Monday, March 3, 2008

Signalling and sequences

One of my daughter's friends suggested that sequels would, on average, recieve lower scores than the original movies - as, at least in her experience, they were invariably worse. I thought I'd just confirm her suspicions so that I could let her know that she was thinking about the problem in a good way.

However, to my surprise the opposite appears to be true. Here is the mean score - the 0.5879992 number (adjusted for various things) for each episode of Sex in the City.

Sex and the City: Season 1 0.5879992 41138
Sex and the City: Season 2 0.5824835 43795
Sex and the City: Season 3 0.6523933 38983
Sex and the City: Season 4 0.7066851 34616
Sex and the City: Season 5 0.7359862 33380
Sex and the City: Season 6: Part 1 0.8097552 33532
Sex and the City: Season 6: Part 2 0.8241694 27914

As you can see the later the sequel the better the result. This seems, at least to me, counter intuitive - However the answer may lie in the second number which is the number of people who rated the movie. It seems that once there have been (in this case two episodes), then people who don't like the movie drop out and don't watch any further. So although less people watch it, they give a higher average rating.

This might be interpreted as some form of signalling. If a movie can accurately 'signal' to its potential audience that it is worth watching then the average rating will be higher. Interestingly this is - potentially - in conflict with the aim of the movie companies who might want to maximise the number of people watching irrespective of what they think of the movie (at least in the short term).


eblair21 said...

Might the ranking of movie sequels (I agree would be less popular) be different that subsequent seasons of a popular tv series? For your Sex in the City example, I fully except the upcoming movie to be less popular than the tv series, and if they make a second movie that it would be less popular than the first.

mathetos said...

I agree, tv series' are a completely different animal than Spiderman 1, 2, 3, or Rambo, etc...

Your theory would be more interesting with a traditional sequel rather than tv series'.

Another interesting angle would be to rate trilogies (Matrix, Lord of the Rings) against sequels, since trilogies are typically one really long story, compared to a sequel which just rehashes the same idea in a different way.

John Barrdear said...

I wonder if it might depend more on artistic continuity. A continuing TV show will often retain the same writer(s) and director. When the same people "make" a movie sequel as made the original (I'm thinking the Matrix and X-Men), it seems to do well, but when the artistic licence is handed to someone else, it tanks (e.g. the last X-Men movie).

Geoffrey said...

With movies, I think your daughter's friend was probably right, but with TV shows (as your example showed) I would think the opposite it true.

I would think that with the majority of recent tv shows, that the earlier the season (thus older), then the lower the rating, simply because tv shows tend to get dated.

With movie sequels I think sequels fall into two categories.

Category 1: Legitimate sequels (same movie stars, etc)
Category 2: Straight-to-video sequels(usually for movies that weren't very successful to begin with, these are the sequels where all of the stars are replaced with lesser known actors)

Stefan said...

The first seasons are usually low budget. As they get more viewers, they get more budget. More budget gives better quality, which is reflected in the higher ratings (together with an increase in viewers, as rumor spreads about "this new cool series").

This is valid until the series has jumped the shark (like Fonzie did).

jacob said...

y'know, have you taken a look at amazon's mechanical turk? That would certainly provide some sort of human element at a relatively low cost.


William said...

Indeed, it would be interesting to do the same exercise by splitting tv series and movie. Is this information available in the netflix database?
+ It would be also interesting to dig deeper into this and to statistically prove it (using ANOVA).
anyway, interesting thought!
good luvk

Nicholas said...

You hypothesize that the ratings are higher for sequels because the people who don't like the first film don't bother to watch the sequel. It seems to me you could test for this by extracting a new average rating for the first film taken only from those who also rate the second film. That is, compare the ratings of the two films only among those who rate both, and see if there's a drop.

Tim said...

I think ratings follow a sort of curve:

1. The rating for the first episode or season is whatever it is.

2. The rating for the second is a function of how "good" the first is (people will try the sequel if they liked the first)

3. Third and subsequent are dependent upon the rating of the second, not the first (If the sequel was a stinker, then people won't try any subsequent ones)

A categorical example I would try analysis on would be Jaws.

Tim said...

I'll add to my post that the curve usually "tails off" with age, for TV series (someone already mentioned Happy Days and Fonzie) but not necessarily for movies (Bond, Die Hard, perhaps Indiana Jones)

Both posts based on the idea of whether people will recommend or want to try the sequels, not whether they actually like them.

Nikos said...

This blog has'nt changed much recently!

Ann said...

But that's a tv show, isn't there some sort of difference between watching tv series verses movie sequels? People become "hooked" to tv shows since there's more time to flesh out characters and overall there is a consistency in quality, whereas movie sequels aren't always guaranteed to have the same amount of quality as the original, especially if the cast and crew is altered.
Just a thought- how many users put a whole tv series on their queue, watch one, then delete the rest because they disliked the one dvd they received?

jason_watkins said...

I agree with the other posts. I'd expect an increase with each season of a tv series due to users who are less interested in the show abandoning it earlier.

As an aside, it might be interesting or amusing to predict the "jumped the shark" moment of various tv series.

Newton Particle said...

It would make sense as people are more likely to give a higher rating to a series due to the greater familiarity and knowledge of the characters.

In addition , unlike a film, a series is made up of multiple episodes so the rating/impression one would have is derived from the sum of episodes in said series.

On top of that, there is the value for money (total length) that a series has over a movie.

karl said...

this is selection bias.

someone is more likely to watch the sequel if they liked the first one. And, if they like the first one, although they might not like the second one as much, they still like it.

to check this hypothesis, you should see how individual users rated sequels compared to the original

umph said...

TV series ratings tend to go up with time as people get familiarized with the characters and story line. There is also a "memetic" factor involved, as more people can be counted among the "watcher" group, the put social pressure on the "non watcher" group.
The curve seems to peak at some point though and that's harder to predict.

Dusty said...

I agree with everyone above who says that the rating of subsequent television series seasons will be much different than movie sequels here's why:

A TV series with multiple seasons has more opportunities to evolve and "come into its own" than a series of 3 movies. With a TV series you have time to see what works episode to episode, and adjust accordingly.

A series of movies has one shot per sequel to improve the plot, characters, etc. The novelty of whatever the original premise was will now be worn off, and it's up to these other factors (or the strength of the original premise) to keep the pace. That's challenging.

It makes sense that a TV series would more often get better with subsequent seasons, and movies worse with sequels. I think you should go back and look at your daughter's friend's suggestion and try it only on movies.

Tim said...

No updates in a while?

Had another thought about sequels... you'd need external data - but with movies, a sequel's quality (and therefore ratings) could depend upon box-office profits from the preceding entry in the series?

Mark said...

While I would in no way consider myself an expert on anything, I do have one observation and it has to do with one simple base in logic.

In your hypothesis a conclusion seems to be made of a generality, but only one sample is produced, it would be impossible to say that your daughter is incorrect (or even correct) based on the model presented.

Mark said...

Clarification, your blog mentioned your daughter's friend and not your daughter.