I see a lot of fans plucking TV ratings out of a hat and making extrapolations and assumptions about them. I've always said that TV ratings in a vacuum are meaningless, if you don't control for all the variables that influence ratings- network, window, competition for viewership, etc. Essentially ratings in a vacuum aren't apples to apples comparisons where you can make a judgement about the TV attractiveness of a team or a conference.

I actually still think the same (not making a judgement about TV draws) applies to this article, but it does do a good job at least trying to make a little more of an apples to apples comparison. Some of the things it does do that I like:

1. Recognize that not all games are rated. Conference network games do not get rated.
2. Clarify that rating and audience are two different things. The rating on a Saturday is not the same as the rating on a Friday night, since the audiences are different.
3. Recognize that not all channels are equal. I hadn't seen a good summary of the network ratings like this:

4. Recognize that more games for some leagues are rated than others. For example, the KU/ISU game is on FS1 and gets rated, while the Indiana- Purdue game is on the Big 10 network and not rated. 75% of the Big 12 games are rated, much higher than the other conferences:


As I said, I still think there are other variables that could be controlled for. If you really wanted to make conclusions about the TV attractiveness of a certain team or conference relative to another team or conference, you'd want to control as many variables as possible- network, TV window, day of the week, relative rankings of the teams playing, etc.

Still, this does a good job of the aggregate data and a good read.

(12-17-2014 09:43 AM)Frog in the Kitchen Sink Wrote:  I see a lot of fans plucking TV ratings out of a hat and making extrapolations and assumptions about them. I've always said that TV ratings in a vacuum are meaningless, if you don't control for all the variables that influence ratings- network, window, competition for viewership, etc. Essentially ratings in a vacuum aren't apples to apples comparisons where you can make a judgement about the TV attractiveness of a team or a conference.

I actually still think the same (not making a judgement about TV draws) applies to this article, but it does do a good job at least trying to make a little more of an apples to apples comparison. Some of the things it does do that I like:

1. Recognize that not all games are rated. Conference network games do not get rated.
2. Clarify that rating and audience are two different things. The rating on a Saturday is not the same as the rating on a Friday night, since the audiences are different.
3. Recognize that not all channels are equal. I hadn't seen a good summary of the network ratings like this:

4. Recognize that more games for some leagues are rated than others. For example, the KU/ISU game is on FS1 and gets rated, while the Indiana- Purdue game is on the Big 10 network and not rated. 75% of the Big 12 games are rated, much higher than the other conferences:


As I said, I still think there are other variables that could be controlled for. If you really wanted to make conclusions about the TV attractiveness of a certain team or conference relative to another team or conference, you'd want to control as many variables as possible- network, TV window, day of the week, relative rankings of the teams playing, etc.

Still, this does a good job of the aggregate data and a good read.

The total audience seems to be the most reliable ranking. It just tells you which conferences has the largest audience, without all of the exclusions, variations, qualifications and excuses and maneuvering of data.
Surprised to see the Acc ahead of the BIG. But that could be because of FSU this season and many Big games being on the BIG network

SEC 197,427

ACC 116,367

BIG 103,167

Big12 97,804

Pac 12 89,505

(12-17-2014 09:56 AM)cuseroc Wrote:  

(12-17-2014 09:43 AM)Frog in the Kitchen Sink Wrote:  I see a lot of fans plucking TV ratings out of a hat and making extrapolations and assumptions about them. I've always said that TV ratings in a vacuum are meaningless, if you don't control for all the variables that influence ratings- network, window, competition for viewership, etc. Essentially ratings in a vacuum aren't apples to apples comparisons where you can make a judgement about the TV attractiveness of a team or a conference.

I actually still think the same (not making a judgement about TV draws) applies to this article, but it does do a good job at least trying to make a little more of an apples to apples comparison. Some of the things it does do that I like:

1. Recognize that not all games are rated. Conference network games do not get rated.
2. Clarify that rating and audience are two different things. The rating on a Saturday is not the same as the rating on a Friday night, since the audiences are different.
3. Recognize that not all channels are equal. I hadn't seen a good summary of the network ratings like this:

4. Recognize that more games for some leagues are rated than others. For example, the KU/ISU game is on FS1 and gets rated, while the Indiana- Purdue game is on the Big 10 network and not rated. 75% of the Big 12 games are rated, much higher than the other conferences:


As I said, I still think there are other variables that could be controlled for. If you really wanted to make conclusions about the TV attractiveness of a certain team or conference relative to another team or conference, you'd want to control as many variables as possible- network, TV window, day of the week, relative rankings of the teams playing, etc.

Still, this does a good job of the aggregate data and a good read.

The total audience seems to be the most reliable ranking. It just tells you which conferences has the largest audience, without all of the exclusions, variations, qualifications and excuses and maneuvering of data.
Surprised to see the Acc ahead of the BIG. But that could be because of FSU this season and many Big games being on the BIG network

SEC 197,427

ACC 116,367

BIG 103,167

Big12 97,804

Pac 12 89,505

That's the total exposure for each conference. What we can't make a conclusion about is the "attractiveness" of each conference for TV eyes. The mix of networks and TV windows affects those numbers, and needs to be controlled for to even start to draw conclusions.

One way to dissect the numbers would be to just compare games within each channel. E.g., compare one ESPN game to another ESPN game, compare one Fox broadcast game to another Fox broadcast game, etc.

The average ABC or CBS game draws more than 5 times as many viewers as the average ESPN2 game, so obviously anyone who has more games on ABC or CBS is going to have higher average ratings, and that tells us nothing about which games would be "most watched" if they all had the same time slot on the same channel.

As for comparing channels, these numbers show that the ESPN people, while they are braggarts, are correct in saying that FS1 has a long way to go to even catch up to ESPN2 ratings, and can forget about catching up to ESPN any time soon. (Fox also has a long ways to go in catching up on production and announcer quality for CFB games, but I suppose that's a topic for another thread.)

While I agree TV ratings just can't be taken at face value, anyone can manipulate statistics to make them say what you want to say by 'adjusting' them for various factors. Any college level statistics class teaches you that.

I tend to look at which matchups a network has to choose from, and which ones they place in the most prime time slots balanced with the least competition (obviously trying to maximize viewership). I think TV ratings are a red herring. This isn't a perfect way to see which games are given the most attention, because of a lot of different factors, but it shows what the networks think will draw the most. When their business depends on decisions like that, I give a lot more credence to it.

(12-17-2014 11:40 AM)Wedge Wrote:  One way to dissect the numbers would be to just compare games within each channel. E.g., compare one ESPN game to another ESPN game, compare one Fox broadcast game to another Fox broadcast game, etc.

The average ABC or CBS game draws more than 5 times as many viewers as the average ESPN2 game, so obviously anyone who has more games on ABC or CBS is going to have higher average ratings, and that tells us nothing about which games would be "most watched" if they all had the same time slot on the same channel.

As for comparing channels, these numbers show that the ESPN people, while they are braggarts, are correct in saying that FS1 has a long way to go to even catch up to ESPN2 ratings, and can forget about catching up to ESPN any time soon. (Fox also has a long ways to go in catching up on production and announcer quality for CFB games, but I suppose that's a topic for another thread.)

AGree and good points.

I think FS1 is making headway, but leagues with a bunch of games on it (i.e. Big 12, PAC-12) are going to be affected.

Here is the numbers of games from Matt Sarz site for each league. Doesn't exactly match the link's data, but pretty close:

ACC:
ABC- 8
ESPN- 18
ESPN 2- 6
ESPN U- 15

Big Ten:
ABC- 20
Fox- 1
ESPN- 5
ESPN-2- 16
ESPNU- 10
ESPN NEws- 5

Big Twelve:
ABC- 9
Fox- 8
ESPN- 7
FS1- 23
ESPNU- 1


PAC-12
ABC- 6
Fox- 11
ESPN- 12
ESPN2- 2
FS1- 11
FS2- 1

SEC:
CBS- 16 games
ESPN- 17 games
ESPN2- 7
ESPNU- 8

Not surprising, in the order of total viewership.

This is all good information. I am not sure what I am supposed to conclude. It was implied there are some misconceptions out there about tv ratings, but did not really list the misconceptions or what to do about it.

If it was implied that some conferences ratings are low only because their games are on a crappy low-rated channel, then that invites the question, maybe there is a good reason why their games are on a crappy channel. But I am not sure if thats what was implied.

(12-17-2014 02:25 PM)goofus Wrote:  This is all good information. I am not sure what I am supposed to conclude. It was implied there are some misconceptions out there about tv ratings, but did not really list the misconceptions or what to do about it.

If it was implied that some conferences ratings are low only because their games are on a crappy low-rated channel, then that invites the question, maybe there is a good reason why their games are on a crappy channel. But I am not sure if thats what was implied.

It was a response to this graphic:

http://texags.blob.core.windows.net/web/...rences.jpg


Even with the more detailed effort made by the original link, I think you have to be careful to make conclusions. There are a lot of variables affecting TV ratings. College football message boards tend to present them and ignore all the variables except for team and conference and make some sort of conclusion like team A is a good draw or conference B is a bad draw. It's much more complicated than that.

I'm fairly certain that if you controlled for all variables the SEC would still stand out, but I've yet to see an effort to dissect that appropriately, controlling for all variables. Certainly the TexAgs effort didn't try to control for anything.

I lumped the networks into the following three groups

Group 1- 5+ million/game: ABC, CBS
Group 2- 2-3 million/game: NBC, ESPN, Fox
Group 3 Less than a million/game: Everything else

Conference: Group1/Group2/Group3
ACC: 17%/38%/45%
Big 10: 35%/10%/54%
Big 12: 19%/31%/50%
PAC-12: 14%/53%/33%
SEC: 33%/36%/31%

The SEC had 69% of its rated games from Groups 1 and 2. That goes a long way as to explaining the gap between the SEC and the other conferences.

The Big 10 OTOH had the highest % from both groups 1 and group 3, and the lowest from Group 2. Most of the group 3 games were on the best of the group 3 networks, ESPN2.

PAC-12 is very group 2 heavy with games on Fox and ESPN. IT had the second highest % of games from group 1 and 2, but most of those were the less watched group 2.

Big 12 had half of its games on the less watched group 3 networks, and most of those were on the less watched FS1.

ACC had the most of anyone on ESPNU, which pushed their average down.

The SEC mix of mostly group 1 and 2 games and only 8 ESPNU games definitely pushes their average up.

There are natural cycles of when restaurants are busy and when they're not, but a restaurant has to pay the lease, utilities, etc for 7 days out of the week so it can't just be open on Friday and Saturday night.

Say the restaurant offers a salmon special on one Saturday night and a kobe burger the next. The salmon sells 150 times and the kobe burger sells 160 times. When offered as a Thursday special the salmon sells 75 units and the kobe burger sells 50 units. The manager makes the Saturday special the kobe burger and the Thursday special the salmon. The kobe burger is the top seller, but which is really more valuable to the restaurant?

That's why raw viewer numbers are misleading.

Your first mistake was actually thinking the texas aggies would produce anything that wasn't misleading and biased in their favor.

(12-17-2014 09:56 AM)cuseroc Wrote:  Surprised to see the Acc ahead of the BIG. But that could be because of FSU this season and many Big games being on the BIG network

ACC 116,367

BIG 103,167

The numbers only show 1 game difference: 96 to 95. So that is a bit surprising.

(12-18-2014 04:00 PM)adcorbett Wrote:  

(12-17-2014 09:56 AM)cuseroc Wrote:  Surprised to see the Acc ahead of the BIG. But that could be because of FSU this season and many Big games being on the BIG network

ACC 116,367

BIG 103,167

The numbers only show 1 game difference: 96 to 95. So that is a bit surprising.

The Aggie analysis shows Alabama first. But when you look at the detail, its easy to see how dominant FSU games have been this year. Bad boys sell. Last year it was Johnny Manziel. This year Jameis Winston.

(12-18-2014 04:00 PM)adcorbett Wrote:  

(12-17-2014 09:56 AM)cuseroc Wrote:  Surprised to see the Acc ahead of the BIG. But that could be because of FSU this season and many Big games being on the BIG network

ACC 116,367

BIG 103,167

The numbers only show 1 game difference: 96 to 95. So that is a bit surprising.

The shape of things to come?

(12-18-2014 04:00 PM)adcorbett Wrote:  

(12-17-2014 09:56 AM)cuseroc Wrote:  Surprised to see the Acc ahead of the BIG. But that could be because of FSU this season and many Big games being on the BIG network

ACC 116,367

BIG 103,167

The numbers only show 1 game difference: 96 to 95. So that is a bit surprising.

The shape of things to come?