Prior to 1984, the NCAA controlled television appearances for all College Football teams, including what games played in what part of the country.  One of the biggest hindrances to teams with this set up was that the NCAA limited the amount of broadcasts any team could have over a two year period.    It is a bit difficult to imagine today, when not being able to see your team’s games is the rarity, but each team used to be capped at six broadcasts over two years.

In 1984, however, the Supreme Court ruled in favor of Oklahoma and Georgia’s claim that the universities owned the rights to their team’s broadcasts, not the NCAA.  This immediately voided all of the NCAA’s television contracts and the College Football Association, which was comprised of 63 schools who also wanted control of their rights, stepped in to begin negotiating agreements for each team.   In a bit of trivia, this CFA was run by Chuck Neinas, who you may remember as the interim Big 12 Commissioner who gelled the conference during realignment and set it on its current path before handing the reigns to Bowlsby.

At the time, cable television was not as prevalent as it is now, so the games that were sold for broadcast were generally on the same “over the air” channels the NCAA used, e.g. ABC, CBS, and NBC who, to this day, still carry the largest audiences.  This effectively became what we now call “Tier One” inventory, which are the highest valued games featured nationally.  Even after this legal victory Tier One inventory still only accounted for a small percentage of the total games in a year.   The vast majority of the games were, basically, never broadcast.

Notre Dame shook up the system in 1990 when they broke from the CFA and negotiated their own Tier One deal with NBC, a version of which still exists to this day.   The SEC would soon follow with a deal to increase their exposure with CBS, who would televise twelve games a year.  Like Notre Dame, a version of this deal remains today, with CBS showing around fifteen SEC games a year, guaranteeing at least one game for each school and a championship game.

Every other conference since then has followed suit.


While more games are now sold in other “tiers”, which I’ll address in future analysis, the way these games in the top tier work hasn’t changed much since Notre Dame and the SEC broke off, except now each of the conferences sell their own inventory, instead of handing it through the CFA.

In the 2014-15 season there were 84 college football games broadcast with a combined audience of about 364 million people. This year there was an increase to 88 in the total amount of games on Tier One for the regular season, primarily due to ABC holding more games during the year.   The remaining three networks all decreased the amount of broadcasts by a handful of TIER2games each.   ABC, on the other hand, increased their total inventory 33% from 2014-15, most of it coming from early season, out of conference, “Made for TV” match ups and increased regional mirrors.

However, individual game viewership overall has dropped significantly since the 2013-14 season.  NBC has seen the sharpest drop, losing nearly two million viewers on average since 2013, but that was when Notre Dame had a championship run.  Since they have much fewer games than the other channels, they rise and fall considerably with the Irish’s fortunes.   Overall averages are also down significantly though.  CBS, the yearly leader in Tier One per game audiences, has lost a million viewers on average in the past two years and all of the Tier One channels have seen a 23% drop in viewership on average per game in that time.

Total viewership, when we add all of the games together, is up 4% from last year, eclipsing 380 million in combined audience for Tier One games.  So, in short, we are seeing a cannibalization of the market.   There are more games providing a larger total audience, but the greater choice in games and the increased amount of games aired is lowering the audience per each game.

Specifically for 2015-16, CBS led all channels in average viewership, averaging just over five million viewers per game.  ABC had the largest audience of all, with over 147 million combined viewers, but averaged a million shy of CBS per game.   The main reason for this is that CBS spends a lot of time and money marketing one game a week and it is normally always the 3:30 pm eastern slot for the “SEC on CBS”.  ABC, on the other hand, broadcasts three times the amount of games and has them all over any given Saturday, which, along with a wider variety of games, causes the averages to drop.

NMTIER3One thing that stands out when looking at the numbers, is that out of conference games, while valuable, are nowhere near as valuable as conference games.  This is why the Big Ten, Big 12 and Pac 12 have moved to nine conference games; they have more premium inventory to sell.  While ABC shows a slightly larger average for out of conference that is due to three big games, two of them in the first two weeks.

Ohio State at Virginia Tech not only had the highest out of conference audience of the year, with 10.5 million viewers, but it was also the only out of conference game to eclipse the 10 million mark. Surprisingly, it was also on ESPN, which is not a Tier One channel. Three out of conference games had an audience around seven million, while the rest were under five million.   CBS showed a 28% increase in viewership during conference season while Fox jumped a stunning 41% between the two parts of the regular season.

The lowest two Tier One out of conference games this year were Nebraska at Miami, with 1.88 million viewers and  Virginia at UCLA with 1.35 million.   Over all UCLA at Utah, a Pac 12 conference game, came in last for all Tier One games with only 845,000 viewers.  Completely anecdotal, but I find it interesting that UCLA, who resides within the second largest metro in the nation, not to mention the largest state, played in the two lowest games.  Contrary to popular belief, location does not bring audience.

Time of day, however, does.  Almost half of the US population resides within the Eastern Time zone.  Matter a fact, just under 80% of the US population resides within the Eastern and Central Time Zones, with only 14.1% of the population residing on the West Coast.

nmtier7Networks schedule their games to land in one of four major inventory slots in their programing, all tied to Eastern Time.   These are games at Noon, Mid Afternoon (2-5 pm, with 3:30pm being the most popular), Prime Time (7-8 pm Eastern), and Late Evening (start times after 9 pm Eastern).

When we look at Tier One games within these slots, Prime Time, as you’d imagine, garners the highest average viewership with 4.55 million per game.  This happens after a rather solid day, with both the Noon and Mid Afternoon slots averaging around four million viewers per game.  All of that college football love drops off considerably after Prime Time, however, as the East Coast starts to head to bed.   Games starting after 9 pm Eastern average only 2.16 million viewers or almost half of the earlier time slots.

I had a lot of mail questions this year asking why the Big 12 was starting to play so many games at 11 am central, e.g. the Noon inventory slot, instead of later in the day, like some of the old schedules, which were closer to 5 pm Central, e.g. 6 pm Eastern.  This is the answer.   The Noon Eastern inventory slot on nearly every major network is the second highest of the day and used to be owned by the Big Ten.  The Big 12 and their media partners are positioning the conference in a new direction.   Instead of aiming it at Texas, it is being aimed toward the Mid-Atlantic.  While it may take several years to build up a new following, targeting your games to over 100 million more people than prior is worth the short term changes.

nmtier6Another aspect that shows up in the numbers is that games are more valuable as the season progresses.   While some games stand out early, like Notre Dame at Clemson in Week Six this year, most of the major games occur in the final four weeks of the season.  Very few conference games occur in the first few weeks of the season, but those that do have very little relevance to conference championship races as the field has not solidified yet.  By the end of the season, as teams are jockeying for final position, be it first or third, interest in the games elevates by nearly two million viewers per game, as indicated by the red trend line in the graph above.

Time of day and place within the season, while having a big impact on audience, are not the primary reason people watch games, it just shows when they like to watch.   The most important indicator toward whether a game garners a large audience or not is the winning percentage of the teams playing the game.

nmtier8This doesn’t mean just one team’s record, it is the combined record of the two teams playing the game.  Below a combined winning ratio of 60-65%, the games hover below 3 million viewers.   However, head over that ratio and the audience explodes quickly.

The largest Tier One regular season conference game this year was LSU at Alabama, which had an audience of 11 million viewers in Prime Time and had a winning percentage of 88%.  Additionally, to give it a little more of a bonus, Alabama upset a top five LSU convincingly.  The second highest Tier One game, losing out by a mere 11,000 viewers, was Michigan State at Ohio State, in which the Spartans not only upset the then #1 Ohio State Buckeyes, but had a spread of 3 points and a combined winning percentage of 91%.

Within the top 25 Tier One games, there were only a couple anomalies to this winning percentage aspect.   Tennessee at Alabama, with a 65% win ratio, had nearly 7 million viewers Mid Afternoon on CBS.   Part of the reason had to do with the game being close until the very end with a mere 5 point spread on the final.   Additionally, both the Red River Shootout and Georgia at Tennessee pulled 4.9 million viewers with only having winning percentages of 57% and 58% respectively.   However, Texas upset Oklahoma and Tennessee upset Georgia by only a touchdown sending both conference races into spins.  (Note: Upsets do have an impact, if it is the right kind of upset.   Overall they add about a half million viewers if a Top 25 team is knocked off.   If that team is undefeated or in the Top Ten, then the numbers increase significantly.) .  The next highest game with a winning percentage in the 50%s was the 71st most popular Tier One game, which had less than three million viewers.

Outside the rare and early “Made for TV” match ups, the recipe for large Tier One audiences seems to be two conference title contenders, matched up late in the season, with a broadcast scheduled for during the day on Saturday or Prime Time.   Make the game close, have it be a long standing rivalry, or add in an upset or two and that large audience can even multiply further.


  1. What makes “little sense” about the AACs numbers is your inability to under stand how averages work. The per team numbers from the second table are directly affected by the number of games that are broadcast on tier 1 networks. Thus, with only 2 games the average goes way down. Given more games it would be much higher.

    • Thank you for the comment, that was actually my point in the article. The AAC looks good, in their four games. However, if you only take the top four from any conference they look fantastic. You need to look at the exposure per team to see the value. To your point on more games, that isn’t true. In any conference the more games you broadcast the lower the average becomes, not higher. The games the AAC would broadcast are already broadcast in Tier Two formats, which don’t support your conclusion, but I’ll get to that in the next article.

      Thanks again!

      • The more games the lower I agree with given these were top games for the AAC which is the only reason they were on Tier 1 but…and its a big but…the numbers wouldn’t be near as bad as the “average” in your table. UH/Memphis for instance would have hit a good number as would UH/Cincy, Memphis/Navy, Memphis/Temple, & Navy/Memphis. Sure games like USF/UCF wouldn’t have drawn flies but the avg would still be nowhere near the bottom and closer to the ACC/Pac12 numbers.

        The point I’m making is similar to what I tell my son about zero’s on his homework and how it affects his grade. Your averages when including all teams gives a zero for 10 of the AAC teams essentially.

        “Overall they only have 0.33 Tier One game per team, making their audience per team three million less than their straight line average.” This is a horrible point and a gross extortion of the data.

        • You’re confusing the two concepts. The first is a straight average, the second is the audience per team. If you played three games at 3 million you have a far larger audience than one game at 3 million. That’s where straight averages distort the picture greatly. The AAC had some good games this year, but, over all, the total audience the Power Five conferences pulled in vastly outnumbered the American’s.

          • To compare the AAC ‘per team’ average to the P5 conferences ‘per team’ average is completely invalid since the AAC only had 2 tier 1 games on while every other conference had many many more . Thus you can get an accurate avg for the P5’s while you don’t for the AAC. Its a simple matter of numbers monkey. How can you not get it?

            Your basically saying that if they added more AAC games to Tier 1 that zero people would watch. Implying that the AAC would actually get an avg of .17 number for in-conference games is ludicrous.

            Look at the conference game numbers. If the AAC had 10 other tier 1 conference games last year…then those 10 games would have only had to average a little more than than half of what the two AAC games that were on ABC for the AAC to have matched the Pac12 in per game average. You would be hard pressed to find anyone that thinks that the AAC couldn’t avg half of what it already actually did. Oh, and those numbers would be sailing by the ACC….not because of the AAC being better but because your ‘per game’ average once again skews due to using only 9 AAC games but dividing by 14 teams.

            In short, your ‘per team average’ is basically a useless number unless each conference has the same number of games. The only reason I can think that you’re using that is that you’re an SEC fan (hello…15 games divided by 14 teams)

          • Actually, you’re the one saying no one would watch, my point is if you look at total audience of Tier One games and divide it by the amount of teams in the conference you get an idea of the exposure each team receives.