
Illustrations by Merijn Hos
My fascination with internet comments began as exasperation. I’d just written a short article that began with a quote from the movie “Blazing Saddles”: “Badges? We don’t need no stinkin’ badges!” After the story published, I quickly heard from readers explaining that, actually, the quote was originally from an earlier movie, “The Treasure of the Sierra Madre.” The thing was, I’d included that information in the article.
This was no isolated case: I soon published another story that mentioned, by name, a program called parkrun, and yet I got about a half dozen emails from people helpfully informing me of this cool thing called parkrun.
These episodes represented only a single type of comment, but they got me wondering about commenting more broadly. Only a small subset of readers ever comment. What compels them to take the time to weigh in? To learn more about the reasons that people comment, I collected data from two sources — an analysis of the comments here at FiveThirtyEight and a survey of more than 8,500 people. What I learned shifted my views about commenters and gave me some interesting insights into the hive mind.
Why comment?
The first thing I wanted to know was, why comment? What exactly are commenters seeking? A survey like ours isn’t perfect since it’s inevitably biased toward the subset of people most inclined to answer an internet survey (and, of course, self-reported results are notoriously unreliable). But it does provide a peek into people’s motivation. Our survey takers gave a wide range of answers, and my colleague Leah Libresco randomly sampled 500 of them1 and sorted them into categories describing their motivations.
CAT | SHARE OF RESPONDENTS | |
---|---|---|
Age | % | |
Under 20 years old | 5% | |
20-29 | 29 | |
30-39 | 28 | |
40-49 | 14 | |
50-59 | 12 | |
60-69 | 9 | |
70-100 | 2 | |
X | ||
Identify as… | ||
Female | 23% | |
Male | 76 | |
Other | 1 | |
X | ||
Comment frequency | ||
Daily | 24% | |
Weekly | 17 | |
Every few weeks | 10 | |
Monthly | 4 | |
Every few months | 15 | |
Once a year or less | 16 | |
Never | 15 |
REASON | PERCENTAGE | |
---|---|---|
Correct an error | 19% | |
Add to discussion | 18 | |
Give my personal perspective | 10 | |
Represent my view | 10 | |
Be funny | 8 | |
Praise piece or commenters | 8 | |
Ask a question or learn | 7 | |
Persuade others | 6 | |
Give my ego a boost | 4 | |
Take part in or interact with a community | 4 | |
Discover and express my thoughts | 3 | |
Vent | 2 | |
Troll | 1 |
Our respondents’ reasons for commenting mirror the results of a recent survey of 600 news commenters by Talia Jomini Stroud and her colleagues at the University of Texas at Austin’s Engaging News Project. In their survey, the top three reasons that people gave for commenting were “to express an emotion or opinion,” “to add information” or “to correct inaccuracies or misinformation.”

The bikes-and-dogs theory
Certain stories seem to generate a disproportionate number of comments, and after years of being on the receiving end of comments, I’ve formed a theory: The subjects most likely to elicit impassioned responses are those that feel personal to the reader (a real-life experience with the subject has made them feel like an expert) and those that hit on identity in some way. It’s based on something a newspaper reporter in Boulder told me many years ago. Back then, readers were still mailing letters to the editor, and they had a seemingly endless appetite to debate two things: who was at fault in conflicts between cars and bikes and whether dogs should be allowed to run unleashed on city trails.
To test this theory, I asked readers about the circumstances that made them most likely to comment. The answers lent at least some support to the bikes-and-dogs theory. But respondents’ reasons were more complex than my one, unified theory; commenters were also driven by a desire to provide their own information or to argue against an idea they disagreed with.
PERCENTAGE SELECTING | ||
---|---|---|
Know something about the subject that wasn’t in the article | 55% | |
Identify with the topic | 41 | |
Have a personal experience to add | 35 | |
Disagree with the author | 34 | |
Find an error | 33 | |
Think it’s a really good story | 31 | |
Agree with the author | 27 | |
Think it’s a terrible story | 12 |

How low do we go?
Since I started down this road after receiving comments from people who hadn’t read (or absorbed) the whole article, I also asked survey takers how closely they read a story before commenting.
Here again, I had a hypothesis: Maybe this commenting-without-reading phenomenon represents a variation of the backfire effect, in which a person who receives evidence that their belief is erroneous actually becomes more strongly convinced of the viewpoint they already held. In this case, the reader sees a headline that catches their interest and reminds them of something that they already know, which triggers them to think about their pre-existing knowledge or belief about the subject and then to blast it out to the world. The article they’re reading doesn’t inform them, it just provides an opportunity for them to reinforce (and broadcast) what they already know. I ran my theory by Brendan Nyhan, a Dartmouth political scientist who has studied the backfire effect. He told me it “seems plausible,” but said he wasn’t aware of any research testing this idea, “so in the spirit of your piece, I probably shouldn’t comment on it!”
When asked if they generally read the whole article before commenting, a few respondents to our survey said they only skimmed or didn’t read past the headline, but the vast majority of them reported that they read the story in its entirety.
That sounds encouraging, but I’m reluctant to take these answers at face value after talking to David Dunning, who’s a psychologist at the University of Michigan and one of the researchers known for identifying the Dunning-Kruger effect, a cognitive bias that, as the paper introducing the effect puts it, causes people to “fail to recognize their own incompetence.” “People are notoriously bad at comprehending what they’ve actually comprehended from text,” he said. “The correlation between what people think they’ve read and what they’ve actually read is quite small.” In a classic 1982 study, researchers asked study subjects to read a text that contained blatant contradictions and found that subjects who failed to find the contradictions still rated their comprehension as high. This could explain all those “stinking badges” comments.

Who’s in the club?
I’ve begun to think that many comments sections, including ours, are like a book club where members routinely fail to finish the book. The reading material is merely a starting point — the real purpose is to gather together to discuss interesting ideas. So who are the most frequent attendees at FiveThirtyEight’s discussion club?
RANK | NAME | NUMBER OF COMMENTS | |
---|---|---|---|
1 | Warren Dew | 552 | |
2 | Norman Shatkin | 360 | |
3 | James Deedler | 308 | |
4 | Fel Martins | 269 | |
5 | Joseph Michael | 252 | |
6 | Glenn Doty | 247 | |
7 | Django Zeaman | 234 | |
8 | Harold DePalma | 230 | |
9 | Nealy Willy | 217 | |
10 | Davey Williams | 214 |
Norman Shatkin is FiveThirtyEight’s second-most-prolific commenter. A retired computer programmer in Rhinebeck, New York, Shatkin leans liberal and calls himself a “politics junkie.” He said that he mostly comments to share his perspective on issues. But he’s also seeking a dialogue with people who have different viewpoints so he can swap ideas with them. He points to Dan Frushour (number 13 on FiveThirtyEight’s top 50 list) as someone whose political views are on the opposite side of his. “We share common values in some areas, and I find it valuable to exchange views with such people, because we can address each other in a common language.” Most of the people he encounters in his community, including the gang he gets together with every morning for coffee and doughnuts, have views similar to his own.
Warren Dew, a software engineer from Somerville, Massachusetts, has posted 552 comments — more than one and a half times as many as Shatkin — and that’s just at FiveThirtyEight. He comments elsewhere too, including cavemanforum.com, Amazon, the scientific journal BMJ, generational-theory.com and at least one bargain-hunting site. He’s most interested in political discussions and he largely comments in order to take part in political arguments with others “on the right of center.” I asked what he gets out of commenting. “It helps me clarify my own thinking,” he said. “I am always trying to test my ideas.” He prefers to lose an argument and learn something than win an argument and finish with the same ideas he had when he started.
Dew is not a troll, but his comments can be pointed. On a story of mine about hormones and athletic performance, he commented, “Sounds like Christie is just looking for an excuse about how she never could have gotten to the top, to justify her not trying.” On the phone, though, he was very pleasant and measured, and I asked if he spoke differently online than in person. “I can have a conversation and argument with people I disagree with on FiveThirtyEight and be totally blunt because I don’t know these people in real life,” he said. “In a social or work situation, I have to be more careful.” Online, he’s “brutally honest.”
Comments often serve as identity badges, said Joseph Reagle, the author of “Reading the Comments: Likers, Haters, and Manipulators at the Bottom of the Web” and a professor of communication studies at Northeastern University. “You see this particularly on social media,” he told me. The comment is meant to tell the world, “This is who I am.” People may also comment to gain approval and solidarity with their social group, he said. “It’s a way of saying, ‘I am like you.’” Of course, offering a correction can also be a means of propping up one’s own ego.
Entering the comments fray can seem futile. “What I have learned from the little commenting I have done is that it accomplishes nothing and provokes the very opposite of rational, thoughtful discourse,” wrote Elisabeth Carroll.
After reading all the survey data and talking to some commenters by phone, my views of comments and commenters have come full circle. Setting aside the trolls, who should, as a rule, be ignored or blocked, I can’t help thinking that on a certain level, commenters want the same thing I do — to have our ideas heard and carefully considered. We all want to have a say, if only we could find a way to stop shouting.
Additional reporting by Dhrumil Mehta and Leah Libresco.