Large viewing figure fluctuations in arc-based TV shows



I find this sort of thing interesting. Here’s the data for Homeland

which is obviously not a great show but it seems like all shows have fairly significant changes in numbers of viewers between episodes of a season, usually with ‘big’ episodes having a larger audience.

Am I alone in finding this odd? Here’s GoT Season 5

There are 3 million people who didn’t give a shit they missed an episode in the middle and 2 million who barely bothered with episodes 2-9. Now I’m not saying all episodes are the same quality, but it seems odd these people apparently never watched the other episodes.

I dunno, another great 0 reply thread probably, @smee.


Ahm oot.


Nah, you’re totally in. I know you.


You still here?


hey theo to the nearest 10k how many of the 8.11m people who watched mothers mercy do you thikn watched none of the other 9 epsidoes?


not really. bunch of normies who just watch stuff if it’s on. maybe there was a pop idol final or something on the same night. they’ll just watch the recap next time.


But I don’t really get how you could just dip into this show and find it remotely compelling if you didn’t want to watch the rest?

The BBC figures used to have some way to include people who watched at a later date so they were compiled later in the week. I mean maybe it’s literally that they stop checking after a week so 2 million people just record all 10 episodes and watch them in a big block?


Okay, I’ll bite. Is it not that the “big” episodes are the ones which people want to see immediately instead of waiting for reruns/torrents/whatever?

Just had a thought, do they make any attempt to account for when groups of people are watching the telly, instead of one? Those numbers are surely all out of place. Maybe “bigger” episodes are more likely to be watched in groups, making those viewing figures lower?


Keep this bullying up and I’ll consider revoking your JDon btw.


Depends how people watch things too - like they might watch some of the episodes via a torrent or at a later date, but want to watch the ‘big’ episodes as they’re screened to be part of the event and the next day’s water cooler talk.

What’s also noticeable is that the average rating for episodes tends to go up as a series progresses and viewers are shed:


But wait, what I’ve said makes no sense. If that was the case then “bigger” episodes would have lower viewing figures? Oh Theo I don’t know I’m not a tellyman.


from what i understand about homeland, it’s a terrible show and people probably don’t watch when it gets especially bad


So the people who aren’t bothered stop watching and the average rises? Although the drop in viewers seems disproportionate. I guess these people were rating very low and later episode get even higher ratings from fans?


Homeland is a great ‘hatewatch’. I have Agents of SHIELD as a solo hatewatch. The TV has Walking Dead.



I have Noah idea what you are getting at.


That’s what largely drives it, although some series (eg Mad Men) build to a more obvious peak plotting-wise every season:


cool little tool that




Yeah. I think it would really benefit from being able to plot the number of viewers and/or the number of reviewers in there too - to test the theories.