Interesting article from 2013 in a statistics journal.
Gala Brand
Posts: 1,172MI6 Agent
The author concluded that Bond 24 (Spectre) would cost $250 million, have an average IMDB rating of 6.8, and make $865 million worldwide.
Well, that's a little eerie.
Apparently, the only two factors that really contribute to a Bond film's financial success are the budget and the actor playing Bond.
http://chance.amstat.org/2014/04/james-bond/
Well, that's a little eerie.
Apparently, the only two factors that really contribute to a Bond film's financial success are the budget and the actor playing Bond.
http://chance.amstat.org/2014/04/james-bond/
Comments
Appears as if he averaged out DC's films gross. The author was not expecting SPECTRE to be an outlier like Skyfall
"I am not an entrant in the Shakespeare Stakes." - Ian Fleming
"Screw 'em." - Daniel Craig, The Best James Bond EverTM
Actually, the author describes his methodology and it sounds a tiny bit more complex than a simple averaging:
"We construct the multivariate analysis of variance (MANOVA) table to assess the significance of each independent variable. We compute the following four multivariate test statistics: Wilks’ lambda, Pillai’s trace, the Hotelling-Lawley trace, and Roy’s greatest root. Using a significance level of 0.05, we perform backwards elimination by removing the least significant variable based on the MANOVA results. Using this approach, we end up with only the Bond actor and the adjusted budget as significant predictors of the bivariate response. In fact, our decisions based on the p-values for each predictor is consistent across the four test statistics. Since our decisions do not vary across the four tests, we are confident in the results.
Our final multivariate regression model treats the adjusted worldwide gross and average rating as the response vector with Bond actor and adjusted budget as predictors. We next analyze the residual vectors for each regression to assess the normality and constant variance assumptions. The Shapiro-Wilk test for normality applied to the residuals yields p-values of 0.2780 and 0.2352 for the adjusted worldwide gross and average user relationships, respectively. The corresponding results from the Levene test for constant variance yields p-values of 0.4533 and 0.4039. Therefore, we claim that the normal regression assumptions are appropriate for this model."
Don't you agree?
"I am not an entrant in the Shakespeare Stakes." - Ian Fleming
"Screw 'em." - Daniel Craig, The Best James Bond EverTM
It would be interesting if the model predicted the gross of the last two films - QoS and Skyfall. I reckon that based on some of the factors including budget, QoS would have grossed higher than CR on that model. It would have been tough to predict an outlier (in terms of gross) like Skyfall
After adjusting the numbers, highest grossing 007 films:
1. SF
2. TB
3. GF
4. LALD
5. YOLT
6. TSWLM
7. CR
8. MR
9. DAF
10. QoS
So basically, the films from the golden period of Bond 60s-70s and the recent releases
All reasonable and largely accurate expectations. It ended up costing $300m instead of $250m but it was a fair extrapolation considering the previous 2 costed $200m and the Bond films tend to get expensive over time. Right now the IMDB rating rests at 7.2/10 (equivalent to Goldeneye). It started at 8.4 but most Bond films are stuck in the 6-range and films tend to drop down in rating over time as the hype wears down.
Not to demean Skyfall in anyway, but it's success was an anomaly (right place, right time), and it's not realistic to set the standard at those levels of sales. The next Bond movie will probably cost less than Spectre (because Spectre was really pushing it, and whereas Skyfall's profits gave a justification to go over-budget, Spectre's won't). It's very well possible for the next Bond film to top Spectre (and perhaps Skyfall, depending on what they do with it) in sales since numbers tend to go up over time.
Apparently the MGM number crunchers had it their way in the end.
I have said elsewhere that if Bond 25 does have a reduced budget as a consequence it could be a good thing. Smaller scale could mean tighter focus and less reliance on costly PTS. My favourite is FRWL and probably cost very little.
The over-the-top PTS in SF and SP are weak points in those films. I don't mind the sequences, but having the most exciting part of each film at the beginning makes the rest of the film feel weak in comparison. The pacing is thrown off by these huge PTSs.