Network News

X My Profile
View More Activity
E-mail Michael  |  On Facebook: Comic Riffs  |  On Twitter: Comic Riffs  |  RSS Feeds RSS Feed
Posted at 11:00 AM ET, 02/ 6/2009

Are Too Many Newspaper Comic Polls a Sham?

By Michael Cavna

It's the ugly little secret that, within the newspaper comics industry and among avid comics followers, is nobody's secret at all:

Namely, that the frequently used Newspaper Comics Reader Poll might be long-standing, but it is hardly upstanding. It has walked on the wrong side of the statistical tracks far too often, stirring so much scorn, skepticism and controversy that it might as well wear a scarlet "A" -- for "adulterated accuracy."

And yet, the much-maligned, oft-distrusted comics poll -- be it an online poll, a phone-in vote or a write-in survey -- keeps being allowed into polite society. Perhaps because someone, or rather something, is always needed around to do some of the dirty work.

The reason I write this is because in recent weeks, Comic Riffs ran an online comics "popularity" poll that had not even the remotest pretense of being scientific. Still, it was somewhat surprising to what lengths that apparent ballot-stuffing subverted how the poll trended for days. Almost overnight, the poll climbed quickly and conspicuously to nearly a thousand votes, as several back-of-the-pack comics surged out of nowhere.

As I wrote in today's earlier blogpost: This wasn't a vote -- this was a movement.

In regards to Comic Riffs, all this is terribly small potatoes. But the dime-store deceit drew the attention of some editors, cartoonists -- and eagle-eyed commenters to this blog (including "elyrest") -- who quickly contacted me to say: Welcome to a perpetual scourge of our industry.

See, the plot twist is, the comics poll still has a mighty place in the marketplace -- and polls, especially because they are conducted online more than ever, are much more easily gamed, stampeded, rigged and sandbagged in recent years, some artists and editors say.

"Cheating in polls seems to be more rampant than ever," says Stephan Pastis, creator of the syndicated comic strip "Pearls Before Swine." "And that cheating appears to have pervaded not only the voting, but in the comments newspapers solicit about comic strips on their site. This would all be 'fun and games' but for the fact that many editors actually rely on these polls, and therefore, it determines our living."

Amy Lago, comics editor at Washington Post Writers Group, echoes that sentiment. "Our sales people try to discourage editors from comic surveys," she says. "It's simply too easy to cheat. And it will usually NOT give them a representative sample of their readership -- or, more important, the readership they'd like to get." (It's worth noting that Lago encourages the "guest comic" concept over polling.)

John Glynn, who is Lago's counterpart at Universal Press Syndicate, says his syndicate also discourages reliance on polls. "Almost across the board, we recommend against newspapers using polls as THE determining factor in deciding a strip's fate," he says. "Polls generally favor older comics and to our mind they're susceptible to manipulation. ... And it's not too difficult how to figure out how to vote multiple times on an online or write-in poll."

(Glynn enumerates a handful of flaws with online comics polling, including the face that "cookies can be deleted"; that people can "figure out a ZIP code in your market"; and that cartoonists involved in the poll can spearhead their own campaign to boost voting.)

The host of factors at issue certainly includes methodology and whether a poll is "scientific" or not (some editors say paying for a polling service or scientific market research is not something they can readily afford in the current economic climate). Also at issue: How comics editors choose to interpret the data, and how heavily to rely upon the feedback. Some syndicate and newspaper editors say it can be an effective tool if used properly, and that write-in comments are more illuminating than a strict election. Others say polling can become a front, an excuse, a cop-out when comic-lineup changes are made.

Scott Adams, the creator of "Dilbert," is among those who believe polls used correctly can be an effective tool -- as long as they're not the biggest -- let alone only -- tool in the shed.


(AP Photo/Marcio Jose Sanchez)Enlarge Comic

"I think comic polls can be helpful to editors as long as they are just one of several factors they consider," Adams says. "For example, if only 20 percent of readers enjoy a particular comic, but they are passionate about it, then that comic probably sells more newspapers than one in which 80 percent enjoy it but not passionately. Editors would have a feel for the passion factor.

Adams cites his own strip's experience: "In the early days of 'Dilbert,' it often ranked near the very bottom in comic polls. But the passion factor was high, and the type of reader it attracted was a good demographic, so editors wisely gave it some time to grow in popularity."

Lynn Johnston, creator of "For Better or For Worse," also sees qualified value in comics polling. "It's only fair to ask the readers what they want. If an editor drops something and the backlash is strong, it's hard to replace the feature right away or perhaps impossible," Johnston says. "I'm all for what makes an editor's job easier and until they come up with something else, I think the polls are useful and necessary."

Often at issue, though, is the science behind the poll -- or the lack thereof. Sherry Stern, who edits comics at the Los Angeles Times, views polls at a faulty fallback position. "The more I think about it, the more I think a poll can be a cop-out," she says. "It allows us beleaguered editors to tell the complainers: 'But we did a poll.' But unless you had Gallup running your survey, the results are not accurate of anything.

"Just ask any editor who had to run a 'mea culpa' after bringing back 'Pearls Before Swine.' "

Stern also notes that the Times recently tested comics as tryouts to replace "For Better or For Worse" and asked readers for e-mail comments. " I found this way to be helpful because people often do more than just cast a vote. You end up learning something about your readers and what they think about many of the comics," she says.

Frank Rizzo, comics editor at the Atlanta Journal-Constitution, says his newspaper recently polled readers online because it was trying to elminate a half-page of comics. "I needed the data so I could go spelunking" and gather information, says Rizzo, adding: "It's about as perfect as we can do without an actual expensive scientific survey. It was a useful tool -- but it'll never be perfect."

After the poll results are in hand, though, Rizzo emphasizes: "Beyond that is where the 'editor' part comes in -- you have to interpret what people want. You have to know your audience."

Shirley Carswell, The Washington Post's comics editor, is in a similar camp: "I think polls are interesting for the insight they provide about strengths and weaknesses of particular comic strips across various demographics -- the open-ended comments are usually the best part. [But] I don't think polls are great as a sole basis for decisions, because the respondents generally are self-selecting."

Adds Carswell: "It's a part of the decision-making -- but not the only part."

At United Media, senior executive Lisa Wilson subscribes to the "know thy readership" approach. "I truly believe polls are a good way to find out how passionate your audience is about comics, but editors need to know their audience. Editors should never take [the poll results] from the readers and just do it. At the end of the day, they should make the decision."

Some cartoonists, though, just can't abide by online polls that might be so easily gamed.

"Stuffing the electronic ballot box has been going on since editors first starting running online polls," says Mark Tatulli, creator of the syndicated strips "Lio" and "Heart of the City." "However, it has never been more widespread than now, when comic-strip space is at a premium and artists and syndicates are desperate to get new or hold onto newspaper space."

Tatulli continues: "Should our income be decided by an unscientific, fixed poll? Is any other job in the word determined in such a way? Editors run these polls [they say] as a way of including the reader in on the decision of which comics to add/drop. The problem is editors, for the most part, are smart people and know these results are flawed ...

"They run these polls as a way of absolving themselves of any responsibility for what is dropped: 'Well, don't blame me, this is what the masses wanted!'

On that point, Wiley Miller, creator of "Non Sequitur," doesn't mince words: "If [editors] did what they're paid to do, they wouldn't need to even think about conducting these moronic polls."


Commenters and cartoonists, readers and editors: If you've got opinions, insights or anecdotes of your own, the e-floor is now yours.

By Michael Cavna  | February 6, 2009; 11:00 AM ET
Categories:  The Riffs  
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   Del.icio.us   StumbleUpon   Technorati   Google Buzz   Previous: The 'Riffs Poll: Fishier Than a Florida Recount
Next: What Comic Dazzles You With Its Art?

Comments

"Tatulli continues: "Should our income be decided by an unscientific, fixed poll? Is any other job in the world determined in such a way? "

Yes! but Life ain't fair.

Posted by: MSchafer | February 6, 2009 12:43 PM | Report abuse

When Lynn says polls are "useful and necessary," I think she is shifting the debate unnecessarily. We can debate the question of whether an editor should use his/her unilateral judgment or base it on a poll another day.

It's sort of like I'm saying that an election has been rigged, and she's saying democracy is a good thing. I don't necessarily disagree with her. It's just not the issue at hand.

The issue now is the scamming of these polls. If we're going to do them, how do we make them valid, as opposed to a scam, which many are now.

Posted by: StephanPastis | February 6, 2009 12:56 PM | Report abuse

Now look what two of the editors are saying about polls right here in this story:

"It's about as perfect as we can do without an actual expensive scientific survey. It was a useful tool"

"I think polls are interesting for the insight they provide about strengths and weaknesses of particular comic strips across various demographics...It's a part of the decision-making"

Even Lynn Johnston: "I think the polls are useful and necessary"

What these folks are not understanding is comic polls, whether write-in, online, or email are generally FIXED by motivated cartoonists and sales reps. Everytime! In a dicey economy there is real value to a creator or his salesmen to alter these results. It is easy to cheat and change results to the point of making the poll completely pointless and irrelevant.

Somebody please tell me how inaccurate, rigged, and completely flawed information is useful and insightful?

Mark Tatulli
LIO, HEART OF THE CITY

Posted by: heartofthecity | February 6, 2009 12:56 PM | Report abuse

For online comics, why not count the "clicks" to open a comic and view it, and then collect the data over time to help average out those who want to tilt the favor by multiple clicks or even just to view one more than once. Perhaps even take the online clicks data to help determine which to print? I subscribe to the print newspaper but often read the comics online because some are in color there versus black and white in the weekday issues and to catch my favorites that are not printed.

Posted by: zerodefect01 | February 6, 2009 1:32 PM | Report abuse

I still haven't quite figured out which strip you are accusing of running an astro-turf ballot-stuffing campaign.

Posted by: yellojkt | February 6, 2009 1:44 PM | Report abuse

Mark, you and Richard Thompson are hands down the best cartoonists in the comic strip biz these days. Keep it up man!

Posted by: jessecline | February 6, 2009 1:54 PM | Report abuse

Perhaps if these editors spent some time reading the comments section of any given Wash Post online story, they would realize that much of the participatory online audience is puerile, imbecilic and prone to carpet bombing to get their point across in relative anonymity.

(Seriously, the Post Online needs to figure out how the NY Times is managing their comments; the difference between Post comments and Times comments is the same as the difference between Milli Vanilli and Mozart )

Fellow blogger Marc Fisher pointed out the other day the 'commenters' were moronic bomb throwers, while his blog participants were Shakespearean by comparison. (My words not his).

So which group is doing the comics polls?

Whose opinion do they think they can capture?

Posted by: JkR- | February 6, 2009 1:56 PM | Report abuse

Michael: By the way, thank you for the opportunity here and for taking all the curse words out of my quote.

One tale I did tell Michael and he didn't use -- probably because it's hearsay (admittedly) but still fun, was this:

There was an editor in Washington state who was getting deluged with calls from people requesting a specific comic. The editor had never run the comic or even suggested to his readers that he was going to run it, so he was confused.

Finally, the editor got sick of all the calls, and called the cartoonist directly.

As you could probably predict, the cartoonist in question denied it vehemently.

But the editor said he never got another call requesting the comic again.

I think both editors and readers would be shocked if they know how prevalent this sort of thing was.

Posted by: JohnGlynn | February 6, 2009 1:59 PM | Report abuse

Echoing John Glynn's comment, I was contacted by one particular newspaper editor informing me that he had caught a cartoonist (who shall go unnamed other than to say he wasn't "Stephan Pastis") cheating so blatantly in a newspaper poll that while Pearls won 10 to 1 in the paper vote, Pearls lost 10 to 1 in the online vote. The cheating was so obvious that they disregarded all of the votes that strip had received.

What needs to happen is this: When a newspaper does an online poll and they are going to base their decision wholly or in part on it, they must require two things: 1) a local address and 2) a daytime telephone number where the voter can be reached (and no phone number can vote twice). Then, when the polling is finished, that same editor needs to spot-check a random sampling of votes for the top vote-getters. When they call the phone number, does an actual person pick up, and did they vote for that strip?

A bit more work for the editor? Sure. But to that I would say this: How would you like your job determined by a rigged poll?"

Posted by: StephanPastis | February 6, 2009 2:03 PM | Report abuse

Here is my question to editors that never gets answered:

Why are comics the only thing in the paper subjected to reader polls?

And if they ever did a poll on anything else, compare it to the response comics receive from the readers.

The very fact that they conduct these polls is an admission that it's the most popular section of the paper, which begs another question... Why do they treat the most popular aspect of the paper, the only aspect that's still attracting and holding readers, in such a dismissive manner?

Posted by: WileyMiller | February 6, 2009 2:56 PM | Report abuse

Tatulli continues: "Should our income be decided by an unscientific, fixed poll? Is any other job in the word determined in such a way?

Duh? The President?

Get a grip you bloated fools.

Posted by: ChillyBoston | February 6, 2009 4:00 PM | Report abuse

Michael,

Thanks for the acknowledgment and the very interesting post. As a longtime follower of polls (insert here cynical and jaded) I often go back multiple times to see how the results are tracking. As the numbers increase, although there are changes in percentages, rarely do they swing widely. To go back to a poll that had been posted for a couple of days and been very consistent in its results and find wacky swings just shouts out manipulation. The drop of "Pearls Before Swine" from a 29% to a 13% in hours just isn't possible without subversion of some type.

The Atlanta Journal-Constitution comic poll was written about on the blog of at least one of the cartoonists involved in the poll. He was encouraging people (non AJC readers) to go and vote for his strip. This poll had 3 votes to place and he wanted them all to go to his strip. It worked. The comic was kept in the paper.

Posted by: elyrest | February 6, 2009 4:40 PM | Report abuse

There's a very grey area in this situation, one in which Mr. Pastis, in defiance to all his posturing, shares just as much guilt. The big problem is that no editor actually knows where these online votes come from, unless they want to spend the days/weeks to cull every IP address to verify their location, if they even have the capabilities to do that (I've been in a lot of newsrooms - most of them don't). I've seen blog posts from Pearls fans from all over the country who spread the word whenever a new comics poll shows up, and they all march over to the poll and do their patriotic duty. Unless Mr. Pastis (or other creators), at the end of any poll he wins, says the following - "And I would like you (Mr./Mrs. Editor) to cull all of those votes and make sure I won fairly and squarely from only local readers," he can only hold the same court with the blatant cheaters. Sorry, but there is no room for high horses in this stable. I'll venture to guess that he happily cashes those checks from new newspaper clients, never bothered by the idea that 'ballot-stuffing' may have helped him, but will surely bellow if he feels it may have helped another. A nice fat fan base does wonders for a person's indignation. He doesn't give the actual marching orders, but ABSOLUTELY knows what is taking place. That which you do not discourage, you merely encourage.

It's like politicians who scream about unfair tax codes while keeping all their money in an offshore account.

Posted by: jmhumphrey | February 6, 2009 5:57 PM | Report abuse

The point is, if an editor feels the need to conduct a poll, it should be done in a professional, scientific manner just as any marketing survey is done, then analyze the data, broken down into demographics, emphasizing the one or two groups of readers that are most important for this section.

This takes a lot more time and money, but that's what will give you more accurate results.

Or, editors could just make their decisions on the comics in the same manner they do everything else in the paper. You know, the job they were hired to perform. You don't see them conducting polls on columnists, do you? Why should it be any different for comics? It ain't brain surgery.

Posted by: WileyMiller | February 6, 2009 6:32 PM | Report abuse

A few years ago, I was given the go-ahead to retool the comics page at the daily where I was working. I wasn't even in the newsroom, much less an editor, but they knew I was the guy who knew comics. (Okay, there's step one in the process. Most papers don't bother finding someone who is actually familiar with the medium.)

We didn't call what we did a "poll." But we told readers we were going to make some changes and wanted their input, mostly to make sure we didn't inadvertently cancel a favorite or unnecessarily keep a strip around that nobody liked. (Step Two: It wasn't a poll and we weren't going to base our decision exclusively on their answers.)

We had 21 strips up for discussion. I divided them into three categories: Old Favorites, Families and Modern Life. Never mind the names, the point was to put all the oldies in one place and then divide the others into the more cutting edge and the less cutting edge. Or something like that. We asked readers to select one strip in each group that they most wanted us to keep, and one in each group they most wanted us to drop. (Step Three: Don't ask them to rate the strips they don't really care about. Find out which ones they love, and which ones they hate. And decide at the outset on a balance of types, then rig the system so people have to choose among those types and not skew things in favor of one.)

We allowed both paper and on-line voting, and we asked for gender and age. But we only publicized it in the paper, because that's where the comics are. (Duh!) The on-line voting generally lined up with the results of the paper ballots. (Step Four: Don't be a putz.)

We used the results as a guide, not a Bible. Most of our responses were from the 55+ age group, which we expected. We honored their opinions, but we made sure the few 18-24s who responded, and especially the handful of under-18s, were well heard, since they spoke for the readers we needed to GET not the ones we already HAD. We made decisions based on maintaining a balance of types, getting rid of ones that a lot of people hated, and not ditching a few we didn't expect people to favor so passionately.

We didn't release the numbers. Fact is, I stopped counting about half way through because (A) the percentages were holding up and (B) it wasn't a poll. (Step Six: Release the numbers and you're asking to be second guessed. Do your job and make some decisions.)

We swapped out a third of our strips and then re-configured our Sunday funnies (which had been a package) so that all our comics ran 7 days. We had about a half dozen complaints, most of them in the "well, I wish you'd kept ..., but I'll give these a try."

Posted by: MikePeterson1 | February 6, 2009 8:32 PM | Report abuse

The thing about online polls of any form - whether it's comics or anything else - is that they are easily skewed by a motivated fan base.

For example, take the WaPo's "Beer Madness" competition from last year. http://www.washingtonpost.com/wp-srv/artsandliving/source/features/2008/beer-madness/index.html

Note the upper right "Specialty and Fruit" section, and the inclusion of Leinenkugel's Sunset Wheat.

When the Leinenkugel brewery - which is near our new home in Wisconsin and actually one of the biggest tourist destinations for the region - found out, they sent out an e-mail to all members their members/fan club (The Leinie Lodge), urging them to vote for Sunset Wheat. I don't know the exact numbers, but I don't know anyone who isn't a member around here (including ourselves).

What Leinenkugel's didn't grasp, however, was that the tasting and results were already set by the judges. The voting included in the bracket was purely for entertainment (which is now listed as a disclaimer underneath the voting numbers).

Now look at the voting rounds, you'll see that Sunset Wheat won by the largest margin in popular voting in the two rounds it survived. Indeed, the voting on it's second round - the round it was booted out - was the highest margin on any vote at 89% to 11%. Had it been all about popularity vote, I have absolutely no doubt that Sunset Wheat would have won the competition handily.

Discussion over Sunset Wheat (first about it's "awesomeness", and then complaints about how it didn't move to the next bracket) is probably in the plurality on the message board for Beer Madness. Many commenters (several who I actually recognized as my friends and neighbors) logged in purely to vote and post in the bracket comments - otherwise, they had no interest in WaPo.

The internet makes it easy for fan bases to skew popularity votes. Methodologies like Mike Peterson's (which are closer to actual market testing or user testing) are more effective for determining general popularity.

Posted by: Chasmosaur1 | February 7, 2009 11:14 AM | Report abuse

if I remember correctly, the San Antonio Express-News ran a poll that required the actual newspaper page, no copies, reproductions, etc. can't remember if name, # required, but as a aprizw was involved, probably most did put in id info. The actual page requirement I thought helped the ballot stuffing problem. What I believe they were really after was least liked, as Lio, Pearls before Swine and my cage(none of which I read)showed up fairly quickly. Can't remember what was dropped as I read so many on-line. Still can't get rid of Peanuts in the paper...I keep trying.

Posted by: ZeldaJane | February 7, 2009 11:52 AM | Report abuse

"The actual page requirement I thought helped the ballot stuffing problem. What I believe they were really after was least liked, as Lio, Pearls before Swine and my cage(none of which I read)showed up fairly quickly. ... Still can't get rid of Peanuts in the paper...I keep trying. "

The first sentence explains the second and third. While the tearsheet requirement solves the problem of Internet voter fraud, it worsens the demographic skew. Heck, I'm always late with bills that require a paper check, an envelope, a stamp, and I'm nearly 60. The folks who will take the time to clip out a sheet, fill it out, put it in an envelope are even older than me, and that's the reason fresh strips like Lio, Pearls and My Cage get nailed in these polls. And, yes, why you can't get rid of Peanuts, or Beetle Bailey or Mary Worth.

Posted by: MikePeterson1 | February 7, 2009 3:01 PM | Report abuse

My friend works at a newspaper fielding phone calls from various concerned readers. I wonder if a widget on the computer desktop to log these comments would be worth the data. (I'm told Lio gets the most complaints)

Posted by: Jumper1 | February 8, 2009 3:07 AM | Report abuse

But I love Lio!

Posted by: Jumper1 | February 8, 2009 3:09 AM | Report abuse

Ever since the days of The Phantom (pulled from our local Lake Charles, LA paper in 1952 (?)) and later (sadly) the disappearance of Beetle Bailey (from the LA Times), I have wondered how this polling works, never once (in my gullibility and innocence) suspecting that comic artists, poorly compensated and overworked ink-stained wretches that they are, would lobby for their own work. Disabused of my naivete, I now recognize that competition is of the essence in comics as well as all other art forms. I see that ballot-box stuffing is alive and well, here as in Illinois, and cast my vote for Lio and Pearls, as well as for my lost and lamented Citizen Dog, Prince Valiant, and so many others from the past. An avid comics fan since before I could read, I will pay closer attention to what is polled. I am aghast at the shrinkage in the number of strips in the LA Times' comics pages, as well as at the shrinkage of the physical size of the strips. I read no other papers, except for the occasional perusal of the OC Register for local sports news, and haven't been aware of the competition. I do know that for years I have voted to keep For Better or Worse, cannot understand why some others run (I won't name those I don't like here, to avoid offending some), but now that I plan on being a regular to this comics feature, I'll be back to post more comments later. By the way, my critiques are rarely content-oriented; if you want to read Mary Worth, go for it. I usually comment on the artistic merit. Comics are great. What ever happened to Gasoline Alley, by the way, and Terry and the Pirates, and, and. . . .

Posted by: CitizenGeorge | February 8, 2009 4:36 AM | Report abuse

I wonder if the statistics for on-line comics pages are available as an indicator of reader preference. What comes to mind is the Internet only strip Pibgorn on gocomics.com. Since it's on-line only, it can't have many readers, yet every day it has a large number of comments -- which, if only to me, seems to indicate a passionate readership. Meanwhile, the most popular comics seem to have few, if any comments, as if the readers are relatively indifferent. (Full disclosure -- I'm a New Yorker, and the NY Times doesn't have a comics, page, so I have to go to three web sites every morning to get an adequate fix.)

Posted by: su10 | February 8, 2009 10:33 AM | Report abuse

CitizenGeorge, I enjoyed your comments as I am a longtime comic reader myself. Gasoline Alley is still out there http://www.gocomics.com/gasolinealley/ - it celebrated it's 80th anniversary last year. Terry and the Pirates disappeared in 1973 (they tried to revive in in the 1990's, but it failed.)

Posted by: elyrest | February 8, 2009 11:46 AM | Report abuse

Oops! Make that Gasoline Alley's 90th anniversary.

Posted by: elyrest | February 8, 2009 11:52 AM | Report abuse

Wish su10 had named the three sites visited as I am looking for a second spot to find RETAIL, as the Seattle Post-Intellingencer may go away. Go Comics and comics.com do not have it, nor AZCentral.com or Houston Chronicle.

Posted by: ZeldaJane | February 9, 2009 8:21 AM | Report abuse

www.dailyink.com has RETAIL.

Posted by: ndancis | February 9, 2009 9:24 AM | Report abuse

I don't know if following the number of hits for online comics would really produce any useful information. I suspect a lot of people do what I do - read the comics in my print newspaper, and only use online for the few others I follow that they don't carry. I never read PBS, Pooch Cafe, CdS, or other fine comics online, because I can get them in print.

Posted by: marshlc | February 9, 2009 9:32 AM | Report abuse

Here a couple of great blog entries by a former writer at the Fort Worth Star Telegram, attesting to the cheating he saw in his comic polls:

http://househunting.typepad.com/

Posted by: StephanPastis | February 9, 2009 12:07 PM | Report abuse

The comments to this entry are closed.

 
 
RSS Feed
Subscribe to The Post

© 2011 The Washington Post Company