The CBS Comedy Fix
If you saw the first version of the R% that was released after finals on Thursday, you saw a rather ridiculously bleak picture for the CBS comedy lineup, that had Life in Pieces below 50% and both The Great Indoors and Mom as huge underdogs. It hadn't been quite as obvious based on the limited Monday comedy data, but this made it abundantly clear that the formula was fundamentally being too hard on the CBS comedies.
A big part of this was the fact that the formula generates its targets using not just the full network average, but a significant portion is also based on the category average (in other words, the network's average with just comedies or just dramas). I found over the summer that it was very tough to use the limited early data to predict what the category average would look like at the end of the season. So for the fall, the formula used a ratio based on what these averages looked like at the end of last season. If I were just using the early weeks of Big Bang and Kevin Can Wait and calling that the network comedy average, it would've been even worse.
But even what I did use has turned out to be a pretty bad assumption... mostly because of CBS' major comedy expansion. Last year, the network aired somewhere from one to three comedy hours for most of the season, and had very few actual bubble-worthy comedies; pretty much just Life in Pieces (and even it was safer at the end of the season), a short order of The Odd Couple and the very few eps of Angel from Hell. It resulted in the CBS comedy average being about 30% ahead of the network average as a whole, and that's the assumption we took into this season. This year, it looks like the network will have much more marginal material, including both 8:30 newbies and both 9:30 returnees. Even Mom had a pretty modest start in True, though I'd expect it to hold up better post-premiere as it did last year.
So what I've done is take out CBS' category adjustment, and will probably not bring it back till at least the winter, which is when the formula is supposed to start using actual data rather than last season's ratios. I hate to have to do something like this, but it was so egregious that I really didn't want to spend the next few months continually apologizing for it. This has led to a double-digit R% increase for most of these shows, and as of week 6, the department now looks like this:
|The Big Bang Theory||100%||3.02||3.01||1.09||2.45||3.58||6|
|Kevin Can Wait||93%||1.61||1.49||1.09||1.18||1.80||6|
|2 Broke Girls||77%||1.29||1.29||1.09||0.97||1.61||3|
|Life in Pieces||59%||1.31||1.16||1.09||0.81||1.51||1|
|The Odd Couple||44%||1.11||1.06||1.09||0.80||1.31||2|
|Man with a Plan||43%||1.44||1.04||1.09||0.66||1.42||1|
|The Great Indoors||40%||1.41||1.02||1.09||0.64||1.39||1|
Now, on to the gainers/losers for week 6.
Biggest R% Gainers:
|The Real O'Neals||1.0||+0.1||1.03||+0.09||70%||+23%|
|Jane the Virgin||0.4||0.0||0.36||+0.03||90%||+17%|
|The Odd Couple||1.1||0.0||1.11||+0.03||44%||+16%|
|NCIS: New Orleans||1.3||+0.1||1.38||+0.17||81%||+15%|
|2 Broke Girls||1.4||-0.1||1.29||+0.01||77%||+6%|
These week-to-week swings are assuming that the CBS comedy fix was applied for all points for the whole season... I did this for this table only just because it would make the growth for The Odd Couple and 2 Broke Girls look even more crazy. It does not use the huge target change that you'll see if you look at their SpotVault pages. Still, even removing that factor, this was a good week for both shows, which were basically steady in True as The Big Bang Theory left the night.
It was also a good week for The Real O'Neals, which gained a tenth as Fresh Off the Boat lost one, and suddenly looked a lot more viable than it did in week two. As I said last week, the R% in the first three weeks are just single-week snapshots, and thus prone to overreaction. Based on this snapshot, we have O'Neals at a 70% and Fresh at just 76%. But I would expect that margin to widen going forward; this is just based on week three and the first two weeks suggested a wider gap between the two. But if O'Neals can stay this tight, it could get interesting.
It may not be the R% winner of the week, but I'd say the winner of the week by a rather wide margin has got to be Superstore, which raged back from the brink with a 1.3 demo even as The Big Bang Theory arrived. A surge of that size in that situation seems kinda ridiculous, but it probably means the show really underachieved last week and probably overachieved this week. The bottom line is the 1.0 doesn't seem to be the norm for this show, or for The Good Place, which is good news. It's now got to survive just two more November airings before a football break, and the landscape should be more favorable when it returns.
Biggest R% Decliners:
|The Vampire Diaries||0.3||-0.1||0.40||-0.11||69%||-16%|
|Once Upon a Time||0.9||-0.1||0.95||+0.00||64%||-12%|
|Fresh Off the Boat||1.2||-0.1||1.06||-0.13||76%||-7%|
|NCIS: Los Angeles||1.2||-0.5||1.13||-0.10||66%||-7%|
Most shows on the list this week went up against the World Series. There were a few cases where it seemed like the True adjustment was not doing well at all with the added difficulty of the Series... for example, Fresh Off the Boat and The Vampire Diaries each dropped by over a tenth in True despite losing just a tenth in raw. (In other words, it was basically ignoring the Series completely.) Things like the CBS Friday shows and Designated Survivor were not as bad, nicked by less in True than in raw, but still perhaps a bit too much. And then there were shows like Quantico and Once Upon a Time, where I think the True formula was pretty reasonable in keeping them steady when they lost 0.1 in raw. (Their continued R% drops are more about what happened the previous week, and the higher points from early in the season dropping out of the average.)
My gut says that World Series competition (at least the early games) probably makes the slot close to a tenth more difficult on average, so maybe I went a bit overboard in depressing how much the formula counts sports competition. Whatever your opinion, fortunately it's just one week (or two for some Tuesday/maybe Wednesday shows), and there won't be big sports interruptions on weeknights beyond this point. So whatever error there might be will be out of the R% picture pretty soon.