I tracked the endings of 7,384 films, and measured the emotions on 1.57 million actors’ faces, to find out whether the movies still send us home happy.
What a fascinating piece. Your research is, as always, excellent. As audience preferences for happy endings evolved, so did the trends in genre films. To illustrate, contemporary rom-coms bear little resemblance to the acclaimed films of Meg Ryan and Julia Roberts. “The Worst Person in the World” and similar films have paved the way for the deconstructed or anti-rom-com, which, confirmed by your findings, is more bittersweet than romantic.
Incredible piece, incredible research, really fascinating, comprehensive insight into endings. Two films I'm developing have endings that are kinda enigmatic, that is 'not quite sure but probably happy' so really useful to know how to place an ending for best audience satisfaction.
Thank you Stephen but I wonder if 10,421 named performers facial expressions is the right metric. Would focusing on fewer actors such as stars playing in movies that were hits, be more telling? Could the "average" be less accurate than the "iconic" and could facial expressions be misleading?
Let's take French movies as an example: Jean Gabin had the habit of dying at the end of many movie:
La Bandera (1935): shot to death by a sniper.
Pépé le Moko (1937): commits suicide with a knife.
Port of Shadows (1938): shot in the stomach.
Daybreak (1939): suicide.
L'Imposteur (1944): shot to death during a battle.
Martin Roumagnac (1946): shot in the back.
Miroir (1947): shot by police.
La Vérité sur Bébé Donge (1952): poisoned by his wife (for a change!)
Speaking of Murder (1957): shot by Lino Ventura.
Les Misérables (1958): dies of illness.
Le Chat (1971): takes a handful of sleeping pills.
No wonder he was nicknamed the doomed man and that everyone thought about French movies as sad. In reality only 10% of all Gabin's movies end up with his death. Yet, it's the sad ones that had success. So the data here would say Gabin's movies are "not sad" but perception would state otherwise. Not to mention that even when happy, he hardly smiled. So facial expressions may not apply in this case. It might not even apply for movies from India, a country where tragedy is often met with a smile. As you wrote... it's complicated.
thanks Stephen, it’s always good to test the industry’s ‘gut feel’ against real data
i write horror, sci-fi and fantasy, and rarely plan to write screenplays with a sad ending
that said, of my nine completed screenplays, two HAD to have sad endings
the first, because it was based on Poe’s Masque of the Red Death – and the second because it was based on the premise that the Simulation Theory is true, and the programmer(s) of the simulation see humanity as bad code and so hit the DELETE key – however, in each case i gave the protagonists a “win” before their final demise
the other seven scripts all have happy endings, but only one of these is unambiguously so – the other six range from “just barely making it out alive” to “life is great now, but the Big Bad remains undefeated”
I find this article bittersweet but dig its rising end ... A+
This is terrific research. Must have been painstaking. Thanks so much! I've always loved Vinnegut's man-in-the-hole metric.
What a fascinating piece. Your research is, as always, excellent. As audience preferences for happy endings evolved, so did the trends in genre films. To illustrate, contemporary rom-coms bear little resemblance to the acclaimed films of Meg Ryan and Julia Roberts. “The Worst Person in the World” and similar films have paved the way for the deconstructed or anti-rom-com, which, confirmed by your findings, is more bittersweet than romantic.
Incredible piece, incredible research, really fascinating, comprehensive insight into endings. Two films I'm developing have endings that are kinda enigmatic, that is 'not quite sure but probably happy' so really useful to know how to place an ending for best audience satisfaction.
Thank you Stephen but I wonder if 10,421 named performers facial expressions is the right metric. Would focusing on fewer actors such as stars playing in movies that were hits, be more telling? Could the "average" be less accurate than the "iconic" and could facial expressions be misleading?
Let's take French movies as an example: Jean Gabin had the habit of dying at the end of many movie:
La Bandera (1935): shot to death by a sniper.
Pépé le Moko (1937): commits suicide with a knife.
Port of Shadows (1938): shot in the stomach.
Daybreak (1939): suicide.
L'Imposteur (1944): shot to death during a battle.
Martin Roumagnac (1946): shot in the back.
Miroir (1947): shot by police.
La Vérité sur Bébé Donge (1952): poisoned by his wife (for a change!)
Speaking of Murder (1957): shot by Lino Ventura.
Les Misérables (1958): dies of illness.
Le Chat (1971): takes a handful of sleeping pills.
No wonder he was nicknamed the doomed man and that everyone thought about French movies as sad. In reality only 10% of all Gabin's movies end up with his death. Yet, it's the sad ones that had success. So the data here would say Gabin's movies are "not sad" but perception would state otherwise. Not to mention that even when happy, he hardly smiled. So facial expressions may not apply in this case. It might not even apply for movies from India, a country where tragedy is often met with a smile. As you wrote... it's complicated.
thanks Stephen, it’s always good to test the industry’s ‘gut feel’ against real data
i write horror, sci-fi and fantasy, and rarely plan to write screenplays with a sad ending
that said, of my nine completed screenplays, two HAD to have sad endings
the first, because it was based on Poe’s Masque of the Red Death – and the second because it was based on the premise that the Simulation Theory is true, and the programmer(s) of the simulation see humanity as bad code and so hit the DELETE key – however, in each case i gave the protagonists a “win” before their final demise
the other seven scripts all have happy endings, but only one of these is unambiguously so – the other six range from “just barely making it out alive” to “life is great now, but the Big Bad remains undefeated”