UPDATE – See bottom for an important update given a bug found and fixed
“Fantastic Four” leads the new set of movies at our disposal in Fantasy Movie League this week and an interesting set of pricing gives legitimacy for some combinations that do not use a full set of eight screens (and the $2M per unused screen penalty). At the low end of our choices, Jurassic World is offered at FB$16, which is atypical given that most weeks there are a variety of choices under FB$8 with which to fill the final slots of each combination. That isn’t the case this week, which means those penalty-laden combinations deserve consideration even though I’m not picking any myself below.
Also new this week is a tool I created to help analyze Fantasy Movie League entries. My nerdy bin packing estimate comparator uses ProBoxOffice.com estimates as a baseline and offers anybody who wants to use it a set of sliders so you can test different scenarios for possible outcomes. Feel free to try it out yourself:
Best Performer Analysis
Since my analysis relies so heavily on ProBoxOffice.com and they only list their predictions for the top 10 box office producers any given week (although they did 11 this week), I was taken completely surprised last week by “Mr. Holmes” taking home the Best Performer bonus. We have a potentially similar situation this week with “The Gift” which has had nice reviews but is STX Entertainment’s first distributed film, so it carries some risk.
If you take the ProBoxOffice.com estimates at face value, “The Gift” will be the Best Performer:
However, if that estimate is off by even $255K, “The Gift” drops and “Pixels” becomes you the Best Performer:
By similar logic, if you keep “The Gift” steady at its original estimate and “Pixels” gains just $180K, “Pixels” gets Best Performer. This demonstrates the power of my nerdy bin packing estimate comparator given the ease at which you can try out these different possibilities.
If you play with those scenarios enough what you’ll find is that in order for “Southpaw” to emerge as a Best Performer candidate, both “The Gift” and “Pixels” would have to fall off 15% of their estimates. Even if “Southpaw” does 15% better than its estimate and “The Gift” stays steady at it’s estimate, “Southpaw” can’t take Best Performer. Similarly, “Trainwreck” and “Vacation” would have a tough time passing both “The Gift” and “Pixels”.
With that, combinations that maximize “The Gift”, “Pixels”, or hedge between them make the most sense.
My Fantasy Movie League Picks, Week 12
The fiendish statisticians at the Fantasy Movie League have made this a tough choice this week. Do you bet on the first movie for a new studio in “The Gift”, should that fail do you think a horribly reviewed movie in “Pixels” can hang on, or do you hedge between those two situations? It’s not an obvious call.
#1 – I’m not messing around this week after getting burned last week hedging between two Best Performer candidates and I’m going to trust ProBoxOffice.com’s estimates. If you do that, the best combination is:
#2 – If you don’t believe in “The Gift” at all and want to go all in on “Pixels” as Best Performer, this is your best bet, although it is risky given that it is comprised of all poorly reviewed movies, save one:
#3 – A variation of #2 that doesn’t rely on “Fantastic Four”, which as I type has a 21% Rotten Tomatoes rating:
That’s all for this week. Feel free to put your thoughts about my nerdy bin packing comparator and combinations you found useful in the comments or get my attention on Twitter @nerdguru.
8/6/15 9:30a PDT UPDATE
Crowd-sourcing is awesome. Within these first 24 hours of launching my nerdy bin packing estimate comparator, I’ve received great feedback, but none greater than this one:
@nerdguru Am I missing something or wouldn’t 3 MI5 and 5 Gift be the best lineup based off of projections?
— Robby Hill (@robbyhill20) August 6, 2015
Thanks to Robby kindly pointing out my error, I adjusted the candidate combinations so that they include anything that costs above FB$900 and is in the top 6,000 when scored based on the default estimates. That gives a much better data set to work with. Thanks again to Robby for sparking this improvement and I’m sure there will be more to come.
As a result, I’m amending my pick to be as smart as Robby: