1st debate TV ratings

[cinyc]

Here's some early Nielsen data that includes the cable news networks:

http://money.cnn.com/2016/09/27/media/debate-ratings-record-viewership/index.html?sr=twpol092716debate-ratings-record-viewership0245PMVODtopLink&linkId=29248123

So, sounds like they're saying total TV viewership was ~81 million, the same as the 1980 Reagan-Carter debate.  But since there were also some people who watched it online, total viewership would be higher than the 1980 debate.

Of course, the US's population is a lot higher today than it was in 1980, so 80 million might not be as impressive as it once was.

There were at least 2 million people watching the NBC youtube stream, and this number doesnt include PBS and CSPAN (I think)

The NBC YouTube stream doesn't count for TV ratings, and I doubt CSPAN is rated by Nielsen.  PBS might be included in the final tally when we get it around 4PM eastern, though - but I don't remember if those will be Nielsen numbers or self-reported PBS numbers.

The overnights we got this morning are always subject to revision, especially for live programs like the debate.  They usually don't take into account the pre-prime time viewership on the West Coast, when the debate aired.

[cinyc]

CNN is reporting total viewership of 80.9 million before PBS.  I'm not sure if that includes time-adjusted numbers for the broadcast TV networks.

From the Hollywood Reporter:

You must be logged in to read this quote.

[cinyc]

So how does this factor in places like bars and other public gathering sites? I know a lot of people who went out to watch the debate in public places.

Nielsen ratings don't count those people, just in-home viewership.  I don't think they ever have counted out-of-home viewership, so it's an apples-to-apples comparison.

[cinyc]

How much higher do you guys think these numbers would be if it included that segment of people?

That's impossible to say. There is no way to account for who was watching in such environments. Plus, think about all the homes where multiple adults were watching, which there is no feasible way to account for that, either (?)

But given the size of these ratings, and those 2 points above, it might be fair to say that a majority of the country tuned in or at least will inevitably catch parts of it over the next week.

I think Nielsen can track how many people are watching in a Nielsen household.  Nielsen provides ratings for both households and individuals watching.  The universe for individual ratings in most reported results is usually Persons 2+, not adults.  Nielsen also provides sub-group ratings to networks who buy them, like Adults 18-49 (the key money demo), adults 25-54 (the key news demo), etc.

[cinyc]

I think Nielsen can track how many people are watching in a Nielsen household.  Nielsen provides ratings for both households and individuals watching.

How big is their sample, though? They aren't in everyone's home. I mean how accurately can Nielsen judge how many people are watching together in homes out there from what limited number of households they are in?

That's a secret.  Some sources say there were about 25,000 Nielsen Households in 2013, others say there are about 50,000.  Nielsen put out a press release saying they increased the number of Nielsen families in 2014, but didn't disclose how many total Nielsen families there were.

Statistics don't require a very large sample to get an accurate rating, though.  25,000 randomly-selected households should be representative of the TV watching universe, in theory.

[cinyc]

No, the 80 million is supposed to be people, not households. 

To be fair, I only meant 'households' as in the way they determine those statistics. They mean people, but the way they count it for most of that is by household, right? For instance, for each cable customer who uses their TV to tune in to some network to watch, they count that as 1 person, even though they could have a whole family watching with them. There is really no way for the companies to know how many people are actually watching. The viewers over the internet could probably be more accurately assumed to be 1 person, but still.

Anywho, given how media is consumed, we'll probably never know how many people truly watched last night. It could be considerably higher when factoring in groups/families, no?

Yes, Nielsen does know who is watching in a household - at least in theory.  Family members are supposed to check in on the electronic people meter (or whatever technology Nielsen is using these days) when they start watching a program, and check out when they stop.  I don't know whether Nielsen has automated the process with new technology like they use for video games that allows boxes to tell how many people are in the room yet.

Nielsen provides both household and individual ratings.

[cinyc]

There's no press release from Nielsen yet, but the final broadcast network adjusted figures are in, and various media outlets are reporting an audience of 81.4 million.  My guess is that figure doesn't include PBS (but the Nielsen press release likely will).

[cinyc]

Nielsen's press release is out.  84,011,000 viewers watched the debate on TV, including PBS.

[cinyc]

Nielsen's press release is out.  84,011,000 viewers watched the debate on TV, including PBS.

Nearly 100m people saw Trump's disastrous performance, great!

84 million << 100 million.  The media expectations of 100 million TV viewers were wildly exaggerated.

[cinyc]

Nielsen's press release is out.  84,011,000 viewers watched the debate on TV, including PBS.

Nearly 100m people saw Trump's disastrous performance, great!

84 million << 100 million.  The media expectations of 100 million TV viewers were wildly exaggerated.

I think if you included streamers and people dvr-ing and watching it later, you'll hit 95-100 million, just not 100 million live TV watcher.

I doubt it.  YouTube only had 2.5 million live streamers, which covered many of the network broadcasts.  And very few people dvr live TV like political debates.  Maybe you reach 90 million after streaming and dvring.  I doubt you'll reach 100 million.  While Nielsen does measure dvr-delayed TV watching, it doesn't measure streaming.

Plus, the media was expecting 100 million TV viewers, not including streaming, anyway.  The debate ratings didn't come close.

[cinyc]

K, and your point?

It's still the most watched debate in over twenty years ...

My point is although it is the most watched debate EVER (not just in over 20 years) in terms of total viewers, the media hype of it getting 100 million TV viewers was greatly overblown. 

I also put up a poll yesterday for Atlas users to predict the TV viewership.  ElectionsGuy came closest by guessing 85 million.  I was second at 82 million.  The most common chosen range was 80-90 million, although the median and mean were higher than the mode.

[cinyc]

Anyone know what kind of international audience this thing got (personal anecdotes are fine)?

I don't know. Probably very low.  The Canadian broadcast networks didn't air it last night, and I can't imagine the ratings are very high in Europe for a debate that started at 3AM CET. 

You must be logged in to read this quote.

http://www.thewrap.com/trump-clinton-presidential-debate-tv-ratings-history/