Project Overview: This dataset was used as a learning material for an Udemy course on PivotTables. I decided to explore a little further and used what I learnt to look deeper. Some of the concepts I applied were “Rank”,
Which genre of films reaped the highest profits in 2015?
Comedy films generated the most profits in 2015 followed closely by Adventure films. When I dragged in Gross Revenue and Budget, it seemed like Comedy films earned the higher profits than Adventure because of its relatively low budget (almost 3 times lower!).
Crime films, however, made a huge loss of $41 million.Their budget is over 100 million but their gross revenue is low.
What year was the best for Action films? For Adventure? For Comedy?
2014 was the best year for Action films with a profit of over $900 million,
contributing to 21.3% of total profits earned by Action films since 1920.
Meanwhile 2012 was the best year for Adventure films, with 12.03% of the total profits. However, it was not a huge jump from its 2010 performance which pulled in 11.31% of total profits by Adventure films.
Lastly, I looked at the performance by Comedy films throughout the years. Oddly, the best year was almost two decades ago; their highest-earning year was in 2003 and it contributed to 41.3% of the total profits by Comedy films. Even though we saw earlier that Comedy films was the best genre in 2015, turns out that when we looked at the performance over the years 2015 is only ranked 9 out of 45 years.
Which countries are each genre of film most popular in 2015?
U.S. dominated the box office for all genres except Thriller in 2015.
Do longer films have a correlation with budget?
I was wondering whether there would be any correlation (NOT causation!) between duration of films and budget. Logically a longer film means more scenes which might imply higher budget required right? When I did a very simple PivotChart for films released from 2011 onwards to 2015 inclusive, turns out the relationship is stronger than I thought.
A simple CORREL on the two variables showed that the correlation is pretty strong at 0.72. The relationship still holds when slicing for Documentary (CORR = 0.81) and Fantasy films (CORR = 0.74).
Despite the strong correlation, the relationship may not be causal (yes econometrics time). Firstly, longer duration do not always mean more scenes. For instance, monologue or dialogue scenes can be longer and these scenes may not always cost more (as opposed to explosive actions scene we always see in Hollywood films). Hence, omitting information about the type of scenes in the movie may cause some bias. Secondly, even if there are more scenes those scenes may not always cost more to produce. Without further in-depth analysis, there is no way I can ascertain this correlation as a causal relationship.
Do Facebook likes affect profits for films released in 2015?
I wondered if social media, particularly Facebook likes would matter when it comes to profits earned. I created a new calculated field which summed the Facebook likes by cast, actor and director and I then rank each film from 1 (being the most likes) to 15 (with the least likes). I then did the same for profits. The line graphs below can be difficult to interpret but ideally, if there is a film with high FB likes it could mean there are many people supporting or wanting to support and thus, generate a high profits. In this scenario, both lines should increase or decrease together as they would have a high correlation. There could be reverse causality but regardless, I am not interested in causal effects here, just correlation and in this case there appears to be no or weak correlation here.
Lessons learnt:
- PivotTables really are quite amazing and I have to admit there is so much more to PivotTables that I didn’t know until I took the course. So I guess knowledge is infinite and there is never an end to learning?
- Always always understand what you are interested to explore before doing the actual dirty work.
Cinematic Frame of Mind 