Tuesday, November 29, 2011

Too many words about James Bond

The new Bond film 'Skyfall' is due to be released in November 2012. There are two reasons why this is the first Bond release that actually excites me,

1.  Ralph Fiennes is in the film and he has shaved his head. Now lets look at some facts. The last two films dealt with worldwide conspiracies committed by criminal organization QUANTUM (a clear ripoff of SPECTRE). The current synopsis for Skyfall indicates that MI6 will be under attack and that there will be some personal cost to Bond. Only three times in the 22 films has there been any personal cost to Bond - when Felix Leiter is seriously wounded (and his wife killed) in Licence to Kill, when Vesper Lynd is killed in Casino Royale, and when his wife Tracy is murdered by Blofed in On Her Majesties Secret Service. The Daniel Craig films have probed into Bond's past, and Blofeld is the leader of SPECTRE. Therefore this means that Ralph Fiennes may very likely be playing Ernst Stavro Blofeld.

You only live twice, Mr. Bond.

2. Skyfall is a return to the single word, slightly odd titled Bond films that are generally quite good (see: Goldfinger, Thunderball, Goldeneye. No. Seriously. See them. They will be the best thing you watch all week)

Now this second point got me thinking about the quality of Bond films vs the number of the words in the titles. Personally, my favourite film is You Only Live Twice, and my least favourite is Quantum of Solace. So from a min/max subjective perspective, the theory that quality drops as the number of the words in the title increases falls apart. But, surely, there is a way to figure this out in a nonsubjective way.

SPOILER ALERT: There is. And then I did it. Because I'm awesome. And also because I've got this useless math degree I've not been using. But mostly the awesome thing.

Before we get into results, let's talk about methods, because that's how you write a scientific article and this is legit science. Legience! Our domain includes only the Eon Productions films. So those two earlier Casino Royale films and that one Sean Connery film Never Say Never Again (which is just a remake of Thunderball and I will never watch) don't count. This leaves us with the 22 Bond films that are considered canon. For a complete list, click THIS LINK

I took the ratings for each Bond film from the approved critic score on Rotten Tomatoes. I used the approved critic's ratings because the general public has NO FRIGGIN IDEA WHAT A GOOD MOVIE IS. Basically, the critic's ratings seem far more accurate to me than the general public's. Maybe that counts as confirmation bias but maybe I'm choosing to count everyone with bad taste as an outlier because seriously, fuck those guys. I assure you that 'fuck those guys' is a valid statistical method.

I'll let you decide what this
bull is up to.

Other methods employed were just your basic calculations for means (averages) and standard deviations (how far away from the average a random film is likely to be). There's also a logarithmic trendline and some r-squared stuff but not very much and I'll explain it when we get to it. It's not terribly complicated but just sounds that way.

Now, lets get into looking at the data. Sexy, promiscuous data, wearing that dress just to drive you crazy.

Hey there.
Let's start with the basics. The average James Bond film rates at 70.8 %. Here are the top and bottom five, their corresponding scores, and the starring actor. Also included at no extra cost is some commentary from me (personal bias).

The Best
 Dr. No (98%, Sean Connery)
 Goldfinger (96%, Sean Connery)
 From Russia With Love (96% Sean Connery)
 Casino Royale (94% Daniel Craig)
 Thunderball (88% Sean Connery)

As you can see, Sean Connery dominates the top Bond films. As he should. Four out of the six films he starred in are in the top five, with only the recent Casino Royale beating him out. As I mentioned before, my personal favourite is You Only Live Twice -  a Bond film written by Roald Dahl simply has to be the best thing on Earth. However, Dr. No has my favourite scene in the entire series. It's the first clip from this video


It gives us a look into not only how ruthless Bond can be but also how that ruthlessness can make him a shitty spy (he probably should have gotten the information he wanted before shooting the dude, is what I'm getting at).

The Bottom (haha bottom)
 A View to a Kill (39% Roger Moore)
 Moonraker (47% Roger Moore)
 The World is Not Enough (51% Pierce Brosnan)
 The Man with the Golden Gun (53% Roger Moore)
 Tomorrow Never Dies (54% Pierce Brosnan)

The worst films are split between Moore and Brosnan (the sixth worst is Die Another Day, another Brosnan film). This isn't surprising - both of them are in some pretty terrible films. In fact, Brosnan's popularity as Bond is pretty much just coasting on the quality of Goldeneye. But, I'm rather fond of both of them, so it's disappointing to see them so poorly thought of. Also disappointing is the fact that Quantum of Solace is not on this list. Let's get something straight, there are good Bond films, there are bad Bond films, but all Bond films have something in them that's memorable. Except for Quantum of Solace. A View to a Kill is rightfully the lowest rated but it had Christopher Walken playing the shit out of being a Bond villian and that makes it entertaining and worth watching. Quantum of Solace had nothing. The only memorable moment in that film is when the Bond girl is coated in oil but that's just a ripoff from Goldfinger.


 This first graph is the average rating of films based off of how many words are in the title. The horizontal red line is the average rating for all James Bond films.

Sometimes I wonder why girls don't like me, and then
I do something like this and the world makes sense.

This definitely follows the trend that quality decreases as words increase, but there is an anomoly when it comes to two films. This is due to the fact that there are only two two-worded Bond films: Dr. No and Casino Royale.

This next graph uses a logarithmic trendline to show that, it's not an absurd conclusion to assume quality decreases with the number of words in the title. The blue diamonds each correspond to the top of the bars in the previous graph. The trendline is logarithmic because it had the lowest r^2 (or r-squared) coefficient. An r-squared coefficient is basically a summary of, on average, how far away your trendline is from your data. So if you have really high r-squared then your trendline is terrible. Just the worst. You should be ashamed of your trendline. The lower the r-squared is, the better the trendline fits the data. A logarithmic trendline had the lowest r-squared (as opposed to exponential or linear or other options), so that is what I used.

Ladies love it when you show
them your trendline.

Clearly, there is a decreasing trend. You can tell because the beginning of the trendline is higher up than the end of the trendline. Hypothesis proven. Chalk one up to science. Mission Successful. If I were James Bond, this would be the part of the film where M rings to congratulate me but I can't be bothered to talk to the guy who signs my paychecks and gets me million dollar cars (that I abruptly destroy) because I am too busy with sexytimes. But I'm not James Bond. So this is the part of the film where I go 'whelp, I've got all this data, what else can I do with it'.

I decided that what I could do with it was make a boxplot comparing the popularity of the various actors who have played Bond. While doing this, I came to the same conclusion that I reached three years ago. Boxplots are fucking stupid. So my graph doesn't used the traditional 'median, 1st and 3rd quartile' bullshit that typical boxplots do. No. Mine is better. The coloured boxes represent one standard deviation away from the mean. The mean is where the coloured boxes touch each other (as they show one s.d. above and one s.d. below the mean). The whiskers stretch out to the maximum and minimum values. In a couple of cases, the maximum and minimum fell within one standard deviation of the mean. I set these whiskers as equal to zero because Excel totally lost it's shit when I tried to input negative numbers. Never use Excel to graph things. Never. It is the worst.

It's very important that we talk about George Lazenby for a moment here. Lazenby only starred in one Bond film, On Her Majesty's Secret Service. The problem with this is that it totally fucks up my data. You can't calculate the standard deviation from one datapoint. Also, OHMSS is a really good film. So Lazenby has a disproportionately high rating. It's doubly awkward because the worst part about OHMSS is that it stars Lazenby. He isn't a very good Bond and he wears a puffy shirt. James Bond would never wear a puffy shirt. Bow ties are cool, puffy shirts are a sin.

We've now entered the part where I feel like talking about two fan theories about James Bond.

Theory 1:

James Bond 007, is a codename. One Bond dies and another moves up to replace him.

This is a really cool theory. Really super cool. You could probably make a couple interesting films based around this idea but there is one small problem: The other films don't reflect this fact. The actors change, but the character stays roughly the same. All James Bonds love girls, take their martini's shaken, and irritate the piss out of Desmond Llewlyn. 

I've had enough of
your shit, Bond.
If you're going to have six different spies take on the same designation, you've got to expect a greater variance in character then that. Sure there's some variation, Moore is the joker, Craig is angry and inexperience, and Dalton was a burnt out killer but that's not good enough. The character's are still far too similar to believe they are fundamentally different people. There's also one other major issue with this theory. Diamonds are Forever begins with Bond hunting down Blofeld for killing his wife in OHMSS. But Connery plays Bond in Diamonds are Forever and Lazenby plays Bond in OHMSS. So either they were in a polygamous relationship together with this girl or they are, in fact, the same person. This brings us to theory number two.

Theory 2:

James Bond is a Time Lord.

In Doctor Who, a Time Lord is an alien species that is capable of 'regeneration'. When they die, it is possible for them to regenerate their energy into a new body (given that they aren't killed instantly). This is how Doctor Who explains away the multitude of actors that have starred in the show. It's an incredibly clever device because it allows for a change in the personality and appearance (goofy, serious, arrogant) of the main character while being able to keep the characters memories and overarching philosophies (martinis, that time Blofeld killed Bond's wife).

However, like with the previous theory, OHMSS fucks everything up. George Lazenby was Sean Connery's replacement and OHMSS was the first Bond film that didn't star Connery. But Lazenby only signed on for one film and the reception for it was really bad. So Connery was brought back for the film that followed OHMSS. Using the Time Lord theory, this doesn't work. Sean Connery can't regenerate into George Lazenby and then regenerate back into Sean Connery before becoming Roger Moore. That's not allowed.

At least, that's what I thought. Then I realized, holy shit, Time Lords can travel through time, duh. So clearly, at the end of OHMSS, Lazenby travels back in time and warns Connery about Blofeld. Connery then hunts Blofeld down, preventing the events in OHMSS from ever occuring (and ensuring that his wife lives, even if they aren't married and will never meet now). Then, later on down the line in his incredibly dangerous job as a spy, he regenerates to Lazenby who does nothing important enough to make a film out of and then regenerates into Moore.

There are two more things that connect Bond to Doctor Who. The first is this picture right here
This man. I want to be this man.
Timothy Dalton has played both James Bond and Rassilon, the founder of the Time Lords. So that basically proves it. The other point is just a striking similarity.

Ernst Stavro Blofeld is Bonds archenemy. He is an evil genius and in his last appearance in the series (For Your Eyes Only), he was confined to a wheelchair and dropped down a chimney stack - presumably resulting in death but at the very least resulting in a huge amount of scarring and deformation.

Now, Stavro is pretty close to the name Davros. Especially when you consider how few words in English contain the sequence 'avro'.

Davros is The Doctor's archenemy. He is an evil scientist (genius) who created the Daleks. He is confined to a wheelchair. Here is a photo comparison of the two

Both very attractive men.

I don't think that it's too unreasonable to think that Blofeld and Davros are the same person. Thus lending the Time Lord theory more credibility. I am done now.