[bluedino]

Luckily, IMDB also shows Metacritic ratings next to their own user ratings. I take those into account before I make my decisions on what I'm going to rent/stream.

Far too often do you find a movie (usually a somewhat recent release) that's just flat out terrible despite having a > 7 score on IMDB (and usually a Metacritic score in the 30's or 40's)

[edanm]

Do you have any interesting examples that you've found?

[jebus989]

Excuse the plugging of my own wares, but by coincidence I did some analysis the other day of the disparity between User/Critic ratings of movies on RottenTomatoes: http://benjaminlmoore.wordpress.com/2014/05/05/what-are-the-...

Example: "Spy Kids" got a 93% fresh rating from critics, but 45% from audience ratings.

Code attached should you want to investigate for yourself.

[icebraining]

If you look at the Average Ratings, the difference is smaller, though: the critics gave it 7.2/10 while the audience gave it 5.2/10.

[edanm]

That's really really awesome, thank you for sharing.

[canvia]

The Avengers was particularly egregious. It was in the top 250 movies of all time for opening weekend. A huge % of ticket sales occur in the opening weekend. There is not a doubt in my mind that these ratings are thoroughly manipulated by the corrupt studios. At least IMDB now shows the metacritic score which is more accurate. A lot of Disney/Marvel movies seem to follow this trend.

If I had unlimited time and money I would create a new movie review site that fights corruption. I would have a phone app for doing the reviews and it would require taking a photograph of your ticket stub at the theater and uploading it for proof (w GPS tag). It is far too easy and anonymous to participate in these online movie ratings sites that have a large impact on people's decision making. I have more ideas for the review site if anyone wanted to discuss it further.

[ronaldx]

I believe your observation is correct, that movies are over-rated at first, but I doubt this is wholly - nor even mainly - corruption.

People who are motivated to see a movie early are the same people who are likely to enjoy that movie. This makes it natural for movies to get rated very highly when they are first released. That's just natural.

If you forced people to photograph their ticket stub, I believe this could make the problem worse: mainly people who particularly loved the movie (or thought it was significantly worthy of comment) would be motivated enough to do that. You'll just collect the IMdB's 10s and 1s.

I know many people use the IMdB for their own record-keeping. These people would stop doing that if there was an administrative barrier that made it less convenient than a paper equivalent, e.g. writing in an address book. I strongly believe this makes their votes neutral/genuine compared to other survey methods.

[JoeAltmaier]

But some 10's and 1's are lazy indignant people. Putting an obstacle in the way of registering a rating could filter out lots of noise. Leaving people who deliberately, intentionally review movies. Those ratings would be of an entire different class.

[lyndonh]

But surely they don't use the mean to find the average score ?

Even a basic trimmed statistic would find a much more accurate average. Lazy people leaving 10s and 1s would be removed or given a lower weighting.

[Retric]

As long as it's consistent such ratings are averaged out as a 8.2 is only meaningful in comparison to other scores.

[bluedino]

Anchorman 2 and The Internship were both > 7 near their DVD release but have settled back down into the 6's and have low Metacritic scores.

[personZ]

This mirrors my own experience with IMDB -- it is of dubious merit for big budget recent releases, and I've always suspected (based upon nothing but a sense of "what else could it be?") that Hollywood has long understood the value of crowdsourced metrics like that.