Headlines from the analysis
Supporting analysis will be uploaded over the next few weeks.
Critics and movie fans largely agree on which movies are good and which are not. Statistical analysis shows a single underlying factor – “quality” – which accounts for most of the variation between scores. However much variation between individual opinions remains.
There is nevertheless a separate “critics’ view”, different from movie fans. This is much smaller than the quality factor, but still significant.
The wider public has quite different tastes. There is almost no relationship between what audiences like, as measured by audience research, and what film fans and critics endorse.
There is even less relationship with box office. Whatever measure is used box office success, has almost no relationship with what critics and film fans like.
The list of the 50 highest scoring movies on CineIndicator has a few surprises, and some notable biases. Films from well-known directors, and male-led stories do well. There are several very successful animated films. The list is dominated by older films, for which there are several explanations.
Rough and ready thresholds are useful, but some indicators say more. The distribution of scores (e.g. “What proportion of films score more than7.0 on IMDB?”) inevitably depends on which sample of movies you are looking at. But for a large and disparate sample, the change doesn’t prevent a threshold functioning as a useful guide.