Demystifying Sports Ratings

It occurs to me that there’s a massive amount of ignorance about how TV ratings actually work among those that pay attention to sports ratings, even among those that should know the most about them. I’ve put up a FAQ that I hope will aid people in reading my ratings posts, as I hope to spew out a whole bunch of them next week, but I want to clear up a few misconceptions here in hopes of elevating the discourse over sports ratings.

The level of ignorance is so bad that this sort of nonsense can spread almost unchecked across social media:

But Paulsen is himself part of the problem here: he regularly posts the overnight ratings from sports events when they come out and compares them to overnight ratings from past years, even though they’re next to useless and most people actually within the industry don’t even pay attention to them anymore. I seem to recall reading that he actually knows better, but still posts overnight ratings because networks – capitalizing on the general ignorance of how ratings actually work – will tout them regularly. But I don’t think that applies to SportsBusiness Daily, which posts overnight ratings for sports events on broadcast every Monday, as though anyone beyond its more ignorant clientele cares.

People who talk about general sports ratings often show a disappointing level of ignorance, but on this front they’re leaps and bounds ahead of the sports world. Overnight ratings, which only reflect viewership within the 56 “metered markets”, are ignored so much that TV Media Insights is pretty much the only general ratings site that regularly reports them, at least on broadcast. Most everyone else is willing to wait the few hours it takes for the fast national ratings to show up around 8 AM ET (which some sites confusingly label “overnight” ratings). Moreover, the fast national ratings aren’t always as accurate as some people would have you believe by referring to “final” ratings “according to Nielsen fast nationals”; there are almost always at least some adjustments from the fast nationals to the final ratings for broadcast primetime shows, and for sports and other live events on broadcast in primetime in particular the fast nationals are next to useless because they incorporate what would be on a given station on the West Coast at the scheduled time, so a sports event at 8 PM ET, which is 5 PM PT, would incorporate whatever aired on a West Coast station at 8 PM PT into the fast nationals.

Part of the reason no one pays any attention to overnight ratings is that the total viewership and household rating numbers that tend to be the most widely available, the latter of which is all that overnight ratings supply, are themselves pretty much useless for the purposes that actually matter – a beauty pageant, something to tout in a press release, and little more. Nielsen exists to provide a benchmark for networks to sell ad space, and networks in this day and age are in the business of selling demographics, not general viewers – especially the 18-49 demographic everyone knows is valuable but don’t generally grasp how valuable. TVbytheNumbers has been able, for a few years now, to predict the fate of (openly) scripted shows on broadcast television based solely on the 18-49 rating, without any reference to total viewership or household rating, and perhaps as a result it and The Futon Critic report only total viewers and 18-49 rating in their daily ratings posts, not household rating. Of course, different networks target different demographics based on what audiences they’re targeting, but what matters to the broadcast networks is particularly relevant here because broadcast networks at least nominally don’t target any audience in particular (and sports has to compete for space on the broadcast networks with pretty much any other kind of programming), so the hegemony of the 18-49 demographic is determined by the free market alone, and the boom in sports rights fees is precisely (in part) the result of sports’ ability to attract the 18-49 demographic like little else, the hegemony of which – as I explained in my Nexus of Television and Sports in Transition series – is in this day and age the result of the fact that 18-49-year-olds simply watch less television than anyone else. So when Paulsen says this…

…he’s implying the 6.8 isn’t the “real” ratings number for the NBA Finals, when – from the perspective of the actual decision-makers – it might be more “real” than the household rating he’s referring to. As if to underscore the point about the rarity of the 18-49 demographic, that household rating was a 10.3, meaning the 18-49 rating was maybe two-thirds of the household rating – and the NBA is known as a league that disproportionately attracts 18-49-year-olds compared to other properties. But the only sources that regularly report 18-49 ratings are the general ratings sites I referred to earlier, The Futon Critic and TVbytheNumbers. Anything else comes from network press releases. To my knowledge, no site that regularly talks specifically about sports ratings pays any attention to 18-49 ratings.

This also helps explain why people so often tend to overstate the importance of the broadcast/cable distinction, as though it were still the 90s. Yes, any given sports event will have a substantial drop-off when it moves from broadcast to cable, but teams, leagues, and networks have proven time and again since 2008 that this matters little to them, that the dropoff isn’t substantial enough to overcome the ability to collect subscription fees from cable customers. A na├»ve reading of the ratings for the Stanley Cup Final would look at the total viewer and household numbers – 4.777/3.0, 6.413/3.7, 2.893/1.7, 3.383/2.0, 6.021/3.7 – and conclude that the two games on NBCSN are suffering horribly and should move to broadcast, and the fact that they aren’t on broadcast like all the other big events (except the BCS, college football playoff, Final Four, Monday Night Football, most of the World Cup including quite possibly all the American matches…) reflects poorly on the NHL. But when you look at the 18-49 ratings – 1.90, 2.10, 1.16, 1.32, 2.31 – the dropoff, while still there, isn’t quite as severe, especially if you take Game 1 (which was neither a potential series-ender nor had a Triple Crown attempt in the Belmont Stakes as a lead-in) as the broadcast baseline, and it becomes easier to see why NBC and the NHL would take lower ratings for two games in exchange for keeping people tied to their cable subscription, and keeping NBCSN in demand for cable operators. For a variety of reasons, some obvious some not, the people that advertisers actually want to reach tend disproportionately to be cable subscribers; cord-cutting hasn’t yet caught on enough to change that calculus, and sports fans are disproportionately unlikely to cut the cord precisely because so many sports events are on cable now.

I’m going to try to come up with a formula to try and calculate what rating a sports event on cable would get if it aired on broadcast, but for a number of reasons comparing the popularity of sports events between broadcast and cable directly, or even from one cable network to the other, is in large measure a fool’s errand, and comparisons are best made within one network. (Even on broadcast, observe the trouble Fox has had getting people to watch nominally-marquee college football games.) Just moving from ESPN to ESPN2 results in a pretty substantial dropoff for all but the most can’t-miss sporting events, and NBCSN and Fox Sports 1 are lagging behind both ESPNs substantially despite not really being that far behind in distribution. I’ve observed a trend where, once someone starts watching something, they don’t always turn the set off until a good long while after the event is over; a really popular event like an NFL game can have ripple effects on a network’s ratings for hours afterward. Even middle-of-the-night re-airs on ESPN can beat just about anything on ESPN2 or any other network; NASCAR, college football, and the World Cup are the only things on ESPN2 that can regularly stand up to anything the powers that be decide to put on ESPN. (This becomes really obvious during college basketball season, when there’s no logical reason why games on ESPN should be so consistently far ahead of games on ESPN2, even when they’re both power-conference games with little discernible difference between them.)

TV ratings have become an increasingly watched scoreboard as the financial stakes in the sports TV business continue to ratchet up, but people seem to be unclear on how to read them or what their limitations are. I hope to increase my coverage of sports ratings at least back to the level they were at in mid-to-late 2013 in upcoming weeks (sans the Studio Show Scorecard), and I hope you’re able to recognize what the ratings actually say – and what they don’t – going in.

Leave a Comment