Saturday, June 11, 2011

Alternate vs Actual History Test: Which Really Happened (Alternative History #2)

Previous entry in the alternate history series: Homo Erectus Never Went Extinct

Next entry in the alternate history series: 54'40" Was Fought

No Googling now. If you don't know, guess which of these really happened, THEN click through. No mousing over to look where the links go either, smart guy. In chronological order of when they occurred or would have occurred:

- The U.S. military investigated LSD as a potential chemical warfare agent in the 1960s - but the first intentional use of biological or chemical warfare was in the seventh century B.C., when the Assyrians deliberately poisoned wells with ergot, a fungus which produces LSD-like compounds.

- A lost phalanx of Alexander the Great entered China and crushed the Qin outpost they encountered

- People living along the Mediterranean in ancient Greek times travelled to central Africa and had a hostile skirmish with pygmies controlling an oasis in the Sahara

- A purge of Buddhists from the Imperial Court in Japan

- The Moorish conquest of a Swiss village

- A Norse Kingdom in Sicily

- An Islamic khanate in Siberia which included Uralic-speaking subjects? (i.e. related to Finnish and Estonian)

- A Tibetan Buddhist republic in recent European Russia

- A Tokugawa naval expedition to Mexico

- Sacagawea's son moved to Europe as a young man with his German prince buddy, touring Europe and North Africa

- A United States of Central America

- Monterey, California "accidentally" taken over by American naval forces when it was still part of Mexico, several years before any war broke out, then given back

- A civil war between Mormons and the rest of the U.S., when Mormons considered everything out to California to be part of the state of Deseret

- A Scottish Colony in Panama

- U.S. and Germany fighting each other after being drawn into the Samoan Civil War on opposite sides

- A serious twentieth century effort to overthrow the U.S. government funded by major corporations, which resulted in criminal prosections

Did you click through? Here's a hint: only one of these was made up. The rest are true. The world is strange.

Previous alternate history post: What if Homo erectus still existed today?

Thursday, June 9, 2011

Film Quality vs. Profit: Is There Any Connection, and Does Talent Matter?

Bottom line: film studios are profit-making entities. Film quality as assessed by critics does not seem to closely correlate with profits. So do studios care about quality, and if so, why? To what extent do directors and actors affect critical evaluation and profit, and how is such an effect mediated? At times the film industry behaves inconsistently and seems to make decisions in terms of things other than profit. There are clear analogies to be made with the sports business, in terms of apparent paradoxes that can be resolved by being reminded that profit and winning games are not the same thing.

An article on Slate contains a Rotten Tomatoes-based gadget that tracks the performance, as measured by critical reception, of directors and actors over the last 25 years. (This choice of metric is important. The first question should be more obvious than it is: why should they use critical reception your yardstick? Would studios rather work with a Michael Bay, a director who reliably produces top-selling shiny shoot-em-ups, or a Terry Gilliam, who makes movies which are a tremendous joy to watch, and well-received critically - and which are tremendously expensive and run over budget and schedule? Now you see what I'm getting at. Most industries don't have the opportunity to lose focus on profit after getting confused by the artistic value of their products.) In the same way, a sports franchise with loyal fans can afford to have losing seasons, at least for a while, as long as the team can keep those fans filling seats, glued to TV sets to see commercials, and buying branded jerseys.

Reading through the Slate article about this new career-tracker gadget, you will note that John Ratzenberger (Cliff from Cheers) is the "winning" American actor. That is to say, he is the American actor who has made 10 or more films since 1986 whose films' average ratings were rated the highest by critics, at 76.1%. Compare to Chuck Norris, the worst actor, for whom the same statistic 18.4%.

"It's over, Mr. Anderson...I mean Prime." You'll get it in a second.

This might seem strange because Chuck Norris would seem to have more name recognition than John Ratzenberger, and (at a guess) I bet commands a higher quote. Also interesting is that in terms of total gross of films-appeared-in, the top actor in the United States is - wait for it - Frank Welker! You know, Frank Welker, the original voice of Megatron? In 2006 he passed Samuel L. Jackson with a career gross of US$4.9 billion. Of course in the U.S. we don't regard seiyuu as a separate career, as they do in Japan.

Why Would Studios, or Directors, or Actors Care About Ratings?

If you assume that critical ratings (i.e. quality) and profits are the same thing, then even the few statistics above present a real puzzle. Of course if commercial culture has taught is anything, it's that the just-stated assumption is a very false one, hence the existence of movies like Star Crash and Transmorphers (see point #3 here), which have little delusion about themselves as art but are safe bets as business propositions. At the very least there is likely to be a diminishing return on profits by improving critical perception of quality; a dollar you spend on a movie budget to raise it from just-okay to not bad might bring back more sales than a dollar that raises it from pretty good to critically stellar. Even if the critics care enough to spend a dollar more, the broader film-consuming public might not. As in sports, the film industry's product has a cultural value separate from its sales value, and because the cultural value is more salient to the public, film consumers confuse the two - just as sports consumers are puzzled about bowls and college ratings. But the film industry usually has its head screwed on straight and is focused on the real prize; they're (presumably) composed of materially self-interested agents and is not confused by this. Right?

Not necessarily. On its face the film industry would seem to be maximizing something besides profit, at least some of the time. Assuming Frank Welker's career-film-gross indicates a real contribution to films, and because as a voice actor you could probably get him to work for less than a big screen actor, then it would seem to be a no-brainer to keep using Welker in the Transformers franchise - as opposed to, say, Hugo Weaving, who will undoubtedly cost more and cut into the bottom line. But they still went with Weaving. (Now you get the little joke in the caption above.) What's the justification in cases like this? Did Weaving really want the part, and got his agent to call in a big favor to the studio to get him? Is the studio afraid of looking cheap by keeping the old-series voice actor, and signaling financial weakness to the rest of the industry? Did they actually project how many more tickets and rentals they would get from people who liked the Matrix, to prove that he would pay for himself? Or is it even less rational than that, and people at the studio and film crew just insist on having more prestigious people to associate with (like Weaving) and they're effectively willing to trade away profits to bask in his company? I have nothing against Weaving or his performance in Transformers, but decisions like these are curious from a financial standpoint, and they raise question about what's really being maximized.

Smart, i.e. rationally self-interested studios that win best picture would always auction these off, or melt them down for scrap. Imagine the rational, curmudgeonly studio exec. "Who cares if the academy liked it. I just want to make sure winning this thing doesn't hurt sales."

Going further, you wonder why a studio ever bothers at all with trying to get good critical reception. Yes, the Pixar movies that Ratzenberger is in have done well financially and tend to be highly rated by critics, but Chuck Norris's movies have been financially successful - but only financially successful. If you can sell tickets when Roger Ebert is bashing you, who cares? It's reasonable to think there's some negative impact on sales if the media hates you, but it would be interesting to see the actual relationship. How to measure? Movies are made for different amounts and intended to bring in different amounts; so perhaps compare on one hand each film's profits as a percentage of the film's budget, versus its Rotten Tomatoes average on the other. Either there will be some relationship between the two - or there will be none, or it will be too noisy to care about the correlation. If there's not a clear relationship between critical opinion and sales (or there's one that's grossly non-linear) it's worth asking what the value of the critics is, to the industry and to the public. To make sure we know what their film school wants us to like?

Frank Welker's take-earned-by-films-I've-been-in statistic raises another question. What's the average per film, and more crucially, to what degree was that Welker's influence? There's probably an 82 year-old key grip somewhere with a spreadsheet showing how his own takes are higher than Welker's. But even if you're looking at the average takes as opposed to absolute, what do you compare to? We don't know how much the movie would have made had X been in it instead of Y, and doing an average % take relative to budget wouldn't give us a comparison. That is, even if Welker has a good average %, how do we know that's higher than what the movies would have made otherwise? What counterpart would we use? (Even if we solved that, this is only correlation; the actor might just pick good-selling movies, as opposed to making them good-selling.) If there are measurable effects, do actors or directors on average have more impact on quality and/or take? Analogously, analyze NBA teams, and you'll find that on average their records from year to year are more closely related to who's coaching than who's playing; when I did this, I didn't investigate whether this is from recruiting skill or on-the-court coaching.

Finally, and I have no proposal for how to measure this, even if there are measurable effects from a certain actor appearing in the film, what mediates that? Is the public going because they think they'll get a good performance, or do they just like the actor because they're familiar with him or her? The fact that studios are willing to pay a premium for well-known actors instead of just using unknowns that can act just as well as the people who had a break (which comprise a large portion of the LA population) suggests that the studios believe familiarity is at least part of the effect.

Of course we might assume big studios investing tens or hundreds of millions in projects aren't stupid; they're businesses looking for an ROI, and they must already doing something like these analyses. Then again that assumes that their decision-making process is profit-maximizing, when the choice of actors as discussed above strongly suggests otherwise (status signaling? ego-stroking by association with celebrities vs. unknown actors? quality, among LA's artistic idealists?)

Depressed by how all description of how much the creation of art is dictated by eonomic considerations? Then move to a much less capital-intensive endeavor with smaller teams, like writing. One person risking only their solo time at a keyboard can and does usually produce more innovation.

In conclusion: I'm not the curmudgeon about the value of film quality that you might assume from this post. In fact I'm a huge Darren Aronofsky fan and I'm very much looking forward to his next film Human Nature, which will star George Clooney. But with The Fountain (easily my favorite film of the last decade) Aronofsky came perilously close to Gilliam territory in terms of his production stopping and starting again. I'm glad that he's able to keep making high-quality films but I recognize that he's no doubt compromising what would have been an even better film, all the time, for business purposes. But the mystery remains about why studios care to invest in films like his at all. Whatever un-focused fuzzy calculations distract them from profit for long enough to fund projects like this, I'm glad.

Find the Rotten Tomatoes career-tracker here.

Sunday, June 5, 2011

Economic Freedom and Happiness

I was recently looking at the economic freedom numbers for countries around the world and I wanted to know what the connection to actual outcomes was; in particular, the happiness of the people in those countries. All the rest are surrogates. Political and economic debates sometimes lose focus on this fact.

I looked at the following data, for all countries that had them (evident in the datasets); for each, it was always a clear majority of countries on the planet.

- The Heritage Foundation's Economic Freedom Index for 2011

- White's Life Satisfaction Index

- GDP per capita (IMF data, supplemented by CIA for small countries, non-reporting countries, or inaccurately reporting countries)

- Gini index (from the U.N., or for some countries in the midst of conflict the Global Peace Index statistic)

- Economic growth for countries 1990-2007 (data from United Nations Statistics Division)

As I've done previously, rather than show a bunch of scatter-plots, I'll give you the statistical highlights. I'm happy to share the spreadsheet if anyone is interested, although this is not quite graduate-study-level QC'd data.

Map of White's life satisfaction index in 2006. This is what counts.


1. Interestingly, economic freedom correlates more closely with life satisfaction than with economic growth (R^2=0.239 vs 0.1075). This suggests that economic freedom adds to utility other than through direct material gain. Freedom does correlate better with per capita GDP than with growth (more below).

2. Economic freedom is associated (albeit weakly) with a decrease in Gini, that is with a more equitable income distribution.

3. There were a number of outliers in the plot of economic growth vs. life satisfaction. Most of these were very high Gini countries.

4. In the "unsurprising" cateogry: there were two R^2 that rose above 0.3 were the correlation between per capita GDP and life satisfaction (0.3106, stronger without outliers). The correlation between economic freedom and per capita GDP was even stronger at 0.3873. The curve appears to flatten at the high end of per capita GDP (removing three outliers raised the R^2 to 0.4665) reinforcing the conclusion that once basic needs are met there is a diminishing return. It bears emphasizing that this is a correlation, not a cause; economic growth may CAUSE economic freedom, or they may both be caused in parallel by the same thing.

CorrelationR^2Relationship Means?
Econ free & PCI0.3873Every point increase in econ freedom (range 0-100), raise PCI US$70
PCI & life sat.0.3106Raise PCI US$1,000, increase life sat. 1.14 pts (range 100-250)
Econ free & life sat.0.239Add a point in econ freedom (range 1-100), increase happiness 1.67 points (range 100-250)
Econ free & growth0.1075Add a point in econ freedom (range 1-100), increase growth 0.001%
Econ free & Gini0.0837Add a point in econ freedom (range 1-100), decrease Gini 0.2685
Growth & life sat.0.0544Every 10% increase in growth, get 25 points happier (range 100-250)
Gini & life sat.0.0258Increase Gini by 1, life sat. drops by 0.5 (range 100-250)

Unsurprising, but interesting to see in this form. Adopt policies that expand economic freedom in order to make people happier, partly by increasing growth. The best way to produce happiness is to reach a target high per capita income, but there is a diminishing return. Economic freedom has a weak beneficial effect on Gini, but Gini can offset the happiness effects of high PCI and good growth.

Future questions:
- There are a number of countries that, looking merely at per capita GDP, aren't as life-satisfied as they should be. These countries usually have large Gini; a surface in a 3D scatterplot would show this distortion. It might be informative to see which countries are "off the surface" in terms of how much we expect their Gini to distort their happiness:PCI ratio, and then ask how this effect is transmitted - a first guess to investigate would be degree of media saturation. Prediction: countries with more media and a high Gini will tend to be less happy that those with less media but the same Gini. Seeing how the other half lives forces everyone into the same status game.

- Are certain regions of the world off these curves in predictable ways because of cultural commitments? (See the World Values Survey.) E.g., are Confucianist countries less happy per dollar of PCI? Or are cultures with more family-oriented, traditional values differentially susceptible to the effects of Gini distortion?

- (Added later: personal economic freedom in the 50 U.S. States can be found by category here.)