“I want the numbers!” – Josh Lyman, ‘Bartlet’s Third State of the Union’
All of IMDb’s user-ratings data is available to download as a csv. I was curious about what patterns I could find in the data of my favourite show The West Wing.
I did a quick clean up of the data created a rough bar chart and some scatter-points.
I made some personal decisions about how to use the data. Everyone who is rating these seasons is a West Wing fan so all the ratings are pretty high. I had the graph rating axis go from 6.5-9 rather than 0-10 to give a better sense of the variation between seasons. I have also chose to add Issac and Ishmael – the special non-sequential episode in response to September eleventh – as a zeroth episode of season 3, as did Wikipedia. Where episodes were shown as two-part specials on TV and subsequently as single episodes on DVDs and repeats I have prefered the format that the data used.
The excel data I used is available from my Dropbox if anyone wants to do a better job.
I have some brief observations:
|Season||Average Rating||Standard Deviation|
Season 1 and 2 are brilliant, setting the quality level for the rest of the show. Neither season has more than two below average episodes.
Looking at the scatter graph it seems that season 4 has the largest variation, with a large range of episode ratings. Each subsequent season tightens up becoming more consistent. This may have something to do with Sorkin’s personal problems. His very public arrest for drug possession in April 2001 (just as the final episodes of season 3 were airing) must have made his life increasingly difficult.
Not surprising to me is that season 5 is the worst season with an average rating of 7.7. However, I was surprised there wasn’t a bigger difference between it and season 4 (rating average of 7.8) – the last season Sorkin wrote. I remembered the quality difference as being much greater. Perhaps this is just because I like Sorkin’s particular style more than the average fan or perhaps it is because season 5 seems to start with a run of bad episodes.
I was curious enough about how good Sorkin was to do a separate analysis of whether Sorkin wrote the episode or not. I got the writing credits from this Wikipedia page. From his Wikipedia page I got this quote from past West Wing writers said, “even by the spotlight-hogging standards of Hollywood, Sorkin has been exceptionally ungenerous in his sharing of writing credit.” We can be pretty confident then that any time another writer got a credit they really earned it. I ignored the distinction between “teleplay” and “story”. All the story credits were ignored. Perhaps that is unfair but it is just too difficult to go that far into the analysis.
If it isn’t clear enough by the scatter graphs, there doesn’t seem to be much difference between episode written by Sorkin and Episodes written by others. There is some difference though. On average about 0.13 rating points, or about 1.6% of the average. That is not much, it might not even be statistically significant. Then again one guy who is able to produce consistently popular episodes is actually quite impressive.
It might be useful to add the show episode ratings data to this analysis. Perhaps the interest of less committed fans – interested enough to watch it on TV but not interested enough to log a rating on a website – varied more.