A weekend full of tinkering has left me with a number that's... maybe a little bit better than the one I had when I wrote last week's post! It's still not quite where I'd like it, but it's definitely better than what I had last year. So I'm gonna go ahead and start into these posts. I'll hold the competition portion for later in the week (after viewing/lead-ins) since that's where my efforts remain concentrated, just to give myself a bit more time to tinker around and hope for a real breakthrough. In the event that I actually have some hope I can make it better but am not there yet, I may hold off on the last few posts, but I'm getting to a point where I'm not sure I actually have the mathematical capability to take it much further.
I spoiled the big change in the viewing portion of the number last time, so this post will be a relative quickie to get us started.
"True" PUT
It's not like anyone else in the history of the world has tried to estimate viewing levels by adding up the ratings and shares in a timeslot. So most likely nobody will check anyway. But I still want to note that I'm not using those exact numbers, because the old way Nielsen defined "viewing level" was in my opinion a better gauge of a show's tendency to get viewed. (Rather than re-explain the entire thing I will just link to last year's post.) Since all Nielsen viewing levels are now New Methodology, my only conceivable way of testing last year's adjustment is to look at the year-to-year trends up through March 18 (at which point Nielsen made the methodology change in 2011). Going from the beginning of the regular season through March 18 with the same percentages as last year, my "True" PUT was down 0.5% year-to-year (whereas using plain old PUT would've been up about 6%). Basically, my thinking is that -0.5% is probably not enough to warrant seriously investigating a change, since it's very possible viewing was just down 0.5% in reality!
I am a bit concerned when I do the differences by half-hour (it's generally down a little more toward the end of the evening and down less or even up toward the beginning), so if I have some time later on I may add those small half-hour differences to the ones I already have. I don't think it'll make much of a difference either way since it's never more than a couple percent.
Tendency PUT
As I said last time, I'm centering the viewing portion of the True2 formula around what I call "Tendency PUT," which is essentially an attempt to strip the influence of heavy competition out of overall viewing levels. How many people would be watching TV if nobody were watching broadcast? My best stab at this is the "True" PUT above minus 40% of the combined broadcast ratings (bc). I explained it last time, but let me provide just one simple illustration.
This is just a look at one of the "swingiest" periods of the season, five Sundays in late January and early- to mid-February. From left to right, there's a relatively event-free Sunday (February 19), the not-that-high-rated Pro Bowl (January 29), the Grammys (February 12), the AFC Championship game (January 22) and the Super Bowl (February 5). When I chart the 8:00 thru 11:00 broadcast ratings vs. the PUT, you can see that they line up very well on those five very different days. The multiplier of x on the trendline is very close to the 0.4 number I use. And the constant added on (34.452) is approximately what I would consider the average Tendency PUT of that period.
So the general idea is that about the same number of A18-49ers would "tend" to watch TV on those evenings. The fact that these big events bring all these extra people is not something for which the other shows should be penalized in the viewing portion, and that's what Tendency PUT addresses.
The PUT Adjustment
To get a PUT-adjusted True Strength, I use the ratio of the Tendency PUT to the average regular season Tendency PUT for the full season, an even 30.00.
So here are the steps of the PUT part of the equation:
True PUT = PUT * Methodology Adjustment
Tendency PUT = TPUT - (0.4 * bc)
Viewing True2 = (A18-49 * 30.00 / Tendency PUT)
Viewing Adjustment = Viewing True2 - A18-49
"bc" itself has to be adjusted somewhat (to account for the 10:00 hour where Fox is not nationally rated), but I'll get into that in the competition portion.
Tomorrow, the lead-in adjustment.
True2 Week: Viewing | Lead-ins | Competition pt1 | Competition pt2 | Wrap-Up
No comments:
Post a Comment