Much like many sci-fi fanboys across the country, my eyeballs found their way in front of a District 9 screening last weekend and they enjoyed it immensely. One man, to which Rotten Tomatoes as of 09.13.09 could attest , chose to be the levee against the tide, panning the movie as “the sloppiest and dopiest pop cinema — the kind that comes from a second-rate film culture.” Now, the contents of the review are not without merit and whether or not you agree with Mr. White is a wholly different story. The real story comes in form of the town-hall-like internet mobs trying to get White shelved from reviewing any more movies on Rotten Tomatoes and whether their collective e-shouting is justified. It seems that while White was indignant at the sight of District 9, There Will Be Blood, and Wall-E, he rose up against the philistines to hail the cinematic triumphs that were Transfomers 2 and Death Race. I’ll be the first to admit that, for me, mindless action and gratuitous sex scenes are equivalent to pumping Mountain Dew Code Red straight into the veins of WoW addicts, but even I can make a clear distinction between the two classes of movies.
Some have come to White’s defense but at least one has quickly retreated from that position: Roger Ebert. Ebert stands by his decision to support White’s review of District 9 but conceded, upon further review, that “White is, as charged, a troll.” Then he says something strange. Ebert reiterates, by virtue of posting his original blog post, his support for the “ideal” that is Armond White. You see, Armond White has a TomatoMeter of exactly 50%. Though he tempers his endorsement of such a metric, Ebert nonetheless concludes that White to be the “epitome of the ideal critic, positioned smack dab in the middle of the scale.” David Engber of Slate follows the same argument saying “White votes with the mainstream exactly half the time: He’s neither conformist nor contrarian.” Engber even goes as far as to compile this gee-whiz graph
Without permission from Slate.com
showing that White indeed falls right at the center of agreement and disagreement with his peers. And while he astutely observes that the other critics “happen to be bunched up at around 75 percent on the scale” he cluelessly insinuates that it is the conformity of the general movie critic community that skews White to be depicted as a contrarian.
Despite what I, and hundreds of Rotten Tomato readers, believe to be overwhelming evidence that White is indeed a contrarian, there is an important philosophical and mathematical point to be made here: professional movie critics tend to agree with each other. As an example, let’s consider the sport everyone loves to hate: figure skating. Tonya Harding triumphantly returns to the oval and skates the night of her life. The five judges raise their score cards: 0.5, 0.4, 0.45, 0.5, 10. Now, you don’t have to be a statistician to suspect that something is awry. How could one person, from a panel of equally distinguished and impartial judges (i.e. rational), upon seeing the exact same performance, deviate in opinion so greatly from his peers? In the context of our example, most viewers would quickly divine that the 5th judge is in fact Shawn Eckhardt. But in general, assuming that judges, or in our case movie critics, are professional purveyors of their respective crafts, one can conclude that the 5th judge is irrational; that is, contrary to normal behavior. Let me try to make this point more formally.
Even without a rigorous statistical study of Engber’s dataset, it is clear from the graph that the judges are, as Engber observed, bunched together. Mathematically, this means that the dataset’s variance is small. In an experimental setting, Armond White’s dot represents an experiment gone wrong. The field of statistics says that, according to the data, the experiment may indeed deviate in such a manner but is mathematically unlikely. For real scientific experiments, the variance of our experimental measurement is due to randomness present in each test; perfectly controlled operating conditions are rare. In our case, the experiment is reviewing movies. More specifically, it is reviewing critics reviewing movies. Let’s bring a large group of professional critics together and have them watch a ton of movies and rate them. In the case of movie reviews, randomness surely exists (i.e. a bad day, a boost of caffeine) but for the most part, critics watch the movies under the same conditions. Intuitively, we understand that a rating, alternatively described as a statistic, aggregated from 50 critics is more telling than one aggregated from 2 critics; averaging over a large set of movie ratings, that is, statistics on statistics, this is even more true. Engber’s data say that when we have critics rate movies they tend to agree, which is why people use movie reviews to adjudicate the quality of a movie. What you see in the chart is not conformity, but rather consistency. In fact, the data is a testament to the integrity of movie criticism as a field. Mathematically speaking, dissenting from your peers for a few movies is understandable and not unlikely. This kind of behavior will not markedly change your place on the chart. Repeated dissension means something else altogether. Armond White is an outlier. And I’m not talking about the Malcolm-Gladwell-just-wrote-a-book-about-me-I’m-special kinda outlier. Rather, if we are to believe the other 19 critics to be impartial, well trained, adroit movie critics, the data set can be interpreted in two ways: 1. White is incompetent; 2. White acts in a deliberately irrational manner. Ruling out 1. by his credentials and Ebert’s quasi-support for him, one can safely conclude that Armond White is indeed a contrarian.
