The Poppies of Terra #71 - The Accelerating Expansion of the Mehniverse

In the society of Plato’s Republic only a small elite of “gold” souls rises above the masses of “bronze” souls to achieve true virtue and wisdom, implying that most human endeavors fall short of excellence. Aristotle argues in his Nicomachean Ethics that “excellence is an art won by training and habituation.” In other words, it’s rare. Centuries later, Benjamin Disraeli went on to write that “nine-tenths of existing books are nonsense, and the clever books are the refutation of that nonsense” (1870) and Rudyard Kipling similarly opined that “four-fifths of everybody’s work must be bad” (1890). Theodore Sturgeon famously encapsulated this notion of reigning mediocrity and scant excellence with his “law” stating that “ninety-percent of everything is crud” (1957). Thus, we might say, it was ever so.

Now, if we assume that the proportion of memorable pop culture to disposable entertainment tends to hold steady, the rising volume of movies and shows would simply mean, in absolute numbers, that there are more great experiences for us to discover with each passing year. And I do believe that’s true. But I think a key part of this equation is the “discovery” component. While there may be more higher-quality entertainment being crafted, per unit of time, than ever before in history, it feels like each decade we hear about less and less of it.

Here’s a very quick temperature check. If we look at the highest IMDB-rated film per year between 1990 and 2019 that has a substantial number of votes, and then take the average per decade, we see that in 1990-1999 the top-rated movies averaged a score of 8.66, from 2000-2009 they averaged 8.59, and from 2010-2019 they averaged 8.45. There’s a dozen caveats to this (for instance, rating inflation or vote thresholds), but it does directionally align with the idea that the perception of the quality of the highest-rated films by IMDB users has declined over the last few decades, and done so in an accelerating fashion, dropping by 7% from the 1990s to the 2000s and then declining by double that, 14%, from the 2000s to the 2010s.

So if Sturgeon’s law of proportionality is holding true, and each consecutive decade actually had, on the whole, more films that ought to be rated more highly than those of the preceding decade, it would seem that not as many of these top-tier films are breaking through in a mass way to the users who vote on the site.

This brings us back to the idea that maybe it is objectively becoming harder to find the rising number of top-notch stories out there.

But how much harder? Is there a way to even approximately assess that?

In what follows, I’m going to focus on films, and I’m going to use some simple samples to illustrate how we might build a model to try and quantify the growing difficulty in finding great movies. This is a conceptually simple approximation, inspired by increasing distance of objects in our actual universe due to its accelerating expansion.

First, let’s think about production dollars for theatrically released movies, and let’s compare not-great, huge-budget flicks with highly regarded, lower-budget films.

The following two tables show recent-ish representative samples of each:

Low-Rated Blockbusters (RT < 60%)

High-Rated Indies (RT > 90%)

The reason I want to start here is because I’m interested in how many more eyeballs, on average, the marketing that accompanies the whales can get exposure to as compared to the marketing of the minnows. To estimate this, we’re going to look at a standard marketing measure called Cost Per Mille, or “cost per thousand,” which represents the price for 1,000 ad impressions or views, calculated by dividing total ad or marketing spend by total impressions and multiplying by 1,000.

Using the above samples, we can complete some quick additional math to estimate CPMs for each of these cohorts, using typical industry figures as yardsticks:

• Blockbusters: Production $2.067B total ($206.7M/film); Marketing $1.345B total ($134.5M/film); Avg. CPM $18 (blended: 60% TV/outdoor at $20–50, 40% digital at $5–15; from Gupta Media/Statista 2023–2025 benchmarks).

• Indies: Production $58.3M total ($5.83M/film); Marketing $118M total ($11.8M/film); Avg. CPM $8.50 (blended: 70% digital/social at $5–15, 30% festivals/earned at ~$2–5; from Match2One/Gupta Media 2023–2025, influencer benchmarks ~$4.6).

We can summarize what we have so far as follows:

We can see from the above that even when blockbusters are critically weak and indies are critically beloved, blockbusters still force roughly 5.4× more paid eyeballs through sheer budget size, overcoming comparatively inefficient media buys ($18 CPM) through scale. Said differently, blockbuster brute-force volume (11× higher spend) offsets their 2.1× less efficient CPM, netting a ~5+× exposure edge. Indies punch above their weight via organic amplification, but paid reach lags due to their budgets.

For our second consideration, let’s think about production dollars by cohort. If we assume a finite amount of production dollars per year as our potential spend pie, what slice of that pie is allocated to the tentpole monsters? In the 1990s, a rough estimate is that about 85% of the production share went to blockbusters. That ratio has been creeping up decade by decade.

This brings us to the third and final factor I’d like to bake in before we put everything together. Let’s return to that nagging issue of how many more choices are available when looking for a new film. 

There’s many different criteria we could choose to try and estimate this, but I’m going to take my cues from this recent article citing P. T. Anderson which happens to contain helpful estimates (and, incidentally, some solid film recommendations!).

Using that as a starting point, and digging into the stats a little more, I’ve compiled the following high-level summary as relates to the volume of productions in absolute numbers by decade:

We now have three simple factors we can combine to create an estimated difficulty multiplier.

The following table summarizes the three factors and how they’ve changed by decade:

We’re ready now to perform the final bit of calculations:

In this last table, I included our current decade. We’re only halfway through and some of the stats are a bit wonky because of the pandemic, but I was curious if the general trend was holding.

And it is.

As you can see, if we compare the 1990s with the 2010s-forward, it’s become somewhere between 18x-20x harder to find higher-quality, non-blockbusters than it used to be through non-targeted exposure.

There’s yelling at clouds, and there’s meteorological forecasting. 

Keep in mind, my very simple model looks at only films. If we were to include television, video games, user-generated content on social media, etc. etc, and thought about our time as the finite resource available for exposure rather than assuming that as a constant background factor, we’d probably find our difficulty multiplier would be even higher, since many of these other activities have grown in CPM competitiveness. Also, I took three basic ideas and treated them as independent factors. There are probably complex and subtly non-linear interdependencies among those three scaling parameters. Finally, an indie may be bad, and a blockbuster might be terrific. 

Then again, this is all based on the idea of essentially random exposure, something akin to the notion of “what does the world of movie culture choose to make part of its central conversation.” But in reality, highly sensitive algorithms are largely responsible for driving customized engagement based on historical usage patterns. So folks who take the time to engage heavily with indie-forward news outlets, review channels, podcast discussions and so on, may in fact receive disproportionately more exposure regarding titles outside of the “general” currents than they would have in the past.

But that brings up the question of time and energy, which goes back to the central idea. If we’re willing to deliberately invest more effort to discover interesting stories, our efforts are likely to be rewarded, and incrementally more so today than ever before.

Our findings aren’t necessarily surprising. The expanding universe analogy comes in handy. As space-time expands, the distance between everything grows. Likewise, as more entertainment options become available, the distance we need to cross to find the ones that appeal to us increases. We can remedy some of that through algorithms, pre-curated lists, and so on, but the overall difficulty continues to grow over time. 

Keep in mind, most standard cosmological models today posit that our universe is roughly 68% dark energy, 27% dark matter, and only the small remainder ordinary (baryonic) matter. In other words, 95% dark energy and dark matter, and 5% the kind of matter and energy that gave rise to us and which we can actually observe.

Turns out, Sturgeon may have been generous with his 90% estimate after all.