Frequently Asked Questions
General
What is the point of this app?
There are 33 movies and counting in the Marvel Cinematic Universe, plus over 10 seasons of Disney+ series and 20+ alternate universe movies which are now being tied in through the multiverse. You may have have seen a few of them, and there may be a few you want to see. When you watch the movies you've been wanting to see, you probably will want to have seen a few other movies first so you have proper context for the movies you're watching. This app will give you recommendations for which other movies to watch first.
What does MCU stand for?
The MCU is the Marvel Cinematic Universe, which is the setting of every film made by Marvel Studios, the Disney+ series and specials made by Marvel Studios, and many now-cancelled television shows made by Marvel Television. For more on this, see Wikipedia.
Where do Marvel movies like the X-Men films, Fantastic Four, Deadpool, Venom, etc fit in?
Those movies contain Marvel characters, but they were not made by Marvel Studios, and they do not take place in the Marvel Cinematic Universe. However, in recent years they have been crossing over with MCU stories by way of the multiverse. To include these movies, use the "more universes" dropdown.
A Disney+ Series isn't a movie. How do you count those?
Currently each live-action series is being treated as a single movie. The What If...? animated series is broken down by episode, because it is an anthology series in which each episode draws from different MCU movies.
What Does Each Algorithm Do?
The algorithms are currently undergoing revision - these FAQs will be updated when the revisions are complete.
Most Interconnected
Selecting the Most Interconnected option will get you a set of movies which have characters, plotlines, etc that are all highly interconnected with the movies you've seen, the movies you want to watch, and each other. Note: if the movies you want to watch are not highly interconnected with many other movies, you will likely end up being recommended movies which are highly interconnected with each other, and only somewhat related to the movies you've been wanting to watch.
Most Relevant
Selecting the Most Relevant option will get you a set of movies which have characters, plotlines, etc that are most relevant to the movies you want to watch. These movies may not be very related to each other; they are only guaranteed to be relevant to your selected movies. (Note: If you request more movies than the number of relevant movies to your selections, you will still be served the full number of movies you selected. The extra movies will be movies which are very interconnected with the movies relevant to your selected movies.)
Most Recent
Selecting the Most Recent option does not simply recommend you the movies released most recently to your selected movies. It gets rid of any movies that don't include characters, plotlines, etc that are present in your selected movies. For example, the 3 most recently released movies before Black Panther are Guardians of The Galaxy Vol. 2, Spider-Man: Homecoming, and Thor: Ragnarok. But nothing in these movies is relevant to Black Panther, so the Most Recent option will skip those and instead recommend you Avengers: Age of Ultron, Ant-Man, and Captain America: Civil War.
How does the code work?
The set of films is modeled as a weighted directed acyclic graph, where each vertex represents a film. Connections between the films are represented as edges between the vertices, with stronger connections having higher weights. The way the films are selected from this graph is dependent on which algorithm you choose.
How are the weights chosen?
The weights, which represent the relevance of each film to each other film, have been standardized in order to minimize subjectivity. For the full set of weight rules, click here.
How are ties broken?
Within the algorithm, ties are generally broken by the little microconnections between movies, which are given a weight of 1 for this purpose. However, sometimes 2 or 3 sets of movies end up being equally good based on the metrics used here. These ties are broken by a series of 3 metrics, each of which only comes into play if the previous metric fails to break the tie:
- If you selected all movies from the same series, which set of recommendations will contain the most films from that same series?
- Which set of movies has the highest weight if we ignore the movies you've already seen?
- Which movies have the highest Rotten Tomatoes scores? (As this is designed for more casual viewers, if 2 sets of movies are equally relevant, you might as well pick the movies considered the best.)
- Which movies are connected to the most other movies, including movies you aren't going to be recommended?
Does this project use Artificial Intelligence (AI)?
No. While AI models can give different results every time and rely on mountains of training data, this project uses precisely defined mathematical graph theory algorithms which will give the same results every time.