What's the idea?
For every film, we list the
ratings of several prominent critics. Some are the usual
suspects, from major publications, but we also try to use
critics from different parts of the country and from
different media in order to give a more representative
idea of the national Critical Consensus. The "median"
-- the middle or average rating -- is represented by
an orange frame around the critic's name, and under the
Dueling Critics we include two quotes, one from the
critic who gave the film its highest rating and one from
the critic who gave it its lowest rating. There are many review
hubs on the web, incorporating all sorts
of rating formats, from spoiled produce to traffic lights.
But we humbly submit that ours is by far the most
innovative in implementation: where critical opinion falls
on a particular film is evident at a glance since the star
ratings are basically a pictograph; more positive reviews
mean more 5-star ratings, which results in a fatter column
of stars.

How do you decide what the average rating is?
There are three ways statisticians calculate the measure
of central tendency: The most common way is the "mean" --
also referred to as the "arithmetic average" of a list of
numbers or data. If you remember your high school math,
you calculate it by adding all the numbers in a series and
then dividing by how many numbers there are in the series.
The mean wouldn't work in terms of calculating the average
rating for a particular movie since it would not show
where critical opinion tends to cluster. The "mode," the
second way of figuring out averages, is the value that
occurs most frequently in a series. And, finally, there's
the "median," our preferred -- and, for our purposes, the
soundest statistically -- measure of central tendency: It
is simply the value that appears in the middle of a
series. In other words, half of the critics' ratings are
above it and half are below it (more for masochists:
See That Tricky Median: A Practical
Explanation).
Do you assign the star ratings?
Most of the critics we include in our critics' ratecards
assign their own star ratings -- or their editors do it
for them. Several do not, and so we must assign the star
rating ourselves by reading the review and deciding how
much or how little a critic liked a movie. Star ratings
are controversial with some critics, and a few absolutely hate them. But star ratings do
provide the best graphical representation of critical
opinion. So, where a critic does not assign stars and we
have to do it, regard them as our interpretation of his or
her opinion. We use a 5-star scale (that's from 0 to 5
stars) at ½ increments, which is the standard in the U.S.

How do you choose which critics you will feature?
Let's face it: no list of critical opinion could be
considered complete without a few prominent critics like
Roger Ebert or whoever writes for the New York Times right now.
But prominence is only one criterion we use for selection.
We also strive for geographic diversity, so we can offer
you a representative picture of the national critical
consensus. Circulation is another consideration, since a
critic is bound to be more influential if he has a large
audience. That said, we also try to mix media and
therefore include print publications (newspapers and
magazines) and web sites, even some small web sites that
do not have a huge audience but do have a loyal following,
and hence credibility. On the other hand, we try to avoid
publications whose principal focus is celebrity coverage
and not serious film criticism, no matter how large an
audience. There are also a couple of practical
considerations that prevent us from using several critics:
Since we try to complete the critics' ratecards on the day
a movie opens, we are forced to exclude several worthy
media outlets (mainly sites of monthly magazines) that
publish their reviews days or even weeks after a movie has
been released. Then there are many sites (even sites of
major newspapers) that do not maintain an easily
accessible database of reviews, or use frames, which make
it hard to link to them.

Do you only feature U.S. critics?
Mostly, yes, although more for pragmatic purposes than any
misguided nationalistic principle. There are two major
reasons: Most major movies open stateside first, and since
we're a U.S. site our readers are mainly English speakers.
As the web grows and changes it may become easier in the
future to list critics from other countries.

What do I do if a link to a review doesn't work?
Some web sites have the nasty tendency to move their
articles and to change the URLs, thus making a mess of the
first principle of the web, which is infinite hyperlinking
between sites. If you spot a misguided or missing link,
please email us at .

What's your copyright policy?
Our features are copyrighted. That means they belong to
Critics Inc. That extends to any editorial content and
editorial constructs, as well as the website design,
format, programming scripts, code, stylesheets, etc.
Furthermore "Critical Consensus" & "Dueling Critics" are
Service Marks of Critics Inc. (plus we may claim
additional trademarks and service marks not listed here).
At the same time, we're not copyright fascists: yes, of
course you can print a page for your own use or to share
with friends; yes, you can e-mail a page to a friend; yes,
go ahead and link to our site (please!); and, yes, sure,
go ahead and look at our HTML code and be inspired and
improve your site, However, any commercial reproduction of
anything appearing within this site without our written
permission is illegal and in really bad taste -- and, yes,
we will come after you.

What is your privacy policy?
Our privacy policy is very simple: we do not collect any
personal information on our members and we do not share
the e-mail addresses of those who write to us, whether
members or readers, with any other party.

What is your ad policy?
Ads appearing on Critics.com are delivered to you by
third-party companies: We allow third-party companies to
serve ads and/or collect certain anonymous information
when you visit our web site. These companies may use
non-personally identifiable information (e.g., click
stream information, browser type, time and date, subject
of advertisements clicked or scrolled over) during your
visits to this and other Web sites in order to provide
advertisements about goods and services likely to be of
greater interest to you. These companies typically use a
cookie or third party web beacon to collect this
information. To learn more about this behavioral
advertising practice or to opt-out of this type of
advertising, you can visit
http://networkadvertising.org/managing/opt_out.asp.

How do you choose your ads?
A word about our online ads: We hate distracting and
misleading ads as much as anyone, and many of the pop-up
ads we feature on our site are very irritating indeed.
However, we don't have much control over advertising on
the site since we do not handle our advertising in-house.
Since several online ad agencies handle all of our
advertising for all of our sites, we obviously we do not
control what advertisers they get.

THE
TRICKY MEDIAN

Take the 15 kids in Mrs. Jones' local third grade class
and measure their heights, and calculate the average. The
average height might be, for example, four feet, six
inches (54 inches). The average was obtained by measuring
every single child, adding their heights together, and
dividing by the number of children.

The median height, however, is calculated differently, and
it is intended to give a different view than the average
height. To get the median height, take those same 15
school kids, but first you must line them up carefully in
order of their height, from shortest to tallest. The
median height is then obtained simply by measuring the
height of the 8th child in the line (the one who falls
exactly in the middle). In this case, let's say Joey,
standing in the middle, is just 50 inches tall.

We would then report: "The height of the children in Mrs.
Jones' third grade class is an average of 54 inches, with
the median being 50 inches."

Averages and medians are intentionally two different
calculations, with each giving a different angle of view
concerning the mathematical "middle" or "center" of a
series of measurements. Sometimes the average is close to
the median, and therefore many people think of the average
and the median as meaning pretty much the same thing.

Indeed, sometimes the average and median values are
coincidentally the same. But this is certainly not always.
The two numbers are more likely to coincide when you have
a larger quantity of things to measure, and especially
when you are measuring naturally occurring forms. For
example, in measuring the height of all 500 third graders
in the local community, the average and the median might
both turn out to be 53 inches, or they could easily be two
numbers very close together, such as an average of 53
inches and a median of 52.5 inches.

But more often than not, the median is not the same as the
average; many times it isn't even close to the average.
This variance is proper and expected, because they are two
different calculations for two different things. The
average and median are most often different when measuring
things that do not follow such a natural pattern of
distribution. They include things like personal income, or
business income, and personal opinions -- as in movie
ratings.

The difference in the average and the median height in the
first example above occurs because the calculation of the
average height takes into account the precise height of
each single child by first adding them all together and
then dividing by 15. The average is therefore influenced
by every height, including the unusually tall children way
over at the right side (the ones who are drinking too much
milk), and by the very short kids at the left side (whose
growth has been stunted by smoking).

The median height, however, is not impacted by every
single height, except during the initial stage of
arranging them in order of shortest to tallest. During
this first step (essential to calculating the median),
each height matters because it affects how you line them
all up. But, once the children have been so arranged,
there is now no need to precisely measure any of their
heights except that of the child in the middle. The median
or "middle" height is the height of that one child, but
only because he or she happened to fall into the center as
a direct result of the pre-arrangement process.

Here it is important to note that the median is not
influenced by any unusually tall or short children that
might appear at either end of the line. For example, if
you replaced the child at the right end (60 inches) with a
much taller child (seven feet), the median would not
change one bit. The median isn't impacted, really, by
anyone but the kid in the middle, except to say that every
other kid's height during the pre-sorting process
contributed to making one particular child fall exactly
into the middle. The child so chosen receives the glorious
honor of representing the median height for that
particular class.

In most elementary schools, the child in the middle would
be flanked immediately on either side by children of very
similar height. As you move away from the center toward
the children at the either end, the heights would
typically tend to slope upwards or downwards rather
gently, with maybe a few small jumps here and there,
especially if the children are all similar in age. This
assumes the class is not a wild conglomeration of kids of
widely different ages (which would impact natural height
distributions).

The median and the average are both useful measures of the
"center" of a set of values. But a median, as noted above,
is much more stable than an average. In the same way, the
median rating of a film by a group of 15 movie critics is
far more stable than an average rating. Some people prefer
the average rating because it is a more familiar number,
while some (like us) prefer the median for its greater
stability.