I'll present Power Ratings in two parts--one part today and the final part next Monday. By splitting the topic into two parts, I can provide a bit more detail for those who are interested.
What's a Power Rating? And how is it different from a Nielsen rating?
For the uninitiated, a Nielsen rating is the percentage of TV households tuned to a particular program. Nielsen Media Research estimated that there were ~110 million TV households in the United States during the 2004-2005 TV season. If 11.0 million TV households watched a show last year, then that show received a 10.0 rating since 11 million is 10.0% of 110 million.
A Nielsen rating just tells you how many TV households watched. That rating tells you nothing about a show's power. Nothing about whether that show can:
1. entice a household to watch a network it normally doesn't watch.
2. increase audience size on a night when people usually keep the TV off.
3. attract more households than the shows that precede and/or follow it.
The essence of the Power Rating is simple. It measures a show's power to change viewing habits (and the show's importance to the network). Not all 10.0 ratings are created equal. I'll use Nielsen Ratings to calculate values for the qualities listed above.
Today I'll briefly discuss how I quantify a show's ability to entice households to watch a network they normally don't watch. I'll finish with two lists of results. The numbers in these lists
are not the Power Ratings for these shows, but the numbers are a good start.
You really want to know "How large was the audience for one show relative to all other shows on the same network?" To do this comparison, you have to correct for the week the show aired. A 10.0 rating in the summer is very different than 10.0 rating during Fall sweeps month. If you want to read the mathematical description, you should expand the section below.
(+/-)Click here to read the details
For every week of Nielsen Ratings, I group the shows by network and then calculate an average and standard deviation for household ratings for that network for that week. I then take a show's rating, subtract the average rating for that show's network for that week, and then divide by the standard deviation for that show's network for that week. The result tells you how different a shows rating was relative to the ratings of other shows on that network for that week. Call this result Value1.
Now do this same calculation to find Value1 for every airing of every series. Group the results by series over the time period you want and calculate the average value of Value1 for each series for that time period. And that's it.
A positive number means the show had a larger than average audience for that network. A negative number means the show a smaller than average audience for that network. The larger the number, the larger the difference between the show's average value and the network's average value.
From the start of the 2004 TV season through Aug 7, 2005, which series had the largest rating relative to other shows on the same network?
Here are the top ten series:
2.24 ''American Idol''
2.14 ''Dancing with the Stars''
2.06 ''WWE Smackdown''
1.89 ''Monday Night Football''
1.85 ''So You Think You Can Dance''
1.58 ''CSI''
1.46 ''Brat Camp''
1.32 ''Desperate Housewives''
1.25 ''7th Heaven''
1.19 ''ER''
Who woulda thunk that ''Dancing with the Stars'' dominated ABC's summer schedule almost as much as ''American Idol'' dominated Fox's schedule earlier this year. And that number for ''Dancing with the Stars'' even includes a few low-rated repeat eps! I now understand, but don't approve, why ''Skating with Celebrities'' is being filmed.
Here are the bottom ten series:
-1.49 ''SI Swimsuit Model Search''
-1.48 ''The Will''
-1.35 ''BMOC: Big Man on Campus''
-1.30 ''Last Comic Standing''
-1.29 ''life as we know it''
-1.27 ''Wickedly Perfect''
-1.26 ''Complex Malibu''
-1.17 ''The Mountain''
-1.11 ''The Benefactor''
-1.07 ''The Cut'' (not including the last four airings)
No surprises there. It's interesting to note that ''Scrubs'' had the lowest value (-0.92) for a returning series. It's also interesting to note that no UPN series are listed--no single series flopped badly compared to their other shows.
