Chris Mooney is proclaiming that Expelled isn't a flop and is a success (see here) while Greg Laden has said nope (see here). I'm a bit demented and I like statistics so I wanted to see what the relationship was between the number of theaters of film opens in and the amount it grosses in its first weekend. Above is that graph on a log scale (lots of variability so this helps it look nice) and shows the data for political documentaries (the data Chris used to proclaim Expelled's success). This data is for the top 16 minus expelled. The R-squared is pretty high and the relationship is fairly linear (the dashed lines are confidence intervals for the regression).
I'm guessing that expelled isn't as much of a flop as I thought..., apologies for my late night math error and thanks to Duae Quartuncia for pointing it out.
oops... this part was wrong --> As you can see based on the number of theater's that expelled opened in, it is a flop of extraordinary proportions. Although this isn't the best way to do it (I don't have time to sit and play with this), the residual of the expelled data is -16 standard deviations away from the predicted line. I'd say that's a flop of gargantuan (well at least significant at P <0.000000001)>
6 comments:
Wow. Visual impact of that is dramatic. Kudos.
Wait a minute. That can't be right! I take it back.
The data point you have for expelled has the number of theaters as about 10^3, which is right. But it has the opening weekend as about 10^0.3, which is about $2.
You should have the opening weekend as about 10^6.5.
You've made some kind of horrible mathematical error here.
Straight lines on log-log plots always make me wince. . . is that saying that the Expelled point is half an order of magnitude below the regression line?
Note: I have not had my caffeine yet today.
It is -0.31 below the regression line. Which puts it earning about 1/2 of what would be expected. The residuals here are logs of the ratios of obs/predicted so 10 raised to the -0.31 is 0.49. Based on the regression the movie should be grossing about 6.415 million dollars.
S. Walker says... "Which puts it earning about 1/2 of what would be expected."
But that is only true if the standard of comparison is limited to the 16 top grossing documentaries.
If you put all the data from the table in the plot, it is very close to the line indeed; just a bit below.
Basically, this is not surprising to me. The producers of Expelled have thrown an enormous amount of money at this, with four PR firms. They pushed to open in as many theaters as they could, I think without much regard for the profit. They just wanted an impact, no matter what it cost.
With a ready supply of willing dupes, this was bound to be something of a success in absolute terms. The total return, and the total number of theaters, both put the film high in the rankings for documentaries. The rate per theater is fairly low; but that is not a statistic used much in the popular press.
The real down side for them has been the reviews. It's been panned across the board by reviewers, and I think the credit for that goes primarily to NCSE and all the bloggers who took up the matter and made the flaws in the film readily available to anyone who looked.
Special hat tip to PZ Myers; whose blog on his expulsion is still showing up as #1 in the Google blog search for "Expelled". He and I disagree fairly strongly on the approach to religion and science; but he's framed the Expelled matter to perfection. His message on the film has consistently been "This is a dreadfully bad movie made by dishonest scoundrels"; and that is the frame that got picked up in the professional film reviews.
I suspect the film will do badly in coming weeks. I hope so. But it is a mistake to blame the limited success of this film on the noisy atheists giving it free publicity... an idiotic claim being suggested in some quarters.
The success it has achieved was hard won by very aggressive marketing. The blogsphere reaction has countered that by making it easy for the middle of the road folks to find out about its flaws.
I think the science community has handled the matter very well and very effectively.
Post a Comment