EDIT (8/29/13): For the 2013 updates to the formula, please see The 2013 True Tweaks post. At the moment, I'm NOT adding the 2013 updates to the formula listed below.
Let's put together all of the adjustments I have laid out over the past four days. I don't 100% guarantee that this will be the final formula when the blog returns to full-strength mode in about 2.5 short weeks, but I am so worn out on this that it seems pretty likely. First, one last tweak that (at least for this year's version) doesn't really seem to fit under the other three banners:
Fall Hype
Last year, I said that for all the work I have done on looking at timeslot numbers, there is still for some reason a considerable inflation on ratings at the beginning of the season. It's basically inexplicable by any combination of overall viewing or competition. People just plain have more of a propensity toward broadcast programming in the early season. I don't necessarily like that I can't explain it because I fear that it's just a "me problem." But it exists.
The thing about this is that it's not about shows declining faster than the rate of overall viewing; if that were the case, I would be more convinced it was a "me problem" because I'd think I just wasn't capturing the rate of decline correctly. But during the period of viewing decline (the last three months or so of the season), the shows decline about how I expect them to. The real problem is in the middle of the season (January/February), when viewing is going up to its peak levels yet most shows' ratings are staying the same at best. That suggests that there's just some "hype" factor in the early season that inflates the ratings in that period. (For a clearer illustration of this process, I recommend last year's much more detailed "Ratings Across the Season" post. I did something pretty similar this year.) Considering the amount of money that goes into the broadcast networks' fall lineup promotion, I suppose I shouldn't be so bothered by it.
This time, I'm pushing the "fall hype" factor a little later in the season - to essentially the end of November.
This number is a lot bigger than last year. I'm concerned that it'll be too much, since I've gotten a sense that this spring was simply a lot worse than most springs. Maybe next summer I won't have to totally overhaul the formula and can look at this from more of a multiple-season perspective. Another concern: the two-season shows (Survivor, The Amazing Race, The Biggest Loser, Dancing with the Stars) are somewhere from 2% to 10% stronger in the spring than the fall after this adjustment. That leads me to believe perhaps it's less a calendar thing and more about how show seasons develop. (Then again, with no fall hype at all they're 3% to 7% weaker, so there's at least something to it.) But hey, if all the True2 scores shoot up in the second half of the season, I have a feeling nobody (except for me) will be particularly distressed by that! It's fairly clear that the fall hype number from last year was way too small, and maybe I have to miss in both directions before settling on the right number.
The Full Formula
I'm using some color coding just to try to make it a little clearer how these legions of numbers connect. Still, this stuff will be incomprehensible without reading this week's other posts linked above and below (and likely still even then!). Reading the generally much more detailed 2011 posts (despite their flaws) would probably help too. Since this whole thing was just five posts this year, I don't think it's as necessary to write a detailed summary like I did last year.
Fall Hype = 92.5% from beginning of season through December 2, 107.5% from December 3 through end of season
True PUT (TPUT) = PUT * Methodology Adjustment
Tendency PUT = TPUT - (0.4 * bc)
Viewing True2 = (A18-49 * 30.00 / Tendency PUT)
Viewing Adjustment = Viewing True2 - A18-49
Lead-in Impact = 1/(4 + length in half-hours)
Lead-in Baseline = 5.5% * TPUT (* 0.25 if CW) - Lead-in last half hour ( * 0.75 if sports)
Lead-in Adjustment: Impact * Baseline
Competition Impact = 3.75%
Competition = Sum of ratings for all OTHER broadcast shows + 10:00 Hour Adjustment + Impact * (A18-49 - 5.5% * TPUT (* 0.25 if CW))
bc vs. Predicted (bcvP) = (0.9% * TPUT^2) - bc
Normal Tendency PUT = 28.3 at 8:00, 29.5 at 8:30, 30.6 at 9:00, 31.3 at 9:30, 31.5 at 10:00, 29.8 at 10:30
PUT vs. Predicted (PvP) = (Normal Tendency PUT + 0.4 * bc) - TPUT
Holiday Number = bcvP + PvP (if sports, + 0.5 * A18-49)
Expected Competition = 9.82 (times 4.00/3.25 if CW)
Competition Adjustment = Impact * (Competition + Holiday Number - Expected Competition)
True2 = (A18-49 + Viewing Adjustment + Competition Adjustment + Lead-in Adjustment) * Fall Hype
So there you have it. Perfect? Not at all. Better than last year? I happen to think it's significantly better, but we'll see. I will be using True2 in the fall preview starting on Sunday, and I also think I'm going to try to churn out a quick set of True Power Rankings if I can get ahead of the 8-ball on the fall preview. I think I need something like that to illustrate the differences between True Strength and True2 in a more practical setting. Again, no guarantees I won't make a little tweak in the coming weeks, but I'm pretty sure this is pretty close.
True2 Week: Viewing | Lead-ins | Competition pt1 | Competition pt2 | Wrap-Up
No comments:
Post a Comment