is a start. If we think (and I do) that seasonality will have the greatest effect on delist multiple, then the next step is to figure how how to normalize it for your genre study. There could be other factors (box office, demographics) that we can layer in layer. But for starters, I would create a genre for African American comedies and African American dramas. I think those will likely have similar delist performance and may be different from genres that are mainstream. I would also create an arthouse genre, that will likely include films like nebraska that had a wider release further into its release. King's Speech is another example. One could argue that once a movie hits 20 mil at the box office, its no longer arthouse, but I suspect that these titles will have a higher delist multiple so they may need their own genre.
I looked at Roger's average delist by month and caculated values based on the assumption the average delist is 2.7. One thing that looks screwy is November, since you expect Thanksgiving titles to have a higher delist (that may or may not be true, thats my assumption). The other thing is I would probably reduce the data to the last five years. And maybe look at seasonality by week (i usually use month, but that november number looks counter-intuitive so we may need to look at seasonality by week).
Finally, I would throw out any genre with 1 or 2 titles. Especially docs.
I tend to let the data define the groupings, but in this case there are already established genres, but the data may allow us to create new genres (hence my recommendations to create AA Drama and AA Comedy and Arthouse). One new genre could be "blockbusters" if non kids titles that open over $50 mil (random threshold) they may have similar delist regardless if they are sci-fi or action.
| Month |
|
| Average Delist Multiplier |
2.7 |
| |
|
| January |
|
| 2.69 |
1.003717 |
| |
|
| February |
|
| 2.46 |
1.097561 |
| |
|
| March |
|
| 2.63 |
1.026616 |
| |
|
| April |
|
| 2.55 |
1.058824 |
| |
|
| May |
|
| 2.57 |
1.050584 |
| |
|
| June |
|
| 2.87 |
0.940767 |
| |
|
| July |
|
| 2.92 |
0.924658 |
| |
|
| August |
|
| 2.75 |
0.981818 |
| |
|
| September |
|
| 2.46 |
1.097561 |
| |
|
| October |
|
| 2.69 |
1.003717 |
| |
|
| November |
|
| 2.34 |
1.153846 |
| |
|
| December |
|
| 4.27 |
0.632319 |