Feedback Request: Starting a Movie Blog

Introduce yourself to the community or chat with other users about whatever is on your mind
ShogunRua
Posts: 3449
0 Ratings
Your TCI: na
Joined: Sun May 17, 2009 3:18 am

Re: Feedback Request: Starting a Movie Blog

Post by ShogunRua »

CarsonWid wrote:ShogunRua, you make some good points.

In regards to studios and their 50 variable equations, I think most of that has to do with the financial success of a movie and very little to do with the critical success. I had originally started this as a prediction of financial success but I didn't find it very interesting. I had various seasonal indices and factors in it but I wasn't getting good results and as mentioned before, wasn't really what I was trying to do.


It's also a far more meaningful, objective, and less fuzzy indicator of success than the RT score or IMDB rating.

CarsonWid wrote:I have a limit on what data is available to me, so as much as I'd like to add more variables I also have to consider that this is just a for-fun after work project of mine.


Most of this data is readily available. Production budget, P&A budget, runtime, time of the year film is released, etc.

CarsonWid wrote:I've thought of things such as release date, budget, production company, etc. but none of these really held a strong correlation to the critical success of the movie.


Have you actually run the statistical analysis to determine this, or are you just guessing? The whole point of mathematical investigations is NOT to just trust your gut instincts, which can frequently be wrong. If you want to ever discover anything meaningful in your analysis, you need to start approaching every problem this way. In this particular case, I also strongly intuitively disagree that the production company does not have an influence on the critical success of the movie.

Maybe if you average it out they don't, but my feeling is that something from the Weinstein Company or Sony Picture Classics is going to generally be higher rated than, say, the average offering from Relativity Media or Warner Brothers.

Carson Wid wrote:I'd love to have a job running analytics of this nature but alas I do not have it, nor do I have the tools or databases to run this at a super effective level.


This might have been true back in the 1980s. Nowadays, no one has this excuse. Even the weakest laptop can run all these Monte Carlo simulations with ease, and there is a wealth of free resources and software online.

Carson Wid wrote:My objective for this blog is to show data and analytics at a high level


Which you have not.

Carson Wid wrote:in a simplified way that people who don't like math can digest it.


Getting people who don't like math to like math seems counterproductive, no?

Carson Wid wrote:The objective is not to 100% accurately predict a movie's score,


Have you heard of "confidence intervals"? You speak very-unmathematically for someone supposedly eager about the subject.

Carson Wid wrote:So, final point/question: What easily-accessible variables do you think would be most relevant when predicting a movie's critical success?


This is a horrible attitude to have for a researcher or analyst. You shouldn't be asking someone else; you should be investigating this yourself.

I still don't understand the point of your project. Predicting movie success in a manner far inferior to what even the most brain-dead 19 year-old quant can cook up in a couple of hours playing with MATLAB?

CarsonWid
Posts: 50
3424 Ratings
Your TCI: na
Joined: Sun Sep 13, 2009 5:56 pm

Re: Feedback Request: Starting a Movie Blog

Post by CarsonWid »

Geez, tell me what you really think. You'd think I'd have insulted your family and way of life the you respond to this.

My confidence intervals are somewhere in the 70-75% range. Not exactly something to brag about.

I tested some of those other variables (seasonality, production company, and budget) and got lower correlations (0.5 to 0.6) than I did for Directors and Actors. I'm only starting to use other creators (Cinematographers, Designers, Editors, etc.) so I can let you know later how those compare. I didn't create confidence intervals for those predictions, just correlations. Quite honestly, I couldn't expect any of those 3 variables to predict whether it's good or not, otherwise the Hobbit movies should have been phenomenal.

Getting non-math people to like math? No, getting non-math people to pay attention to numbers and stats that come out of math. People like numbers and stats and rankings so I don't think it's that tough of a thing. There have been positive responses in this thread alone and I think it's a worse attitude that you have that non-math people shouldn't be pushed to pay attention more to simple stuff life this.

IMDb and Rotten Tomatoes aren't the best indicators of whether a movie a good, I agree, but they are measures of success that involve a lot of different people from a different perspectives. It's a standard that can be compared against whether you agree with the rankings or not.

In regard to tools and resources. Yes, I am limited. Mainly by the data entry component. If you know of databases that can be easily exported then yes, I can do a lot even with just Excel...but as of now it's taking me a long time to compile information and enter it manually. Even this so-called easily accessible information such as budget and run-time is hard to get for movies that are 3-4 months away from release date. There would be less value to predicting movies coming out sooner than that because by then there have been screeners and people who have actually seen the movie so I could just listen to them.

TL;DR - Hey don't be a jerk. I'm obviously just someone who likes numbers and stats who is doing this at like an hour or so a week. I know many different reasons why this isn't mathematically correct. You could use your knowledge of this subject in much more productive ways. For example, just be like 'hey some other variables might be budget, run-time, other contributors, etc....and some metrics you might want to use would be confidence intervals, t-tests, and here is a good program to help you do that'. That's pretty much all I was looking for, not a lecture on why I'm mathematically dumb and shouldn't be attempting this and how 19 year olds could do this better than me,

CMonster
Posts: 689
1444 Ratings
Your TCI: na
Joined: Thu Mar 24, 2011 12:22 am

Re: Feedback Request: Starting a Movie Blog

Post by CMonster »

Be careful Carson, pretty soon he is going to start dropping lines like "I'm an economist." or "I'm a scientist."

ShogunRua
Posts: 3449
0 Ratings
Your TCI: na
Joined: Sun May 17, 2009 3:18 am

Re: Feedback Request: Starting a Movie Blog

Post by ShogunRua »

CarsonWid wrote:Geez, tell me what you really think. You'd think I'd have insulted your family and way of life the you respond to this.

My confidence intervals are somewhere in the 70-75% range. Not exactly something to brag about.

I tested some of those other variables (seasonality, production company, and budget) and got lower correlations (0.5 to 0.6) than I did for Directors and Actors.


I think you're doing something wrong. r = 0.5 or 0.6 is a very high correlation for any one variable. Probably your sample size is too small.

Carson Wid wrote:There have been positive responses in this thread alone


You're getting politeness confused with genuine interest.

Carson Wid wrote:and I think it's a worse attitude that you have that non-math people shouldn't be pushed to pay attention more to simple stuff life this.


No, it's quite a good attitude, because your approach satisfies NEITHER group. The people who dislike math will still dislike/not understand your diluted misuse of statistics. And the people who understand math will dislike it, too.

Carson Wid wrote:IMDb and Rotten Tomatoes aren't the best indicators of whether a movie a good, I agree, but they are measures of success that involve a lot of different people from a different perspectives. It's a standard that can be compared against whether you agree with the rankings or not.


You completely ignored the substance of my arguments. Your counter is essentially "yeah...but still!"

Carson Wid wrote:TL;DR - Hey don't be a jerk.


Oh look, another sensitive, special snowflake who has been told all his life by parents and teachers that he is special and awesome!

I thought you wanted genuine, helpful feedback. Guess not. You wanted compliments and attention. With that attitude, your blog will die off in a few months without garnering any interest and you will have learned far less than you could have from reading an introductory statistics textbook. But hey, at least you will still have your self-esteem!

Carson Wid wrote:I'm obviously just someone who likes numbers and stats who is doing this at like an hour or so a week.


This is pathetic. If I was passionate about something, I would devote multiple hours per day to it, not an hour a week.

Carson Wid wrote:For example, just be like 'hey some other variables might be budget, run-time, other contributors, etc....and some metrics you might want to use would be confidence intervals, t-tests, and here is a good program to help you do that'.


I gave you the broad strokes. As for the specifics, I'm not going to do this project for you. I'm really curious what age and profession you are to have such an attitude. Even undergrads doing research for a professor wouldn't expect the mentor to do the project for them.

Carson Wid wrote:That's pretty much all I was looking for, not a lecture on why I'm mathematically dumb and shouldn't be attempting this and how 19 year olds could do this better than me,


16 year olds, 16 year olds.

CarsonWid
Posts: 50
3424 Ratings
Your TCI: na
Joined: Sun Sep 13, 2009 5:56 pm

Re: Feedback Request: Starting a Movie Blog

Post by CarsonWid »

Update: http://movienumberz.tumblr.com/

So I took some feedback in regard to design for this one. Used some of the technical people instead of just actors. ShogunRua made a good point about not satisfying either the hardcore data people or the non-math people (y'know, cuz there is only two groups of people in the world) so I'm leaning more towards the non-math side...showing some of the trends and people involved that you might not know by the trailers and media releases.

For my next post I'm going to do a The Revenant vs. Bridge of Spies thing and might work on adding a category or two.

paulofilmo
Posts: 2586
2428 Ratings
Your TCI: na
Joined: Mon Apr 07, 2008 3:40 pm

Re: Feedback Request: Starting a Movie Blog

Post by paulofilmo »

Much easier to read! and I love the logo.

The typeface terrifies me. The kerning is strange, and the numbers sometimes seem as though they're wobbling on a different line.
Maybe unify the design of the graphicy bits. Have a theme of shapes/colours.

The 'have they worked together' section basically boils down to Rush (2013). So maybe just have a title like 'Previous connections', and then show Rush, with its vitals. That way ppl might see the poster and think, hey I liked that, maybe i'll like this.
(but you have a red line from the editor to the DP.just to fyi - both worked on Rush)
(you also have red line from the designer to the DP when they both worked on 28 Days Later!) -- I feel like computers should be doing all this stuff for you.


Couple of extra thoughts

The composer for this film has done some interesting work. Creepy little horrors/thrillers. I'm sort of curious why Howard would choose him when he's used fuck-off hero composers like H.Zimmer in the past.

I didn't give a shit about the film until I saw Ben Wishaw and Cillian Murphy were in it (also CM was in 28 Days Later -- ppl be talking). These two are probably the UKs most talented young male actors.

The DP has done some interesting stuff. You mention Slumdog, but it would have been cool to know he'd been a D.Boyle regular, as well as some legends of Dogme 95. Just giving me an imdb average doesn't give me a sense of the visual potential of this film as compared to my recollections of Slumdog, Festen. Maybe a picture of the talent, and then a row of movie posters, perhaps with a mouse-over or somesuch for the data analysis. Click-on to imdb page.

Have you modelled the genre bit yet (modelled-? whatever the word is)? The data table scares me. Is there a way to present it visually?

I don't understand the value of the predictions, but maybe i'm stupid.
Why won't the film take any risks? I think this sort of value judgement thing should be self-evident. Maybe personalise/brand it more simply: 'numberz thumbz up!' or give it so many hashes: (lolmstlyjkingdntdothis) (butalsomaybedothis)

ImageImageImageImage

(but, you know, in a way that doesn't look like the decaying remnants of leper dick)


^this is all brain-farting, immediate thoughts. I wouldn't spend this much time on it if I wasn't excited that you're doing things. I could be reading john donne, or masturbating, or hitting my children.

theficionado
Posts: 293
1908 Ratings
Your TCI: na
Joined: Wed May 02, 2007 4:31 am

Re: Feedback Request: Starting a Movie Blog

Post by theficionado »

Might be worthwhile to check out how previous research has approached this and similar problems. For instance:

http://haas.berkeley.edu/groups/mors/pa ... entity.pdf

mattorama12
Posts: 887
3086 Ratings
Your TCI: na
Joined: Mon Feb 13, 2012 3:05 am

Re: Feedback Request: Starting a Movie Blog

Post by mattorama12 »

Just read an interesting article that discussed a model for predicting profitability. Interesting to note that from the examples given, it seems like this model is much more pessimistic than the one Shogun referred to earlier (though that could be selection bias on the part of the writer). Unfortunately, the study that's cited is behind a paywall.

http://www.futurity.org/movie-algorithm ... -1104882-2

paulofilmo
Posts: 2586
2428 Ratings
Your TCI: na
Joined: Mon Apr 07, 2008 3:40 pm

Re: Feedback Request: Starting a Movie Blog

Post by paulofilmo »

Ahh, it's sad he didn't crack on with this. His In the Heart of the Sea prediction was close.

Could be pared-down to the two predicted ratings. Perhaps highlight an interesting connection. Simple enough to create intrigue on social media.

CosmicMonkey
Posts: 594
1271 Ratings
Your TCI: na
Joined: Sun Apr 05, 2009 6:52 pm

Re: Feedback Request: Starting a Movie Blog

Post by CosmicMonkey »

I understand little about math so I'm afraid I can't give any constructive criticism, but I love it when other people more mathematically-inclined than me apply math and statistics to real world topics like pop culture and politics. Anyways, I've started following you on tumblr, keep up the good work!

Post Reply