HOLLYWOOD STOCK EXCHANGE
THE ENTERTAINMENT MARKET
THE ENTERTAINMENT MARKET
Only to see what relationships there are between early indicators and delist.
Anything to do with genre and demographics is based on my judgment - so it's subjective and not scientific. My methodology is in the Info Worksheet. The sub-genre info is incomplete. I haven't gone through and identified every sci fi movie, for example.
i think genre and target audience overlap a bit. Males see more action movies, adults see more drama, women see more romance, families see animated etc. If I had to pick one factor I'd go with target audience. The key thing to understand is the moviegoing habits of each group. Families take their kids to the movies on the weekend, and are more likely to pick a weekend that suits their schedule. Teenagers go see whatever is opening on Friday nights and by the next they've moved on etc
i wouldn't use theatre counts to predict delist as I only have data on theatre #s for the opening weekend. As movies that lose the most business from OW to W/E 2 are the ones that lose the most theatres later in their run, I'd use a factor like Friday to Friday, or weekend drop to give a better idea of the future.
i wouldn't worry about movie length either. It matters on opening weekend, especially for sell-out movies. But for predicting delist the audience interest in seeing the movie is the big driver. Titanic, Avatar, etc, while there are plenty of awful 90-minute movies that no one went to see at any of the ten showings per day.
Announcing the beta version of the HSXsanity box office database, with adjust and delist estimators for you to try. Feb 24, 14:20
um, holy crap this is amazing? {nm} Feb 24, 14:38
Wow, good stuff!! {nm} Feb 24, 18:35
finally had a chance to take a peek...holy smokes! {nm} Feb 24, 22:41
oh...and thanks for doing this! {nm} Feb 24, 23:34
That is awesome. You've really outdone yourself. Thanks! {nm} Feb 24, 22:43
2 competitor kaigee {nm} Feb 25, 06:58
Hmm, every interesting. I could implement a nice regression model on all that data, and do BO predictions. {nm} Feb 25, 07:01
So you've run Linear Regressions? I guess it a matter of identifying the right set of features to get the best fit. Feb 25, 15:33
Only to see what relationships there are between early indicators and delist. Feb 25, 17:04
(So I think you can do it with two factors - target audience and level of audience interest) {nm} Feb 25, 17:07
(Actually, Three - I forgot "time of year") {nm} Feb 26, 03:52
Wow..that's a lot of work...thanks for the effort...much appreciated! {nm} Feb 25, 09:51
WOWZER - That's Impressive. A Tip O' The Hat. {nm} Feb 25, 19:03
