The Only Wheel in Town

TV's toughest running drama of suspense and conflict draws a small but influential audience. At Lindy's in Manhattan, bookies take bets on the outcome of each month's cliff hanger, The Ratings. Across town on Madison Avenue, admen ante up with million-dollar chips, and on the same fateful turn, TV shows stake their survival, performers their jobs and networks their reputations. Every eye in TV is firmly fixed on the numbers that do what the batting average does for baseball, the big board for Wall Street and Debrett's for the peerage.

Last week, while executives tensed like the knots in their pine paneling, the latest chapter flashed through the industry. Nielsen's top ten network TV shows for early December were:

1) I Love Lucy (CBS) 46.3*

2) G.E. Theater (CBS) 41.2

3) Ed Sullivan Show (CBS) 38.1

4) Alfred Hitchcock Presents (CBS) 37.5

5) $64,000 Question (CBS) 36.3

6) December Bride (CBS) 35.3

7) Arthur Godfrey Scouts (CBS) 33.6

8) Perry Como Show (NBC) 33.4

9) Jack Benny Show (CBS) 32.9

10) Gunsmoke (CBS) 32.9

This did not mean that these were the ten best shows on TV; indeed, they were by no means a roster of merit. It simply meant that, by one of the systems in TV's way of counting the house, they drew the biggest audiences. By this alone, rating systems such as the Nielsen work the most ruthless tyranny in a nervous industry that looks to its audience for leadership instead of providing its own. As big-time TV enters its second decade, the ratings are more powerful, feared, hated--and needed--than ever before. The sponsor has always demanded omens that his money is well spent. With the money going ever faster (a weekly half-hour show can now cost a sponsor close to $3,000,000 for a 39-week season), he demands swifter omens of how his investment is faring. In 1956 sponsors dropped some 50 network programs because the ratings fell so low that, by Madison Avenue's sacrosanct formula of "cost-per-thousand," the price of reaching a given number of viewers rose correspondingly too high.

"I Die a Thousand Deaths." Almost as much as sponsor pressure, the telecasters can blame themselves for playing Frankenstein to the rating monster. It was the networks and the performers who began using the advertiser's yardstick to beat the drums of publicity, plugging ratings from whichever system made them look best and playing up rating feuds, e.g., CBS's Ed Sullivan v. NBC's Steve Allen on Sunday at 8 p.m. They made the rating seem even more potent than it really is--and believed the illusion themselves. Since NBC began trailing in the ratings, it has sensibly (if belatedly) dropped the habit of regularly publicizing the figures, but the habit clings where the ratings are higher.

Even those on top of the rating heap suffer--from fear of falling. Long a frontrunner, Sullivan has written: "At 9:30 a.m., I hear my Trendex rating for the show the night before. I tell you I die a thousand deaths in the thirty seconds between saying hello and hearing my rating . . . Your rating is the entire expression of your work." Sullivan's whimper is mild alongside the outrage and scorn that ratings evoke from their real victims. Walter Winchell's anguished demand for a congressional investigation of the rating services (TIME, Dec. 3) was only the latest lament in a line that includes George Jessel's bitter charge that he was "ruined" by ratings and Writer Goodman Ace's sardonic crack: "Polls are fascinating. They are read by everyone, from the farmer in the field all the way up to Tom Dewey, President of the U.S."

How Do Ratings Rate? Almost everyone finds fault with how the raters get their ratings. Indeed, all the services are vulnerable in varying degrees and none meets all standards set by the Advertising Research Foundation in 1954. But the most common complaint against rating systems is groundless. Independent statistical experts dismiss the popular notion that a mere 1,000 homes are inadequate as a sample of millions. For its own monthly national surveys, the U.S. Census Bureau uses a sample of 40,000. But the bureau has convinced itself that it is possible, with a small, predictable statistical error, to draw national estimates from a sample of only 400--if the 400 are properly chosen as a U.S. cross section. That still leaves the big question of how good a sample the rating systems take.

Another widespread complaint is that the different services often get contradictory results, sometimes agree on no more than three shows among the top ten. Example: Nielsen rated Fireside Theater as falling from 27.5 to 23 from one month to the next, while Trendex showed it rising from 18.5 to 27. Such confusion nettled Chrysler's Sales Director Richard Forbes, who spends an estimated $17 million a year on TV shows, into demanding that the industry develop a single comprehensive rating system. NBC's President Robert W. Sarnoff agrees that this would be an ideal solution. Without really satisfying these critics, the raters can usually explain away the contradictions by pointing out that, for the most part, they measure different phenomena in different areas. Another reason for the clashing results: the leading four national services use different methods.

P: The A. C. Nielsen Co., the most trusted of the services, uses electronic meters hooked to TV sets in 900 U.S. homes. Called an Audimeter, the 20-lb. gadget costs $300 to make, records hieroglyphics on film giving a minute-by-minute account of when and where the set is tuned. In most cases, says the service's meticulous founder and president, Arthur Nielsen, 59-the family does not even know what the Audimeter does, regards it vaguely as a gadget designed to improve TV quality. Every two weeks, the family gets a new film cartridge that will yield two 25-c- pieces when it clicks into, place. The old cartridge is mailed to Chicago, where analysts read the code signs and put the results into IBM machines. Together with what Nielsen knows about his 900 families. they produce not only a measurement of audience size, but masses of data that enable Nielsen to report, for example, how many viewers of a particular show own Cadillacs. Nielsen prefers to call his product "audience research" rather than a mere rating; it is the only service that checks tuning all month long, traces audience ebb and flow during a single show, uses the same sample all the time (touching it up to match population changes). Nielsen's sample is the costliest to produce. His service is also the most expensive, costing advertising agencies annual fees ranging between $15,000 and $90,000 and networks up to $100,000. It probably grosses more than all its competitors combined.

P: Trendex Inc. uses the telephone method pioneered for radio ratings by the late C. E. Hooper. To spot-check viewing while the TV set is on, it farms out the phoning to 400 women who call from their homes in 15 major cities where all three networks are on the air. A rating for a half-hour show is based on an average of 1,000 completed calls. The telephoners use a set list of questions and a pattern for picking numbers at random out of the phone book. Supervisors telephone to keep tabs on them. If telephoners try to fake returns, Trendex can spot it in plotting a statistical curve based on their reports. Trendex is neither a national survey, a true cross section nor even a meaningful gauge of audience size--nor does it pretend to be. Instead it offers an index of relative popularity in the 15 big cities. Its main appeal is speed. It is the only service that can operate normally overnight.

P: American Research Bureau Inc. uses the viewer-diary method, placing its diaries with a different 2,200 families each month. Every batch of families is carefully chosen to make up a national sample. ARB representatives usually visit the homes to explain how to fill out the daily viewing record and later to check that the diary is being kept properly during the week's viewing. Against the criticism that a self-kept record of family viewing is no more reliable than an expense account. ARB recently ran an experiment in eight cities. The company used the telephone-while-viewing technique to check on the diary results for the same period. The results: "virtually identical."

P: The Pulse, Inc. uses the technique of the doorbell-ringing personal interview. Its interviewers, all married women (men might incite neighborhood gossip), visit a cross section of homes in 64 markets. When they establish that a family has watched TV that day or the day before, they jog the viewer's memory by displaying a program schedule for the period and asking what was seen before or after normal household activity, e.g., shopping, dishwashing, linked to particular hours of the day. Every tenth interview is checked by a letter to the family from Pulse. For the average half-hour nighttime network program, Pulse gets its rating from some 7,000 interviews, puts together a national rating from its local surveys. Of the raters Pulse claims the largest sampling.

The most eloquent, hard-hitting critics of the ratings are the services themselves--when speaking of their competitors. Nielsen, for example, argues that human error, bias and forgetfulness work against the accuracy of the others' methods. He says also that their samples are usually unreliable. In special surveys, he has tested the accuracy of the other methods by the yardstick of his own and says that all three fall wide of the mark. Nielsen's rivals--who also rap each other's techniques--seize on the fact that Nielsen's national system measures the tuning of sets, not the number of viewers, and does not account for the chance that the set might be playing to unheeding householders or even to an empty living room. ARB insists that this is a big factor; Nielsen insists that it is negligible.

What Ratings Can't Do. The ideal system, according to CBS President Frank Stanton, would be "to get a Nielsen rating the morning after." The industry now waits almost a month for the reports. Nielsen has devised an Audimeter that can transmit readings instantly to Nielsen offices by leased lines and, at the request of TV brass, is preparing an estimate of what his service would cost on an instantaneous or overnight basis. He holds hope that it may be economically feasible.

Even if ratings were 100% accurate--which they can never be--there are some things that, contrary to widespread fallacy, they can never do. They are no true test of a performer's popularity, because they are influenced by too many other factors, e.g., the programs that come before and after, the ones on competing stations, the weather, the time of day--even the season. Red Skelton's history on TV shows that a performer may be dropped because he is overwhelmed by his competition, later become a hit in another time slot.

Ratings are also worthless as a guide to what the sponsor is really after: sales. I Lone Lucy was riding high when it was dropped by Philip Morris; it was evidently not paying off at the cigarette counter. Though virtually all automobile sales fell in 1956, high-rated Ed Sullivan's sponsor Mercury suffered a bigger decline than most. (Adman's rebuttal: Who knows how badly sales might have fared without Sullivan?) Arthur Godfrey argues that sponsors ought to judge shows by how the product is moving. But, say admen, most products are affected by too many variables for TV to get the credit.

However they may be improved, the rating systems and their tyranny over TV hold no hope for the viewer who believes that the verdict of democracy is fine for government but folly for television. The viewer is both the unwitting culprit and the ultimate victim of the tyranny. The ratings discourage worthy programs that might make a deeply favorable--but un-measurable--impression (including sales impact), but that do not attract measurably large numbers. They inspire imitations of high-rated programs, touching off cycles of the second-rate.

Thus, in a mediocre season that owes much of its mediocrity to them, ratings remain a controversial, confusing means of doing imperfectly only part of what the TV advertiser needs done--and spreading unhappiness in the process. But until something better comes along, TV finds them indispensable. "I'm not bitter and I'm sure they're as honest as they can be," said one unemployed comic last week. "But it's like the story of the gambler who played the roulette wheel in this little town and kept losing all night, until a fellow came up and said: 'Look, bud, that wheel is fixed.' 'I know,' said the gambler, 'but it's the only wheel in town.' "

* Each figure is the percentage of the nation's TV homes within reach of the program that are calculated to be tuned in during an average minute. Lucy's 46.3, for example, suggests that 17,040,000 sets were tuned in for the show of the night of Dec. 3.

This file is automatically generated by a robot program, so reader's discretion is required.