I've long felt that the individual film critic's job isn't really to give you a thumbs-up or thumbs-down on whether a movie is good or not. You'd be better off looking up on Rotten Tomatoes an aggregation of critics' ratings to even out the random perturbations.
Now, there are a tiny number of unjustly-overlooked movies that I've helped call to public attention -- "Idiocracy," of course, but also Stephen Fry's Evelyn Waugh adaptation "Bright Young Things" and bring back to the conversation John Huston's "Man Who Would Be King" during the Afghan War, but, in general, I'm too old to care whether I'm succeeding in imposing my personal tastes upon the world. (Which is good, because I'm not!)
Instead, I see my job in my movie reviews as adding value. Some critics do this by being amusingly snarky, but I'm more earnest. I go read the book, Google the history, think about the issues the movie brings up. For example, my review of the Oscar-winning "Capote" came to the same conclusion as everybody else's: Philip Seymour Hoffman is great! I don't have the skills to explain much about why Hoffman was great within 735 words (but not too many other critics do, either). I do think I made some contribution by reading Capote's In Cold Blood and pointing out that the film screenwriter's unthinking bias against capital punishment made the film unfair to Truman Capote and his small but influential place in literary history.
Anyway, I don't think it's very hard to give a reasonably accurate thumbs up or thumbs down on individual movies, and most critics aren't bad at that aspect of the job. As I wrote to The Audacious Epigone:
There is a fair amount of agreement between critics and the public on movies within various classes: e.g., that "Saving Private Ryan" was better than "Flags of Our Fathers," or that "Gladiator" was better than "Alexander." Movies tend to "work" or not work, rather like a good band within a musical genre is pretty clearly better than a bad band.
Fortunately, he took me up on this and crunched the numbers on the correlation between critics' ratings and box office dollars. In "Movie Critics and Movie Mavens," he reports:
Steve's comment is borne out for the most part--critics do well ... by genre, except for horror (which they despise) and romance (which, not surprisingly, they just don't seem to get). Criticism and cynicism are often mistakenly thought to be synonyms, and this provides some justification for that confusion, as these genres are the two most susceptible to it. In action and science fiction, they do very well (.55 and .59 r-values, respectively). The disdain for horror extends across the board, as it received the lowest average score of all of the genres considered. For drama and children's movies, they perform about as well as movies at large.
Drama films are most likely to be heavily politically or culturally ideological, which is probably why critics don't do so well in this serious genre. It's similarly a tough one for the general public, as the characters and the actions they take are often judged in many gradations, open to more interpretation than say, concluding that Scar is a bad dude.
Fat Knowledge, in wondering how the general public fared compared to the critics, pointed me to Yahoo, which constructs an average user score based on thousands of online ratings. The folks obliterate the critics. Rotten Tomatoes' critic scores, the most reliable relative to box office receipts of the different services looked at, correlate with revenue at .295 to the Yahoo users' .405.
While the critics do a little better by genre than by all movies in general, Yahoo users beat the critics in every genre, excepting action and science fiction, by slim margins. Romance is the most glaring. Critic scores and box office performance correlate at a meaningless .06 to users' .77. That blossoming could never happen in real life! But two cowboys... Horror was similarly divergent. Apparently stuffy critics cannot degrade themselves enough to review horror movies with any seriousness--they all belong in the garbage bin!
The genre-to-revenue relationships (r-values) for professional critics, Yahoo users:
Genre, Pros, Yahoos
Action: .55, .48
Comedy: .41, .58
Drama: .28, .38
Horror: .09, .61
Kids: .27, .79
Romance: .06, .77
Sci-Fi: .59, .55
Thriller: .31, .42
Simply put, if you want to know how a movie will do, ask the moviegoing public that will go to see it. Not only do the movie-maven plebeians do better than the putative experts at predicting actual box office performance, they're a lot more stable across the board, always providing at least a moderate amount of insight. Of course, uppity critics would hate to be amalgamated as a group, so unique are their individual opinions! Find a critic that you feel to be insightful, and his criticism becomes valuable.
The data, via Swivel, is available here.
One addition that could be made to his analysis would be to make it a multiple regression analysis and add in budget as independent variable. Critics tend to grade on the curve, demanding much more from a $100 million movie than a $1 million dollar movie. (Perhaps we assume that the marketing budget is proportional to the production budget, so we can help a small budget movie more than a big budget one.) For example, the tiny Irish musical for heterosexuals "Once" got huge ratings on Rotten Tomatoes, with 97% positive reviews. On an absolute scale, it's really not that good, but it's the "Citizen Kane" of $150,000 movies. And the public seems to agree: it's made $6.5 million domestically, which is a huge return on investment.