I HATE the Meanings of Netflix Star Ratings !

[Serge]

Netflix uses the following meanings for its star ratings:

5 Stars = "Loved It"
4 Stars = "Really Liked It"
3 Stars = "Liked It"
2 Stars = "Didn't Liked It"
1 Star = "Hated It"

So what's wrong with these meanings ? Here's my main gripes:

(1) What does "Really" in "Really Liked It” supposed to mean? That when you rate a movie as “Liked It” you are just kidding ? Or that you speak like a 8-year old, and express an intensive like of a movie as "Really Like" ? But than how would you rate a movie that you really, really, really, really, really liked ?
(2) The rating is not symmetrical: we have “REALLY Liked It”, but not “REALLY Didn’t Liked It”. Subsequently the intervals between 1 and 2 Stars are larger than between 5 and 4 stars.
(3) Is “Hated It”, really opposite of “Loved it” ? Can’t you hate a movie you love ? Isn't indifference the opposite of love ?


As an alternative I propose the following meanings:
5 Stars = "Loved It"
4 Stars = "Liked It MUCH"
3 Stars = "Liked It MODERATELY"
2 Stars = "Liked It LITTLE"
1 Star = "Didn’t Liked It AT ALL"

Here’s the advantages of the meanings above:

(1) It’s more accurate and consistent, since it measures only one dimension, “Likeability” of the movie, and it doesn’t concern itself about other feelings like “Hate” the viewer might have while watching the movie.
(2) It’s symmetrical. The intervals between stars are roughly equal.
(3) It forces the viewer to recapitulate all they liked about the movie they’ve watched, and the intensity of that feeling. Rather anything they’ve hated it.

While Netflix spends millions on developing a better recommendation algorithm, it should give more thought about the meaning of its rating system: make sure it’s not confusing, ambiguous or overlaping. It must be simple but not juvenile ( “really liked it” and “hated it” do sound juvenile to me).

[Pihk]

You're not obligated to use Netflix's meanings. Unless you are really dependent on their recommendation engine, and even then, I guess it would be fine so long as you are consistent.

So basically, you can interpret those 5 stars however you want.

I used to "grade" movies on IMDB's 10 point scale (9-10 = A, 8= A-, 7=B+, etc) just like academic marks a few years back and then divide by 2 to translate to Netflix while griping that they had no half stars.

Netflix has stated that 1/2 stars confuse most of their subscribers. My first thought when I read that was Wow, they must really think they're subscribers are drooling morons. Then I joined NFC and after 2 years, I realized that Netflix really understands the bulk of their customers very well.

These days, I just accept their system. Yes, most of my ratings are 3 or 4 but I realize there's little point in agonizing over the precision of that or desiring another increment in system for symmetry, etc.

The most precise way to evaluate a movie is to write a review and not an arbitrary point or star system. IMO

[Serge]

See. ideally, everybody should use the same rating meanings. Else the movie recommention engine will not work as well.

[Serge]

Besides, due to the rampant grade inflation in college, even a mediocre movie deserves an A-, or at least a B+ (or that what would most of my college professors would give when it comes to mediocre students) If you are slightly more than mediocre, like myself, you are expected to get straight A's on most college campuses. And so I did.

So because of my college experience, my perception of what deserves an "A" is greatly inflated, and like my college professors, I would be very reluctant to assign a "C" to a movie, else I would hurt its self-esteem. ;-)

[Knaldskalle]

Else the movie recommention engine will not work as well.

How do you know this?

[Serge]

The movie recommendation engine uses an algorithm that factors the ratings of millions of users, and if different users interpret the star ratings in different ways, the algorithm would not able to generate very good matches.

Analogically, if various drivers interpret traffic signs in different ways, even if similar, you won't expect great traffic flow would you ?

[Knaldskalle]

That's why there are laws governing traffic and not Netflix ratings.

People already perceive movies differently, hammering out specific rules for each rating is not going to change anything.

[Dr. Anton Phibes]

Star rating system that is universal...

☼☼☼☼☼ I would trample small children to watch this movie again.

☼☼☼☼ I would injure someone with a cell phone texting habit if they sat next to me during this movie.

☼☼☼ I could be made to sit through this movie again if a promise of sex were involved during or immediately afterward.

☼☼ If I have a choice of watching it or cleaning the roof gutters I'll flip a coin.

☼ If I ever even accidentally see this movie again I will be forced to boil my eyeballs in bleach

[Cheryl]

☼☼☼☼ I would injure someone with a cell phone texting habit if they sat next to me during this movie.

I would injure anyone texting during a movie whether or not I liked the movie. It's just damn rude and distracting.

[Dr. Anton Phibes]

good point.

[thatboy]

I'm surprised someone so picky about definitions isn't equally picky about spelling and grammar.

[Serge]

The primary reason I am concerned about rating definitions, is to enhance the movie recommendations, not because I am being pedantic.

[Knaldskalle]

How does changing the description of the star rating change the rating?

Does the 3 star Amazon rating ("okay") differ from the 3 star Netflix rating ("liked") in any substantial way?