[uf-discuss] hReview feedback
sfam at cyberpunkreview.com
sfam at cyberpunkreview.com
Thu Jan 19 21:10:59 PST 2006
Well again, Rottentomatoes does this every day (up-converting ratings from
whatever the author intended into their own rating system). Go there and
at random, check out a movie listed on their front page, look at the
number given for a review and then read the review itself - often you'll
see a difference. So if Ebert does a 2.5 star review, Rottentotatoes
doesn't automatically give this a clear corresponding number. They also
read the review to see where exactly it falls on their 100 point scale. We
all know they do this and nobody minds (so no, it doesn't matter), as the
final score is an aggregation across 100 or more movie reviews. The only
major difference on an individual review for them is whether the movie is
60% or above, in which case its considered to be "fresh" and gets a
corresponding fresh tomato symbol, or whether its 59% or below, which then
requires a "rotten" symbol.
As for where the bulk of the movies end up, on a 5 star scale, this
depends on the reviewer. Some make liberal use of the bottom end of the
scale, but most tend to put the bulk in the 2-4 range. Most reviewers
reserve their top mark for ones that will make their yearly "best of"
list. The other interesting thing is even if you give people a 10 point
scale, they will STILL want to uses half-stars. This is especially true at
connoting differences between the very best movies (meaning the use of the
half-star changes on a 20 point scale). So on a 20 point scale using 1-10
with half stars, you'll see lots more 8.5 and 9.5 star reviews than you
will say 6.5 star reviews (people generally don't care very much in
defining the difference between a 6 and 6.5 star movie), whereas on a 10
point scale going from 0-4 (in Ebert's case, for instance), you'll see 2.5
used lots more than 1.5 or 3.5. This is true because 2.5 seems to be the
bar for "decent" movies to reach. If its 2 stars or below, it's considered
to be a bad movie (this is again different from RT which definees 60% as
I will pose this question on a few boards though, and see what others
think of this. As I understand it, your big concern is wondering whether
explicitly defining the bottom number is critical (my guess is YES, most
certainly). Secondarily, you are interested in what scale most prefer?
Best, Noel Dickover
"Paul Bryson" <paul at msn.com> wrote in message
news:<dqnha0$pqi$1 at sea.gmane.org>...
> <sfam at cyberpunkreview.com> wrote...
> > Just a comment on this. I'm pretty well connected to the movie
> > reviewer community, and have been a participant on Rotten Tomatoes
> > discussion boards for quite some time (I have over 10,000 posts on
> > the discussion boards). RottenTomatoes.com (www.rottentomatoes.com)
> > collects the reviews from all the major movie reviewers in the
> > country, and translate it into a 100 point scale.
> You've probably looked at more movie ratings than anyone else in this
> discussion. How would you describe the grouping of ratings for movies?
> (What range are most movies in?) How do you feel about the importance
> of the accuracy of up converting from smaller scales when the
> difference could vary 2-9% for an improperly converted rating? Would
> this be enough to make a significant difference? Would it matter?
> microformats-discuss mailing list
> microformats-discuss at microformats.org
More information about the microformats-discuss