Now that we can tag emotions to a scene, we can look up the various axes and plot on a graph changes to those axes. Like JoyToSadness, we should be able to see rise and fall of that axis from all the hundreds of tagged emotions. This can help us identify common patterns, and check if these patterns differ between genres, etc. Then when using the tool to generate new content, we can randomly select target emotions based on probability based on known common patterns for the current genre or dramatic question.
This should line up with the Sequences, like All Hope Is Lost should have a lot of sadness. But, it also provides a lot of flexibility to deviate from the Save The Cat formula, which is a high level assumption of what emotions should occur at a given point. Instead of forcing every story into the Save The Cat sequences, we could instead identify common emotional fingerprints for the genre + dramatic question.
Now that we can tag emotions to a scene, we can look up the various axes and plot on a graph changes to those axes. Like JoyToSadness, we should be able to see rise and fall of that axis from all the hundreds of tagged emotions. This can help us identify common patterns, and check if these patterns differ between genres, etc. Then when using the tool to generate new content, we can randomly select target emotions based on probability based on known common patterns for the current genre or dramatic question.
This should line up with the Sequences, like All Hope Is Lost should have a lot of sadness. But, it also provides a lot of flexibility to deviate from the Save The Cat formula, which is a high level assumption of what emotions should occur at a given point. Instead of forcing every story into the Save The Cat sequences, we could instead identify common emotional fingerprints for the genre + dramatic question.