowledge that the preferences of the Academy are probably a bit more highbrow than those of IMDB voters. But, they are probably more aligned than the ones of the Academy and the average person.