[what_ever]

But most of the people who has given the rating have not seen the movie. They are giving the lowest rating just to put their point forward. This is not a documentary to put the truth on the screen (I am not sure how much of the movie is fiction).

I think the author calls data erroneous as the rating does not represent the actual quality of the movie - a thing which IMDB wants to achieve. He doesn't call it erroneous for actual errors in the process of collecting ratings.

I think, as you said, you are just inclined to seek more nuances.

[thaumasiotes]

> But most of the people who has given the rating have not seen the movie.

In the (paraphrased) words of Andrew Gelman:

"I can't knock this guy for slamming my book without reading it. For example, I have never read the autobiography of Uri Geller, but I'm confident it's full of crap."

If you don't think Gonojagoron Moncho is lying to these people about the point they're taking offense over, how is it relevant that they haven't seen the movie?

[nkozyra]

> But most of the people who has given the rating have not seen the movie

I'm sure it's possible, but no rating represents the "actual quality of the movie." A rating is based on a subjective amalgam measurement that encompasses a myriad biases, even when watched and with total scrutiny.

[jorleif]

While this is true, I would argue that there is a social expectation that an IMDB review is written by someone who has seen the movie. Especially, it is expected to be true on average with the usual "few bad apples" thrown in the mix. In this case, this expectation does not hold, and in this sense these reviews are truly anomalous (they are produced by a different process than the other ones). From the perspective of the expected usage of the reviews this anomaly is indeed possible to see as an error, rather than just slightly different behavior.

[talmand]

While true, the ratings can be useful if there is some form of understanding as to what people are rating the movie for. Most ratings cover the simplistic "I saw the movie and this is how I feel about it" kind of data. In that case, the ratings can be useful.

But if a group with an agenda comes in and rates the movie on a different set of criteria without letting the typical reader know renders the data useless.