Questions: Given the huge amount of critic and audience movie reviews, is it possible to identify when an individual or aggregate class of reviewers has more or less predictive power for critical success of a movie?

For instance a movie reviewer who is finely attuned to the quality of horror movies but does not have a refined reviewing taste for other genres of film. If this information is known, can an "accuracy score" be assigned to individual reviewers or aggregate reviewers as measured by their ability to predict critical success of a movie.

Dataset: Large movie review aggregators (Letterboxd, IMBD, Rotten Tomatoes)