As a follow up to the wildly popular “Hubbert Peak Theory of Rock” article on the decline of rock music (and U.S. oil production), I decided to take the logical next step in pop culture quantitative analysis: Greatest Movies Lists. Only this time, no spurious correlations–just straight up number crunching based on these lists.
UPDATE: I can haz arithmetix? There were some errors in the math in the original draft of this post. Details below.
Did movie quality peak in the ‘60s and ‘70s? Has it all been downhill since The Godfather in 1972? Some statistics you can’t refuse, after the jump.
One of the main criticisms leveled at the “Peak Rock” theory was the very nature of the Rolling Stone 500 list. Clearly, the RS 500 skews far towards older songs. It has only a small handful of hip-hop, a genre that came to maturity much later than pop/rock. In short, too many great songs from the 80’s and 90’s were overlooked by the hopelessly retro-focused folks surveyed by RS.
But what about movie critics? Are they, too, hopelessly stuck in some bygone “golden age” of cinema and disregard everything since then as inferior and derivative?
Let’s take a look at the American Film Institute’s “AFI 100 Years, 100 Films” list of the 100 greatest movies of all time, as determined by a 1998 survey of industry professionals:
Notice that the single year with the greatest number of movies in the AFI 100 list is 1939 and that the trend line shows a bell curve that peaks roughly in the middle but tails off towards the end. The median year of the distribution is 1959.5, which is 39 years prior to the release of this list in 1998.
Oh noes! Is his proof that we’re running out of good movies? Not so fast. Fortunately for us, AFI updated its list 9 years later, in 2007, and here are the new results:
Nothing shockingly different, right? But notice that after 9 years, the median year (1964.5) for top movies has increased by, you guessed it, 9 years. Clearly, the outcome of the new rankings reflect an accommodation of modern tastes. New classics were crowned (Titanic, Lord of the Rings), while old ones fell off the list (The Jazz Singer, Frankenstein…wow).
The fact that the distribution’s median kept pace with the 9 years between the lists is an indicator that, in spite of all the insipid comedies and formulaic blockbusters that seem to have dominated the modern cinema since Jaws and Star Wars, studios still manage to produce a few truly great films at a respectable rate. Eight or nine years from now, hopefully AFI will do another update to their list, and I wouldn’t be surprised to see that median year move up eight or nine years as well.
So that’s what the critics and movie professionals have to say. But when you ask the movie-going, internet-surfing public what they think, you get a very different picture. Let’s take a look at that most democratic of movie ranking systems, the IMDB Top 250, as determined by users ranking individual films on a 1-10 star basis:
Note: data reflects the IMDB as of Sept 30, 2008.
That’s no bell curve. The trend definitely shoots upward starting around 1990 and continues through the current decade. How could this be? Isn’t this the era of Not Another Teen Movie and Pearl Harbor? Isn’t this the time when Jerry Bruckheimer and Michael Bay unleashed their celluloid crap onto the masses? Why, yes it is, but it’s also the time of the fanboy, that over-exuberant, internet-savvy, slavishly devoted fan who comes home from the midnight screening and *ZAP!* fires away a 10 star salute to the work of incredible genius that he just saw. Why else would The Dark Night be the Number Four Greatest Movie of All Time according to this? For a few weeks after its initial release, it was even ranked Number One, over longtime favorites The Shawshank Redemption and The Godfather. Some other recent entries on the list that are of questionable “all-time-greatness” include The Bourne Ultimatum, The Prestige, Sin City, V for Vendetta, Shaun of the Dead, and Pirates of the Caribbean. Don’t get me wrong. These are all fine films, but they’re also the types that get fanboy-inflated ratings.
All of this brings us to two fundamental question. First, what’s the difference between movies and rock/pop music that seems to allow for continued output of “great” movies, while rock/pop music produces fewer and fewer “greats” in the post-golden-years?
To be fair, people do complain about the decline of both the movie and music industries. Both are derided for churning out predictable, derivative fare. However, I think there is a widely held belief that Hollywood can still make a great film of staggering genius, but the music industry will never produce anything approaching the originality and greatness of the musical greats from the ’60s and ’70s. To explain this perceived difference, I would point to some fundamental differences between movies and music. First, although the barriers to both movie and music making have been dramatically lowered, it’s still relatively harder to churn out a truly professional movie versus a well-polished album and get it out there. Distribution of feature films is still limited by the number of big screens out there, which can’t multiply as fast as the number of iPods and iTunes downloads. Consequently, the music market is more saturated than the feature film market. With less competition to choose from, it’s easier to pick out and identify “the greats” from the dozens of major feature releases each year versus the thousands upon thousands of albums that make their way onto the market.
Second, on the artistic side, I still hold to the theory that I proposed in “Peak Rock” that rock/pop is an inherently limited form. There’s only so much you can do with the guitar/bass/drums rhythm section and the confines of Western popular music. The motion picture, on the other hand, is virtually limitless. The two hour feature film gives incredible leeway in storytelling, both in the crafting of the story itself as well as in the visual and aural representation of that story. Movies have visual effects, dialogue, physical action, silence, sound effects, musical scores, and countless other variables. Rock/pop music has just the music, an inherently limited form at that.
The second, and more existential, fundamental question: is all of this list-making a pointless exercise? With something so subjective as quality of cultural creations, who really has the authority to make lists of greatest songs or movies of all time? Rolling Stone? AFI? The Public? None of the above?
My first instinct is, yes, it’s pretty pointless. I don’t think there’s a whole lot to be gained by arguing why Frankenstein deserves a spot on the AFI Top 100 list versus Titanic, or why Green Day got left off the Rolling Stone 500 list while Dave Matthews Band made the cut. But everyone has the right to their own personal favorites. I keep a running list of my top five or so movies, concerts, albums, etc. And at their core, these lists are just a bunch of peoples’ personal favorite lists, just aggregated and compiled by an entity that thinks it has enough credibility to speak on the Greatness of Culture authoritatively. The IMDB list is a little different in that it uses compiled rankings on a movie by movie basis, and that it’s open to the general public as opposed to industry insiders, but it still is a distillation of a large number of individuals’ subjective pop culture tastes.
In conclusion, I don’t think we’re in some sort of crisis of quality pop culture production. Rock/pop music may continue to be deritative of previous works, but that won’t take away from my enjoyment of the next Killers/Ben Folds/Fountains of Wayne album. And they still give out Oscars every year, right? As for these lists, if you don’t like what you see on them, make your own pop culture. Or support those who are putting their creativity on the line and making their own pop culture. Or at least post some comments here and overthink this with me.
UPDATE 2008-10-27: Turns out that anyone with simple math skills would have noticed that I incorrectly calculated the difference in median years between the AFI 1998 and 2007 lists. 1964.5 – 1959.5 = 5, not 9, and that does do damage to my argument that movie tastes had kept up with the advancing years. In fact, it seems like they hadn’t, at least by AFI’s standards. Nonetheless, I would argue that the median advancing by anything at all is a good sign that the march of progress is continuing, at least a little bit, in the movie world. Also, I doubt subsequent Rolling Stone 500 song list would shift that much in almost a decade’s time due to the undue influence of the Motown hit machine on the list. Motown was truly a unique phenomenon, and I can’t think of any analogous movie hit machine that would skew critics so heavily towards one time period when making these kinds of lists.
I found this data to be very interesting, but not unexpected. I think we have to remember that IMDB’s top 250 list is not a reflection of the quality of the films, but rather of their popularity, and will thus always be skewed towards recent films.
Amen, Jorbex. The IMDB list is comprised of votes by Joe Sixpacks AND film connoisseurs, not just the latter (which is the AFI list). As such, it’s probably safe to assume that a lot of the movies AFI would list don’t make the top 50, let alone 100, on the IMDB list because voters haven’t seen them.
But Lee, I have to interject a little. How does the saying go, “There are only nine plots in existence,”? Something like that. The point is, there are only a certain number of ways every story can go, and all have been used and exhausted in novels and films time and again. So this doesn’t really make the film industry any different than the music one, in that sense. What makes either a quality film OR song is *how* it does what it is trying to do, not WHAT it is trying to do per say. To allude to the other article a bit, neither has to necessarily be “innovative” in order to be “good” or “golden,” either. If it’s done WELL, then hey, why not?
Before people who actually know statistics start pouncing on me, I should probably point out that the “bell curves” I produced in the above graphs are actually “4th order polynomal trend lines” that I made in Excel. Bonus points for anyone who can explain the difference.
Gab, Jorbex, I agree with your interpretation of the IMDB lists’ bump in the more recent years, but I do stand by the “fanboy” phenomenon whereby someone gives a 10 star rating to a new movie without the thought that such a judgment deserves. Seriously…10 stars? Those should only be reserved for once-in-a-lifetime, truly iconic movies. Based on this scale, I’d give a well though out 8 for “The Dark Knight,” given that “The Godfather” gets a 9 on the IMDB scale, and that TDK, though freaking awesome, has its weak parts and is probably not a better movie than “The Godfather.” Others, though, are not so discerning. They just come home, still blown away, ZOMG JOKER ZOMG ZOMG, and hit the perfect 10.
Give us your current top 5 concerts Lee.
i don’t know… yeah, the data are sparse in the ’20s and ’00s… it looks pretty flat but noisy to me rather than a normal distribution — although i guess a flat distribution would be a normal one. how does it look if you combine the two AFI lists (i.e. the ’07 list plus the 23 that they took out of the ’98). what do the data look like when binned?
what if one were to express the frequency in relation to the number of films (and where do we draw the line there?) produced in the given year? the ’80s would look like even more of a cesspool of film…
i’m lazy and i won’t attempt to answer any of these questions. sorry. for my money, the late ’60s through the ’70s produced the best crop of movies.
“Those should only be reserved for once-in-a-lifetime, truly iconic movies.”
And this is YOUR rating system. If 10 somehow becomes “off-limits” (since obviously no one could ever think The Dark Knight (as an example) is a better movie than the Godfather) then 9 becomes the new 10. What is someone supposed to give a movie that they really really like, even though it may not be as popular with other people? Roger Ebert is said to have been the critic who has given out the most 4 star reviews. Would you call him a crappy critic?
I will not deny there are fanboys who rate movies in terms of 10s and 0s, but I also cannot deny that certain movies resonate with certain people. If the Dark Knight is on the top 250 on fanboyism alone, then I think a lot of other movies would be on the 250 as well, the least of which being Transformers.
I would also say that the imdb top 250 is skewed towards more recent films because the more recent films have a higher number of votes (for example The Dark Knight has 291,000 votes whereas the classic Chinatown only has 59,000). Given the formula that imdb uses to rank come up with a movie’s “weighted average”, it gives more weight to a movie has a higher number of votes.
Another major factor is imdb’s skewedness is the demographic of its users. It is used extremely disproportionately by males between the ages of 18-29. For example, looking again at The Dark Knight, of the 291,000 votes, only 38,000 of them were women (thats only 13%!). And 214,000 were between the ages of 18-29 (thats 74%). Gender and age are big factors in what people think about a movie. (Of course a 25 year old guy is going to like Fight Club, Sin City, The Professional, Goodfellas, The Usual Suspects, etc. more than a 75 year old woman)
I have not checked it out yet. But there is also a list of worst movies on IMDB. Maybe i am wrong but i am pretty sure that there more recent movies on it then old movies.
I think the fanboy theory is true but if the Internet would have existed in the 70s like it does today there would be no great curves.
At least just considering the IMDB list.
I am looking forward to see your article on that in 30 years.
You are missing a very, very fundamental difference between rock and film, and as a consequence, are comparing apples to pears (is that an idiom in English? I am German, and it’s a German idiom but I have a feeling it works in English too).
The thing is, rock music is a music genre, much like action movies would be a genre of film.
You’d really have to find an equivalent film genre (one that first popped up around the same time? Like… I don’t know what that could be, if anything) and compare that to rock music, or compare film to music in general – including symphonic, opera, Jazz, Pop, Musical, Folk…
Or maybe compare only standard Hollywood Studio fare to mainstream pop & rock? That would leave out everything independent on both sides… well I guess film and music are kind of hard to compare, really.
I’ve been pretty actvie on IMDb for over a decade and I will attest to the fanboy phenomenon. Kevin Smith and Tarantino are gods there. The old-time snobs (i.e. me) have our representation there too. The people self-segregate. This becomes obvious when you compare the message board for the Dark Knight vs. the board for Chinatown.
So I think your fanboy theory is pretty water-tight. However, what about its diametric? What about the fanboy equivalent apparent in the AFI list? Does Spartacus hold its weight when viewed by today’s audience? I didn’t like it. Is the Wild Bunch really that good? I don’t think so. Can Intolerance be tolerated (sorry) by anyone? I love the silent era, but this movie was unbearable. Is Hitchcock really that great? Rear Window and Vertigo are popular with *Hitchcock* fanboys, but I can’t stand them. So I guess I’m saying that there is bias toward the old just like there is toward the new.
These movie lists are only shortcuts to finding greatest in the popular culture. If fanboys are the predominant users of IMDB, then wouldn’t to make sense that users of IMDB would want to know what their subculture ranks as a 10-star film.
It makes me recall what Matt Damon recently said. To paraphrase, a good movie is only good if it is still good ten years from now. Therefore, it would make sense that movies are only deemed classics long after they are distributed to the world.
I do wonder why movie fans stay focused on New Releases and only movies in the recent past. Perhaps it is not strictly our tastes that are represented in IMDB rankings, but something from the political adverting book of “astroturfing.” If you don’t know what astroturfing is, look it up in Wikipedia. I’ll try to summarize, basically it is advertising hidden as a grassroots led effort.
One final note, I have come to the conclusion that good movies like all good art including music needs to be a conscientious pursuit because not all 10-star films can successfully marketed (find their audience)in our fast-paced society.
I wonder if that bell-curve might not actually represent the median “impressionable” age of the critics involved in the survey and the 9 year drift is tracking that accordingly as the older ones die off and the new ones come on the scene. Each film critic falls in love with cinema at a certain age and probably considers that time to be a sort of golden age?
On the Hitchcock issue – he invented a lot of the cinematic technique we don’t even notice now it’s been so widely adopted. He’s important if you care about film on any level deeper than escapism.
Markus: I would counter your argument by saying that “rock,” at least how I think of it (and how it’s represented in the Rolling Stone 500 list), is more than just a genre. I consider it more of a broad art form that constitutes different genres: metal, punk, pop, rock, r&b, maybe country; pretty much everything that’s not hip hop, which doesn’t follow the rhythm section/western music rubrik. Jazz, different art form all together. Symphonic/Classical, different art form.
Likewise, the motion picture isn’t a genre in itself. TV, feature films, video art–these aren’t genres, they’re different art forms.
So my analogy is something like this:
Action Movie: Feature Films :: Heavy Metal : Pop/Rock
Romantic Comedy: Feature Films :: Motown : Pop/Rock
Um… 1964.5 minus 1959.5 equals 5 years, not 9. Am I missing something from the analysis? This means the median moved by 5 years between the 1998 and 2007 surveys of film critics. This would indicate a drop off, of about 50%, in “great movies made by time.”
Argh, you’re right Chip. Thanks for pointing that out. Excel makes me bad at plain ol’ math.
I think we’ll need to wait and see if this holds true in the future. Perhaps the newer movies have not yet stood the test of time and are therefore not on the list. Or if truly great movies come out in the future, they will bump the mediocre ones off the list, maybe from the 50’s 60’s and 70’s, thus creating a new bell curve with the current time at the peak.
Isn’t a popularity a required feild for a fine film? Even if it’s the most breath-taking, beautifully filmed masterpeice, somebody still has to see it for it to be considered a fine film. Or am I crazy? The way I see the IMDb list is not as a popularity rating or a critics rating, but more of as the AFI top ten of the people. I mean I’m not doubting that most of the suff on that list ian’t some of the greatest work in the past hundered years, its just that those lists are coming from professional film critics. The IMDb list is really no different from that it’s just that the people criticing it aren’t highly paid magazine editors, but as stated above fan-boys and regular 21st century pop-culture buffs. I know it is based on a rating scale, but why would you rate it highly if you didn’t see some form of quality in it? So really there is no real difference between the rating list and the critic list. And frankly, I’d trust the IMDb over the AFI just looking for a movie because it seems that the IMDb would be geared more to a general intrest instead of the highly sophisticated film attache. But that’s just my opinion.
Standard deviation.
Look it up.
yeah, i agree with the Doctor on this one, in terms of finding a film to watch i would rather use the IMDB’s top 250, just in terms of watchability alone. I trust more of what a regular person recommends instead of what a bloated pretentious film snob thinks is great because what they recommend is usually some artsy snorefest that a regular person cannot relate to. And as for the Godfather, it is a good movie, but its one of those movies that people say is the greatest just because everyone else says its the greatest even if they don’t think it is. If anyone doesnt think the Godfather is the greatest then they get about a hundred people bitching at them, as I’m sure i will for even mentioning this.
how can you say films like sin city, the pretige and others dont deserve to be classed as some of the greatest films ever simply because you don’t agree, if the majority of the public believe these films to be deserving of such status, then that surely means they are that good. Basically if the majority of the public believe a film to be iconic then it is even if just for the reason that so many people like it.
Are there lists like this for video games and novels? I think it would be interesting to see if the distribution holds true for those formats as well.
Being a metal drummer, math teacher and a bit of a film fan I feel I should add my two cents to this discussion.
On the trend lines: you can fit any line to those bar diagrams: a linear curve, a 2nd (or higher) order polynomial curve, and yes, a bell curve too. A bell curve is not a polynomial curve. A bell curve will drop off to zero far from the mean value, whereas a polynomial curve will shoot towards infinity at a certain point. So choosing a bell-curve means you assume a drop-off at a certain point. A high order (6th is the highest in Excel) polynome will probably give you the best fit, but might not make sense if you want to extrapolate towards the future.
All of this doesn’t really matter, because the point is of course how WELL does a trend line fit? Excel can give you the value of ‘R-squared’. 0 being no fit, 1 a perfect fit – so the closer to one, the better it fits. Try some different types of curves in Excel and see which has the best fit.
On pop/rock then: I’ve been a rock/metal fan since ’79 and I _have_ noticed an “impressionable age” correlation. My idea is that the music someone loves when they’re about 16 will stay with them forever. I’ve ‘tested’ this with several of my friends and some other people at birthday parties, and it seems to stay true. However, I’ve also noted that my taste broadens: in years and in genre. So now I like anything from 1967 Jimi Hendrix’s pop/rock to 2008 Krisiun’s brutal death metal. With my band we try to mix what I liked when I was 16 with what I like best of the current albums.
On films then finally (people still with me? :-)
It was definitely with Empire Strikes Back in 1980 (when I was 9) that I fell in love with Star Wars – and I still like it. I still see a lot of films, but seldom will get as excited as I was back then. This supports the “impressionable age” theory, though I’m not sure what age that would be.
So summarising: with rock/pop and films I think it’s all very subjective and dependable on the person’s age when they fell in love with a certain album or film. You see that with the generally older AFI people who like the older films, and with the fanboys who are having their peak now. And you can see the effect of mixing those two in the IMDB top 250. The AFI bell curve with an added fanboy flaring at the end.
There is simply no way putting an absolute quality value on a film or an album – it’s just too subjective. Every age will have it’s favourites.
And math won’t help you with that. :-D
Thanks for all the comments everyone, including those calling me out on my lack of statistical (or even basic math) skills. ;-)
And I think Max summed up this whole discussion quite nicely:
“There is simply no way putting an absolute quality value on a film or an album – it’s just too subjective. Every age will have its favourites.”
If you want to equate popularity with greatness, then I guess High School Musical 3 is the greatest story every put on film…at least this week.
Perhaps the best way to restrain such fervid fanboyism is to make the IMDB poll run like the US Postal Service: No one can cast a vote until 20 years after the film’s release. If the Postal Service had no such constraint in place, philatelists would have to deal with stamps featuring the Captain and Tennille, Debbie Gibson, Paula Abdul and MC Hammer.
How about the impact of generational (re-)release? Some of us were teens in the 1970s and saw “Star Wars” in a cinema, but had to re-rent on DVD so our kids could catch up on the plot AND take them to see the new episodes. I imagine there’d be a similar impact from converting TV shows to movies (e.g. “Starsky & Hutch” and “Bewitched”) as well as remakes (e.g. “King Kong”, and “Journey to the Centre of the Earth”). How about Director’s Cuts?
if afi is the measure, then i don’t want anything to do with this