Presentation on theme: "Comic Book Movies By Anthony Borkowski. Why Base Movies Off of Comic Books? Comic books are an American art form that many children in each generation."— Presentation transcript:
Why Base Movies Off of Comic Books? Comic books are an American art form that many children in each generation have grown up with. Characters such as Spiderman, Superman, Wonder Woman, or She-Hulk represent the values of our society, so it is only natural for a movie to be made about them and further glorify the characters.
Why SHOULD We Base Movies Off of Comic Books? Simply put, movie producers create movies based off of comic books because they know the movies will make money. But the real question is, how much money will the movie make?
How Much Money Will the Movie Make? I honestly do not know how much a movie based off of a comic book will make. I cannot predict the future, I am a statistics student. BUT, I do know a way where we can estimate the amount of money the comic book movie will make: Multiple Regression!!!!!
Multiple Regression? To create multiple regression model, we will first need to select a random sample of existing movies that are based on comic books. However, not all movies are eligible. Since many comic book movies spawn sequels and trilogies, the success on the sequels could be influenced by the original. So, for this, we will only use the FIRST movie in a series. The movie also needs to be live-action (not animated) and the comic English in origin. To define “series”, we mean storyline/plot continuation. The Batman movies, for example, from the 90’s tie into one another, so they count as a separate series that DOES NOT include Batman Begins (the 2005 film). So, both Batman films were eligible for selection
Selecting Movies Using the internet sites Internet Movie Database and Wikipedia, I compiled a list of movies based on comic books. I alphabetized the list and assigned each movie a number from 1-100 starting from the top of the list down. After writing numbers on small slips of paper and mixing them up, I randomly drew 10 slips and chose the comics corresponding to the numbers on the slips.
Initial Predictors Initially, I looked at the following factors to try and predict the amount of money comic book movie would bring in.
Initial Predictors (continued) Years in publish Number of shows the series had BEFORE the film. Number of games for the series BEFORE the film. The number of times I (Anthony “WoodStock” Borkowski) saw the film. The number of films the movie director directed BEFORE this film. The month number the film fist came into theaters. The day number the movie came into theaters.
End Result Predictors After entering all of my data into Fathom, I formed a linear regression model. I arranged the predictors based on P-value and deleted the greatest values until the remaining values were below.05 (5%). The following is a list of the predictors I ended with.
End Result Predictors (continued) Number of games for the series BEFORE the film. The number of times I (Anthony “WoodStock” Borkowski) saw the film. Years in publish The day number the movie came into theaters.
Conditions To validate our multiple regression model, we had to check the following conditions: Straight Enough Independence Does the Plot Thicken? Nearly Normal
Straight Enough Condition Unfortunately, Fathom does not have the ability to create a scatter plot of residuals for multiple regression. For this condition, we will have to assume that the condition is met.
Independence Condition Since the income of one comic book movie will not affect the income of another, there is no reason to think they would affect one another. Also, I chose the movies randomly (as described earlier). This also helps ensure independence.
Does the Plot Thicken? Condition This condition is met due to the fact that all of the individual scatter plots comparing two different variables are all randomly scattered.
Nearly Normal Condition (continued) The data for “Number of Games” is not symmetrical in the histogram nor is it linear in a normal probability plot. If we choose to use it, we should do so with caution. However, after removing “Number of Games” I found the R-square value dropped over 60 points. For my analysis, I kept “Number of Games” and did so with caution.
Nearly Normal Condition (continued) Everything else is either a fairly linear normal probability plot or is unimodal/symmetric histogram so they meet the conditions. Note: Also, be careful of the residual plot for the variable “day”. Some could interpret it as NOT being linear, so we should also proceed with caution when using it as well.
So.. What’s that mean? Under these conditions, and proceeding with caution in regards to “Number of Games”, the multiple regression model is appropriate. The following is the output from Fathom.
Incase you can’t see it, The estimated multiple regression formula is … ^ (Predicted) income = 830,476,102.309 + 38,984,539.9564(games) – 80,223,449.7395(Saw) – 9,745,083.07688 (publish) – 13,067,131.0898(day)
Interpretation The R-squared value is about.881, so my variables predict about 88% of the predicted movie income of a comic book movie. Looking at the bar chart in the multiple regression model, the “Number of Years in Publish for the comic book” variable seems to be responsible with about 65% of the income, followed by about 20% from the day the movie was released on.
Interpretation (continued) The Number of Games produced BEFORE the movie accounts for about 17% of the income, while the number of times I saw the film accounts for about 2%.
Interpreting the Formula If the variables “Saw It”, “years in publish”, and “day” remain at a constant, the predicted income for a comic book movie should be about $38,984,539.9564 greater. If all other variables remain at a constant, the predicted income will be about $ 80,223,449.7395 less for every time that I see the movie.
Interpreting the Formula If all other variables remain at a constant, the predicted movie income should be about $9,745,083.07688 less for every year the comic book was in publish before the movie came out. If all other variables remain at a constant, the predicted income for a movie should be about $ 13,067,131.0898 less based on the day number it is released.
Interpreting the Formula If all the variables have 0 values, the predicted income for a comic book movie should be about $ 830,476,102.309.