6 ms·

"Success in the movie industry relies on a studio’s ability to attract moviegoers—but that’s sometimes easier said than done" No it isn't. Just make a good mov

"Success in the movie industry relies on a studio’s ability to attract moviegoers—but that’s sometimes easier said than done" No it isn't. Just make a good movie. Forrest Gump didn't need any machine learning. The only reason someone might use this is when your strategy is "spray and pray", create crappy movies that insult everyone's intelligence and you need to figure out is it worth launching at all.

yodon 8y ago
As with most simple recipies "just make a good movie" is unfortunately neither sufficient nor always the best path. Word of mouth on movies is far more likely to have a negative impact on theater going decisions than positive impact (if your friend says it sucks, you are far more likely to switch from go to no go than you would be to switch from no go to go if the same friend said it was awesome). In practice, what this means is that because positive word of mouth is so minimally impactful studios need to focus on high marketing value films rather than high story value films, because high marketing value films cause potential audience members to see a trailer or an ad and say "I'll totally go see that!" Yes, negative word of mouth can and will reduce the number who actually go see it, but final viewership is higher if you start with a huge population saying "I'll totally see that" than if you start with a "meh" response to the initial ads and try to grow viewership through word of mouth about how awesome your film is. Given these observations, Hollywood has made the rationally and financially correct but unfortunate for audiences decision to focus on marketability over enjoyability. These economics are quite different for entertainment as a service providers like HBO and Netflix, with the obvious results in quality of output.
- smichel17 8y ago
  The alternative is to build up a reputation, so when a new $studio film comes out, people are excited about it because $studio has a history of making good films. I think Pixar is/was a good example of this.
  - NicoJuicy 8y ago
    Every studio that has focus on children is smart marketing. Children > parents. If you are a parent, don't you go to a child friendly restaurant that is reasonable good instead of one that is very good but not child-friendly.
  - TylerE 8y ago
    That only works if the films are still good though. Pixar has had a couple of bombs, The Good Dinosaur for one.
selflesssieve 8y ago
> Just make a good movie. That seems obvious- but good movie does not mean high grossing blockbuster. Shawshank Redemption made less than the Bee Movie.
- erichocean 8y ago
  Not all good movies can have good/effective trailers made for them, and not all word of mouth can overcome a bad trailer.
- msla 8y ago
  Shawshank Redemption joins Office Space, Fight Club, and It's A Wonderful Life in the ranks of films saved by being shown on TV (whether via cable, DVD, or obsessive rebroadcasts over the holiday season before It's A Wonderful Life got snatched out of the public domain) after having bombed in theaters. They eventually found an audience even if the vagaries of the theater world kept them from making much in box office sensu stricto.
  - melkiaur 8y ago
    Maybe they needed to get a bigger marketing budget. Now, if only the majors could predict which movie need to be marketed how...
  - tokyodude 8y ago
    There are movies I want to see multiple times and movies I truly loved but only needed to see once. I would assume block blusters fit the "multiple times" category which is why they make so much money, because people go more than once. A drama like Shawshank Redemption (a favorite of mine) is not the type of movie I would have paid to go see multiple times in the theatre. Where as I saw The Matrix 6 times in the theatre back when it came out. Every friend who wanted to go see it I went with.
dagw 8y ago
No it isn't. Just make a good movie. Lots of "good movies" (using any definition you want) fail to find an audience. And even more mediocre movies make huge buckets of money. If your goal is to make money then just blindly making "good movies" is probably not your best strategy.
- freediver 8y ago
  Depends on your horizon.
  - yodon 8y ago
    > Depends on your horizon Box office returns dominate the return for a film, set the license price for long term post-theatrical deals, and decay with a very fast exponential from the opening weekend take. It is extremely rare in modern times for a film to turn a poor opening weekend into a high long term return. The film market simply doesn't work that way however nice it would be if it did.
    - rjf72 8y ago
      I read the comment you're responding to a bit differently. I think most people can enjoy a really bad movie. Some t-and-a, explosions, a bit of CG - it's all good fun even all you can think about after the movie is how awful it was. But when really bad movies are all there is, they quickly lose their charm. Instead of being one off guilty pleasures, they become the movie industry. And that's about the point you stop watching movies at the theater. I think this describes many people's transition away from movies. The point of this is that pumping out crap may be optimal in the short run. But in the longrun you end up with suboptimal returns, even when profit is your one and only motivation. Incidentally the same story is true and also being played out in various other industries. Optimizing for next year's quarterlies is often not the same path to optimizing for earning the most money in the next decade.
- TeMPOraL 8y ago
  > If your goal is to make money That's one of the reasons why we have so many poor movies (the other reason is that making movies is hard). If your only consideration is making money, you'll produce whatever crap is cheapest to make that still makes reliable income. This is why e.g. consumer market is full of barely-working near-disposable products which existence is a pure waste of natural resources. Seriously. There's no rule that says that the most profitable product is also the best for the user. In fact, it's almost always the opposite, and A/B-testing things to death leads to the former type of product.
msla 8y ago
> Forrest Gump didn't need any machine learning. No, it just needed a massive demographic the studios could exploit the Hell out of by taking a semi-popular novel and Spielberging all over it. You have a good point, but you used an imperfect film to demonstrate it. Forrest Gump was just about as cynical a film as you could imagine, and succeeded precisely as predicted. That kind of cynical filmmaking, explicitly targeted at demographics with earnings projections in place before casting begins, is precisely what this AI-driven concept will end up with at its worst. A better example would be the works of the Coen brothers, or David Lynch: Even if you don't like them personally the fact remains they have an audience, and they most certainly weren't made to order based on market forecasts. They represent the auteur theory in action, being the products of distinctive creative minds who impressed their equally distinctive visions upon them.
- akhilcacharya 8y ago
  Gump was Zemeckis!
  - saiya-jin 8y ago
    those 2 have quite a similar approach to movies AFAIK
    - SketchySeaBeast 8y ago
      Spielburg produced a lot of his films, so that would make sense.
- 1zee 8y ago
  >massive demographic the studios could exploit the Hell out of who? readers of a novel?
  - msla 8y ago
    Baby Boomers! The film Forrest Gump was basically Baby Boomers: The Movie. It was a Greatest Hits album of everything the stereotypical Boomer would consider great and memorable about their childhood and young adulthood.
ekianjo 8y ago
> No it isn't. Just make a good movie "Good" is not a single-variable problem, you know? Sometimes you have a good script, but you don't have the perfect cast. Sometimes you have decent actors and all but almost no marketing budget so nobody hears about your movie. There's so many things that can go wrong even with the best intentions when trying to make a good movie that "just make a good movie" sounds really naive. If it were that simple, every "good" director would make great movies one after the other. I struggle to find more than a few who could pull that off (and even then, that's highly debatable), among the the tens of thousands of directors who have been active in the past 50 years.
golergka 8y ago
Solid advice. On a similar note, if you want good software, you don't need fancy unit tests or type systems. Just don't write bugs.
- freediver 8y ago
  No, the correct analogy would be don't create crappy products and use ML department to predict how many people would buy it anyway.
  - JetSpiegel 8y ago
    > don't create crappy products Exactly, just don't write bugs. The same way Michelangelo sculpted David, he just removed the marble not part of the statue.
ameister14 8y ago
Tell that to Dredd.
eastbayjake 8y ago
Forrest Gump had a $55 million production budget and a $71 million marketing/advertising budget: https://www.nytimes.com/1995/05/25/movies/gump-a-huge-hit-still-isn-t-raking-in-huge-profits-hmm.html https://www.nytimes.com/1995/05/25/movies/gump-a-huge-hit-st...
Shorel 8y ago
I think the examples to watch are the many good movies that were not profitable and a decade later are "cult movies".
rurban 8y ago
That's unfortunately totally wrong. Good movies rarely get released, for decades. I'm professional film critic and I have a good idea of the data, because I analyze the feedback from festivals and industry experts, who decide what gets released and what not. From the Top 10 typically only 3 get selected. The selection criteria has nothing to do with the quality of the movie and rarely the quality of the trailer, even if it's more important. Most important are the names. Hollywood divides the Actors and Directors into A+, A, B+, B and C. Without an A name you'll have a hard time to get financing and proper distribution. 2nd most important criteria is the marketing power. How much money will be spent on marketing. If the distributor is not sure he'll try smaller tactics, like smaller releases, with the risk to earn much less. Most important is only the first weekend. Really good movies without names rarely exist, which can drive marketing by its own. Only these movies can survive a bad opening and raise slowly over the weeks to get profitable. Hollywood knows exactly the triggers beforehand, that's why they can get away with so many formulaic and simple movies still bring profitable. There's not much need for ML, just counting the numbers and trusting the experts is enough. In this case, if a trailer works for which group is certainly helpful though. With trailers you get better feedback than from test screenings, but not much better feedback than from experts.

yodon 8y ago

As with most simple recipies "just make a good movie" is unfortunately neither sufficient nor always the best path. Word of mouth on movies is far more likely to have a negative impact on theater going decisions than positive impact (if your friend says it sucks, you are far more likely to switch from go to no go than you would be to switch from no go to go if the same friend said it was awesome). In practice, what this means is that because positive word of mouth is so minimally impactful studios need to focus on high marketing value films rather than high story value films, because high marketing value films cause potential audience members to see a trailer or an ad and say "I'll totally go see that!" Yes, negative word of mouth can and will reduce the number who actually go see it, but final viewership is higher if you start with a huge population saying "I'll totally see that" than if you start with a "meh" response to the initial ads and try to grow viewership through word of mouth about how awesome your film is. Given these observations, Hollywood has made the rationally and financially correct but unfortunate for audiences decision to focus on marketability over enjoyability. These economics are quite different for entertainment as a service providers like HBO and Netflix, with the obvious results in quality of output.

smichel17 8y ago
The alternative is to build up a reputation, so when a new $studio film comes out, people are excited about it because $studio has a history of making good films. I think Pixar is/was a good example of this.
- NicoJuicy 8y ago
  Every studio that has focus on children is smart marketing. Children > parents. If you are a parent, don't you go to a child friendly restaurant that is reasonable good instead of one that is very good but not child-friendly.
- TylerE 8y ago
  That only works if the films are still good though. Pixar has had a couple of bombs, The Good Dinosaur for one.

selflesssieve 8y ago

> Just make a good movie. That seems obvious- but good movie does not mean high grossing blockbuster. Shawshank Redemption made less than the Bee Movie.

erichocean 8y ago
Not all good movies can have good/effective trailers made for them, and not all word of mouth can overcome a bad trailer.
msla 8y ago
Shawshank Redemption joins Office Space, Fight Club, and It's A Wonderful Life in the ranks of films saved by being shown on TV (whether via cable, DVD, or obsessive rebroadcasts over the holiday season before It's A Wonderful Life got snatched out of the public domain) after having bombed in theaters. They eventually found an audience even if the vagaries of the theater world kept them from making much in box office sensu stricto.
- melkiaur 8y ago
  Maybe they needed to get a bigger marketing budget. Now, if only the majors could predict which movie need to be marketed how...
- tokyodude 8y ago
  There are movies I want to see multiple times and movies I truly loved but only needed to see once. I would assume block blusters fit the "multiple times" category which is why they make so much money, because people go more than once. A drama like Shawshank Redemption (a favorite of mine) is not the type of movie I would have paid to go see multiple times in the theatre. Where as I saw The Matrix 6 times in the theatre back when it came out. Every friend who wanted to go see it I went with.

dagw 8y ago

No it isn't. Just make a good movie. Lots of "good movies" (using any definition you want) fail to find an audience. And even more mediocre movies make huge buckets of money. If your goal is to make money then just blindly making "good movies" is probably not your best strategy.

freediver 8y ago
Depends on your horizon.
- yodon 8y ago
  > Depends on your horizon Box office returns dominate the return for a film, set the license price for long term post-theatrical deals, and decay with a very fast exponential from the opening weekend take. It is extremely rare in modern times for a film to turn a poor opening weekend into a high long term return. The film market simply doesn't work that way however nice it would be if it did.
  - rjf72 8y ago
    I read the comment you're responding to a bit differently. I think most people can enjoy a really bad movie. Some t-and-a, explosions, a bit of CG - it's all good fun even all you can think about after the movie is how awful it was. But when really bad movies are all there is, they quickly lose their charm. Instead of being one off guilty pleasures, they become the movie industry. And that's about the point you stop watching movies at the theater. I think this describes many people's transition away from movies. The point of this is that pumping out crap may be optimal in the short run. But in the longrun you end up with suboptimal returns, even when profit is your one and only motivation. Incidentally the same story is true and also being played out in various other industries. Optimizing for next year's quarterlies is often not the same path to optimizing for earning the most money in the next decade.
TeMPOraL 8y ago
> If your goal is to make money That's one of the reasons why we have so many poor movies (the other reason is that making movies is hard). If your only consideration is making money, you'll produce whatever crap is cheapest to make that still makes reliable income. This is why e.g. consumer market is full of barely-working near-disposable products which existence is a pure waste of natural resources. Seriously. There's no rule that says that the most profitable product is also the best for the user. In fact, it's almost always the opposite, and A/B-testing things to death leads to the former type of product.

msla 8y ago

> Forrest Gump didn't need any machine learning. No, it just needed a massive demographic the studios could exploit the Hell out of by taking a semi-popular novel and Spielberging all over it. You have a good point, but you used an imperfect film to demonstrate it. Forrest Gump was just about as cynical a film as you could imagine, and succeeded precisely as predicted. That kind of cynical filmmaking, explicitly targeted at demographics with earnings projections in place before casting begins, is precisely what this AI-driven concept will end up with at its worst. A better example would be the works of the Coen brothers, or David Lynch: Even if you don't like them personally the fact remains they have an audience, and they most certainly weren't made to order based on market forecasts. They represent the auteur theory in action, being the products of distinctive creative minds who impressed their equally distinctive visions upon them.

akhilcacharya 8y ago
Gump was Zemeckis!
- saiya-jin 8y ago
  those 2 have quite a similar approach to movies AFAIK
  - SketchySeaBeast 8y ago
    Spielburg produced a lot of his films, so that would make sense.
1zee 8y ago
>massive demographic the studios could exploit the Hell out of who? readers of a novel?
- msla 8y ago
  Baby Boomers! The film Forrest Gump was basically Baby Boomers: The Movie. It was a Greatest Hits album of everything the stereotypical Boomer would consider great and memorable about their childhood and young adulthood.

ekianjo 8y ago

> No it isn't. Just make a good movie "Good" is not a single-variable problem, you know? Sometimes you have a good script, but you don't have the perfect cast. Sometimes you have decent actors and all but almost no marketing budget so nobody hears about your movie. There's so many things that can go wrong even with the best intentions when trying to make a good movie that "just make a good movie" sounds really naive. If it were that simple, every "good" director would make great movies one after the other. I struggle to find more than a few who could pull that off (and even then, that's highly debatable), among the the tens of thousands of directors who have been active in the past 50 years.

golergka 8y ago

Solid advice. On a similar note, if you want good software, you don't need fancy unit tests or type systems. Just don't write bugs.

freediver 8y ago
No, the correct analogy would be don't create crappy products and use ML department to predict how many people would buy it anyway.
- JetSpiegel 8y ago
  > don't create crappy products Exactly, just don't write bugs. The same way Michelangelo sculpted David, he just removed the marble not part of the statue.

ameister14 8y ago

Tell that to Dredd.

eastbayjake 8y ago

Forrest Gump had a $55 million production budget and a $71 million marketing/advertising budget: https://www.nytimes.com/1995/05/25/movies/gump-a-huge-hit-still-isn-t-raking-in-huge-profits-hmm.html https://www.nytimes.com/1995/05/25/movies/gump-a-huge-hit-st...

Shorel 8y ago

I think the examples to watch are the many good movies that were not profitable and a decade later are "cult movies".

rurban 8y ago

That's unfortunately totally wrong. Good movies rarely get released, for decades. I'm professional film critic and I have a good idea of the data, because I analyze the feedback from festivals and industry experts, who decide what gets released and what not. From the Top 10 typically only 3 get selected. The selection criteria has nothing to do with the quality of the movie and rarely the quality of the trailer, even if it's more important. Most important are the names. Hollywood divides the Actors and Directors into A+, A, B+, B and C. Without an A name you'll have a hard time to get financing and proper distribution. 2nd most important criteria is the marketing power. How much money will be spent on marketing. If the distributor is not sure he'll try smaller tactics, like smaller releases, with the risk to earn much less. Most important is only the first weekend. Really good movies without names rarely exist, which can drive marketing by its own. Only these movies can survive a bad opening and raise slowly over the weeks to get profitable. Hollywood knows exactly the triggers beforehand, that's why they can get away with so many formulaic and simple movies still bring profitable. There's not much need for ML, just counting the numbers and trusting the experts is enough. In this case, if a trailer works for which group is certainly helpful though. With trailers you get better feedback than from test screenings, but not much better feedback than from experts.