Here’s one thing I would like to say before the U.S. networks announce their fall schedules: reruns. Avoid them whenever possible. Except in the summer, and maybe not even then. And especially avoid them near the end of the season, in April and May.
Looking at the viewership numbers for a lot of shows this season, there seems to be a clear overall pattern – when a show takes some time off from airing new episodes, and it’s not for an occasion where nobody shows new episodes, then some viewers don’t come back when the new episodes do. Take Glee, which managed to avoid long stretches of repeats last season because of the split-season format it used. This season, it managed to hold steady for most of the season, even surviving a January hiatus with the help of a major launching pad (the post-Super Bowl episode) that gave it heavy promotion. But then it took some more time off in March and April: after the March 15 episode, it didn’t air another new episode until April 19. And since it’s come back, it’s lost 1-2 million viewers. You can find a similar dip in viewership for many shows around the same time – for example, Community got 4.4 million viewers for its March 24 episode, then had some reruns and pre-emptions, and came back a few weeks later (April 14) with a million viewers gone. It was already airing against Idol, which didn’t take that many viewers away from it – but reruns do what Idol cannot. NCIS took a break between April 12 and May 3, and two million viewers didn’t come back when it did.
It’s probably unwise for me to dwell too much on individual examples, since week-to-week or month-to-month declines have other explanations besides the repeats, and sometimes a show can buck the trend and go up if there are other factors working for it. But it’s a pattern, and it’s a pattern that makes sense. We hear all the time that viewers have more options than ever before. We also know that TV viewers are creatures of habit. If a show doesn’t provide a new episode for a couple of weeks, it may not turn us off the show, but it can break the habit – we may discover another show on cable, or we may just start recording it. The danger of finding another show during repeat weeks was less pronounced during the three-channel era because it was less likely that we could find something to give us exactly the same kind of entertainment at exactly the same time. It’s still not the likeliest thing in the world, but it’s more possible.
Networks realized some time ago that serialized shows don’t repeat well and that they needed to find ways to keep a steady flow of episodes: that’s one of the reasons for Fox’s famous trick of not starting 24 until January. But I wonder if we’re not approaching a point where repeats will hurt any show: even if it’s one that gets pretty good ratings for the repeats, it can have a bad, habit-breaking effect on viewer loyalty. It can even take down a whole night: Two and a Half Men reruns get better ratings than many new shows, but since it’s been in reruns, all of CBS’s Monday night shows have withered in the ratings, as the viewers’ attachment to that whole night – the knowledge that there’ll be four hours of new comedy followed by some tropical action – is reduced.
They’ve been able to get away with repeats in the past because advertisers mostly care about sweeps periods, so the idea is to save the steady flow of episodes for those all-important months. But so many things combine to drive people away as the season reaches its end (after Daylight Savings Time it’s especially rough, because people go outside more), and with that and the increased competition, I feel like we may see shows collapse by the time the May sweeps period rolls around. Networks could once prevent this with stunt casting and promotion, but those just don’t do the trick as much as they used to either. Promotion can make a huge difference when the network goes all-out and puts in all its resources – at the beginning of the season, or the beginning of an extended period like the second half of Glee‘s first season. Trying to promote May sweeps, though? I don’t see how it can help enough to make up for the bad effects of repeats in March and April.
This is the moment when I should be calling for networks to cut down the seasons and show 13-episode seasons like cable networks do. I can’t bring myself to do that. The short-season model does work for cable: you promote the hell out of the new season, air it straight through without repeats (well, I mean, except for the “encore” showings on the same network) and repeat again next year. But I think many broadcast shows have too few episodes, not too many, and the ability to produce 22 or more episodes a year – leaving room for more different types of episodes and a more slow-building emotional engagement with the show – is something broadcast would be foolish to give up. In a world with many choices you don’t give up one of the things that makes you stand out.
What I do wish is that networks would just cut May sweeps out of picture for many of their shows. It’s been done before, and it was done partly because of the switch to seasons of 22 to 24 episodes. If you look at episode guides for shows in the ’50s and ’60s, when larger episode orders were common, the schedule was similar to today’s – from September, or October at the latest, through May or possibly even June. Look at the list of titles and air dates for Gunsmoke, the never-ending show.
And then the episode orders were reduced due to the increased expense of making TV episodes. But as Gunsmoke was cut back to 25 episodes, The length of its seasons were also cut back. By the late ’60s and much of the ’70s, television series would end the season in March, not may. This allowed a 24-episode season to run from September to March with few repeats. The September-to-May season started to come back in the late ’70s, for reasons I’m not entirely sure of, but it was a slow thing coming: when J.R. was shot on Dallas, it was on March 21, 1980, and that was the end of a 25-episode season.
I get why it is better to have a successful show run in all the regular-season sweeps periods, and why the programming a network can muster for May will not be of the same calibre. But I think broadcast networks may ultimately have to re-evaluate what they’re doing: they are using the eight-month schedule from the ’50s with the episode orders that were created for a six-month schedule. By doing so, they may be diluting the appeal of the shows and one of the most comforting things about broadcast TV: knowing you can jump in any week and find a new adventure. Something’s got to give, and it might be the schedule, it might be the number of episodes – or it could just be broadcast TV itself.