Twitter  Twitter is public, but who does that benefit? nie.mn/1keiCd3  
Nieman Journalism Lab
Pushing to the future of journalism — A project of the Nieman Foundation at Harvard
nick-denton-cc

Pay attention to what Nick Denton is doing with comments

The NYU professor thinks Gawker is on to something with its attempts to surface quality comments.
Email

As comment bait goes, I’m a Gay Mormon Who’s Been Happily Married for 10 Years is a corker. That’s Gawker’s headline for a piece by Joshua Weed (excerpted from a longer version posted on Weed’s own site) about how he balances his homosexuality, his marriage to a woman, and his Mormon identity. Once up on Gawker, it quickly attracted several hundred comments.

As you might imagine, some of those comments are moronic. One read, in its entirety, “lol mormons.” Some are grating: “…the idea of putting something in a slimy vagina is not sexy” (a comment since dismissed). Some are ad hominem: “This article is bullshit, by a self loathing brainwashed moron.” Of the hundreds of replies, however, the seven currently highlighted on the article page form a conversation between Weed himself and and a handful of other users; the replies are civil, thoughtful, and even, mirabile dictu, spell-checked. These seven replies don’t just happen to be highlighted — they are highlighted because they are part of a conversation.

Most news sites have come to treat comments as little more than a necessary evil, a kind of padded room where the third estate can vent, largely at will, and tolerated mainly as a way of generating pageviews. This exhausted consensus makes what Gawker is doing so important. Nick Denton, Gawker’s founder and publisher, Thomas Plunkett, head of technology, and the technical staff have re-designed Gawker to serve the people reading the comments, rather than the people writing them.

The technical choices here are simple, but their social ramifications are not. The new design dispenses with the tyranny of time order. On most systems, the most prominent comments are posted either by the most obsessed users (when comments are posted oldest first) or the drive-bys (newest first). On Gawker, a user who replies to an existing comment is likelier to get her contribution seen than an earlier user who added another reaction directly to the original post.

Gawker’s default assumption is that most comments won’t ever appear on the article page — like the Slashdot comment system, they are all there, but only accessible with extra work by the reader. This ensures that there is, by design, no way for regular participants (the Commentariat, a group Denton loathes) to use either volume or aggression to maximize attention. On Gawker (and, soon, on its seven sister sites), anyone can still say anything, but it’s no longer the case that anyone can say anything to everyone.

This lets Gawker to do less policing overall. If vapid or aggro comments are unlikely to make the main article page, Gawker can expand its support for anonymous comments, as with its burner accounts (a nod to the phones favored by drug dealers) and its instructions about how to report information anonymously.

They have been rolling out this system in pieces — burners and the ability of commenters to accept or dismiss replies came in April, shifting most comments off the main page happened in mid-June, and comment search is still coming. Remarkably, as the system has rolled out, it has proven to work even retroactively — a lovely Gawker piece by Maureen O’Connor, When My Mother and I Were Obsessed with Death, acquired its comments back in May, under the old system, but when the replies are sorted under the new system, the three featured comments (out of a hundred) add up to a heartfelt thousand-word coda, written by other women grappling with the same issues. If you view the same hundred comments in time order, you come across someone hitting on the author — “maureen you are so attractive, death-obsession is just the icing on the cake” — long before you find anything you’d care to read.

Gawker’s plan might fail, of course. When there are complicated user reactions to big posts, as with the Gay Mormon piece or John Cook’s recent Watergate post, the system buries many replies that are, to my eye, more thoughtful and engaged than the ones that made it to the main page, including many interesting replies by the posts’ authors, one of the things Denton explicitly hopes to highlight.

There are also real design challenges — the reader is supposed to understand that a grayed-out comment placed next to a featured one is an alternate response to the same parent comment. (It was confusing even to write that sentence; the current design doesn’t make it much clearer.) Conversations now have URLs, so readers can send traffic directly to a particular set of replies, but few people seem to understand or use this function. Comment search has yet to be launched. All these things will affect reader reactions.

There could also be deeper problems. On the list titled “This Is Why We Can’t Have Nice Things On The Internet,” Item #1 has always been human nature. The Commentariat, who know little about the subject of any given post, still have a lot of time on their hands, and they may yet come to dominate the new system. It may also be that readers are in fact as horrible as industry consensus has it, that without the bread and circuses of cursing and trolling, commenting will lose its appeal.

Many news sites seemed to have entered an endgame with their comment sections a year or so ago: give up on any real participation, give up on anonymity, fall back to “Login With Facebook.” This reaction has come about in part because comments on news sites have been stagnant for a long time. Gawker is demonstrating that a good part of that stagnation came about because the way readers have been asked to participate has been stagnant for a long time too.

Nick Denton photo by Matt Haughey used under a Creative Commons license.

                                   
What to read next
Mashable_GrumpyCat-cc
Joseph Lichterman    July 22, 2014
The site known for social media and tech coverage has hired nearly 30 more editorial staffers since October and, like BuzzFeed before it, is expanding into more general interest news.
  • Apokrifa

    “mirabile dictu, spell-checked” – as in “The NYU professor things Gawker is on to something with its attempts to surface quality comments.” The limits of spellcheck…

  • Doug Faneuil

    We’re launching an iteration on the design above next week. Stay tuned.

  • http://twitter.com/RexHardbody Rex Hardbody

    It’s a horrible system that is unnavigable and does not do anything remotely approaching “elevating discourse” as Nick Denton arrogantly predicted. Commenters have already been having direct conversations with one another for years now. This system improves upon nothing but has so many defects as to render it a complete and utter failure.

  • http://twitter.com/sjkmcnally Steve McNally

    Denton and Gawker do keep pushing forward – and they don’t seem to mind the backlash that happens with any change at all. 

    As for treating comments as “part of a conversation,” we’ve been doing that on Forbes for the last two years. We’d been doing it at True/Slant the two years before that. Your fellow, Prof. Rosen, has been admiring our work for some time. 

    https://twitter.com/jayrosen_nyu/status/207925286061342721

    Excellent that you mention Slashdot: their Moderation, Meta-Moderation and Karma system for engaging the community in assuring quality is still ahead of most others.  They were “gamifying” curation before either of those were in vogue. 

  • Penelope Floofdoodle

    If a website’s design is difficult and troublesome for users to understand, that design has failed.  Period.  End of story.  Usability is at the heart of all good web design.  By your own admission, Denton’s new commenting system is counter-intuitive and difficult to navigate and understand.

    And let’s not kid ourselves here.  The true purpose of PowWow is to monetize comments, to funnel users into prominently displayed conversations with advertisers.  It isn’t even remotely about elevating discourse or highlighting valuable user input.

  • Ben Brown

    I appreciate the attempt to creep towards greater civility in comment sections on news sites. But I think a strategy like this one comes at the problem from the wrong direction.

    It starts with assuming there is some sort of obligation to honor all comments with publication just because space limitations are suddenly irrelevant. Even in the ancient days before digital delivery, editors would have balked at that idea. Now, when so many publications — especially newspapers — are struggling to defend their brands in the face of so much competition, it’s depressing to watch publishers conspire with digital vandals to undercut their publications’ credibility and authority by anointing all comers with the same rights as their reporters.

    If a publication doesn’t have the resources to actively monitor comments the same way they monitored letters to the editor in the old days, they should end the invitation for comments from readers — who, after all, now have near instant access to broadcasting their opinions all over the world for the price of an internet connection.

  • http://www.spot.us digidave

    TLDNR.

    FIRST! 

    and other such fun comments to leave…

  • http://twitter.com/westseattleblog West Seattle Blog

    Ben suggests it is an error to assume “there is some sort of obligation to honor all comments with publication …” EXACTLY. We do NOT publish all comments. Irrelevant comment? Insulting comment? Profane? Trolling? Perpetuating a tired old argument (around here, it’s “bicyclists vs. drivers”)? Sorry, it’s not going through. And speaking of “tired old arguments,” don’t tell me I can make this work because my site is small, but you can’t. If I can monitor and moderate 200 comments a day, a site with 10 employees can do it for 2,000, and a site with 100 employees can do it for 20,000. If you don’t want to pre-review every comment, then set up a really great filter that will catch 95 percent of the potential problems for your review. I believe it’s a news organization’s obligation to have and enforce rules, and it is lazy and irresponsible to just let the sewer keep flowing. – Tracy in W. Seattle

  • http://twitter.com/fmbandit nick, man

    It’s hard to “pay attention” if you have no idea what’s going on. Enjoyed your use of “bread & circuses” though :) Kudos to those who point out that this is not about elevating discourse or even relieving moderation, it’s about $$$

  • Clay Shirky

    Of course the improvement is about money, but that is as true of the NY Times comment system as it is of Gawker’s. As long as we are talking about commercial media, we can count on commercial motivations. 

    This isn’t in as much conflict with elevating valuable user input as you think, since the thing people visit Gawker for is _to read things that interest them._ Even with the most cynical possible view of Denton’s motivations — he wants to trick commenters into providing the kind of comments readers would be happy to read — this is still consistent with sorting for value.

    This is also where I think you overestimate the effects of the UI issues. Gawker could certainly use a better set of visual cues for navigating comments, but that only applies to readers who want to read comments other than the ones shown by default. When you consider than 90% of users will never do anything other than use the system in its default setting, and that for any given article readers outnumber commenters by at least 100 to 1, the UI issues are minor. 

    I’ll say it again — don’t assume that a commenting system that serves readers instead of writers is a bad idea until you look at that Maureen O’Conner article. The value of the comments between the two systems is night and day.

  • Clay Shirky

    It is unnavigable _by design_, so I don’t think you can regard it as a complete and utter failure. Navigability would that anything anyone writes can be easily seen by everyone; Gawker clearly sees this as a liability.

    You may not like what this redesign succeeds at — elevating a tiny subset of comments, without regard for the dedication or commitment of the commenter — but that is different than failure. Failure would be no one commenting and no one reading comments after launch,  clearly not a problem Gawker suffers from.

  • Ben Brown

    I’m with Tracy. Take control of your brand or go home.

  • http://www.facebook.com/profile.php?id=100002465852145 Vadim McNab

    Bah! Bullocks. The internet’s bread and butter is in moronic comments.

     

  • http://twitter.com/SailRabbits Magister

    You’ve drawn a nice, but somewhat inaccurate picture and though the easiest thing would be to let it slide, we’d be sacrificing one of the values of comments. It seems that any blogger, whether they’re a twenty-five year-old with their first job or a seasoned professor can make mistakes which some people may not see, while to those with more information, it could possibly undermine the post.

    Again, apologies and I often let these things go, but…

    All of your examples of bad or distasteful comments were made by burner or non-registered accounts. Under the system which Gawker had employed for the previous four years, none of them would have ever displayed. Under the current Gawker system, the burners are almost a necessity because there’s no army of “starred” volunteer moderators who can bring good comments into view, but a downside of the burner accounts is that anyone can say anything at any time.

    For example, the default conversation on the Mormon post now has the seven exchanges mostly between the post author and one of the Gawker bloggers timestamped fourteen days ago, while three days ago, a burner decided to grab onto the end with questions and allusions which are much more personal than the original seven. Now of course, The_Weed could dismiss this eighth comment, but it doesn’t look like he’s logged into that account since the post — it may have been created for just that one use, as are most burner accounts– and a mod could delete it, but they’re unlikely to see a post-post comment unless they follow a link.

    By the way, while I’m on the subject of the Mormon post: The seven comments are a conversation, as are many of the other threads. The one which displays as the default was given that status because it has participation from a staff member and I’m guessing it comes first because both the post author and the Gawker staffer replied multiple times, while on the next thread, a staff member only commented once. Under the current system, any conversation which includes a staff member goes into the top spot and one which may have specific information from another source is down the list. Though it may work in this case, it should go without saying that just because a thread has comments from a blogger, it’s not necessarily the best thread.

    Also, it should be pointed out that Gawker changed from what most people considered the best large-site commenting system — crudeness and the type of trolling which dominates other large sites, if they could get through to be displayed, would have resulted in a suspension or ban — in late April. This is how a burner was able to post a come-on to Maureen in May. It wasn’t under the system which had existed for years and it’s a comment that would have only been displayed after the change. Comments made under all of the previous commenting systems still display as they were made. The system defaults to “featured”, so you have to click “all”, but comments from prior to April haven’t been translated to PowWow.

    Now, I could go off into a discussion about who reads comments and how they do it. There seems to be a perception among parts of the industry which isn’t entirely accurate and is full of conceit, but I agree with Denton that the conversation can be improved. I’ve also said on Gawker and in other forums that I like the current incarnation of the system, though there is room for improvement. What he’s trying to do certainly bears watching and if he keeps at it, he could produce another model, but  there have also been recent indications that the system may change again and at this moment, we don’t really know how it will look or function in the end.

  • opendna

    Any idea what method Gawker is using to choose the “best” comments? Is it algorithmic, reliant on a human editor or directed by a crowd-sourced karma system?

    IMO, Slashdot and Kuro5hin (from which DailyKos forked) solved the comment quality problem years ago. I wouldn’t call a CSS hack to prettify 10-year old methods “innovation”, but maybe Gawker’s working on something genuinely new. If so, I’d love to hear more about it.

  • http://twitter.com/RecentrTheRight Martin Long

    Escaping the adolescent tropes that pass as “comments” is every bit as satisfying as the moment when the group of middle schoolers finally leaves the subway car.

    Who wouldn’t want to stand next to the thoughtful (and dare I say, intelligent and introspective) people at a gathering and discuss, well, almost anything?  Thank goodness we might finally, at long last, have a way to tune out the blathersphere to find sentient people.

  • http://twitter.com/holgate Nick Sweeney

    I strongly disagree about Slashdot: the incremental attempt to filter through mod, then metamod, then karma was ultimately a failure, because it didn’t address the fundamental problems of scale. Arrive at a /. thread in its heyday within an hour of posting, and there’d be 3,000 comments, a few dozen standouts, each with a handful of subthreads, and absolutely nowhere to enter the conversation.

  • Clay Shirky

    Magister, 

    A lot of this has to do with the goals of the new system. I certainly agree that the thread with the most replies from the author is not necessarily the best thread — indeed, I made  just this criticism of the comments on both the Gay Mormon and Watergate threads. 

    However, when you say “a downside of the temporary accounts is that anyone can say anything at any time and if they put it in the right place, until it’s deleted, it could be seen”, I’d reply that Denton and Co don’t regard this as a downside. 

    Put another way, Gawker has given up on trying to reduce the writing of bad comments, in favor of reducing the _reading_ of them. They don’t have to care that WowJustWow made an inane comment to O’Connor, so long as no one reads it. Similarly, they regard the reduction in the power of moderation as a victory for the current system, in order to reduce the power of the regulars. 

    (What remains to be seen with the burners, of course, is whether people use them to report news Gawker could not otherwise have gotten. That test is still in the future.)

    Where I will disagree with you most, however, is this: “because the system may change again, we don’t really know how it will look or function in the end.” They have been working on this for some time, and are rolling out the last of the planned features this summer. There will doubtless be tweaks in features and UI, but I think we can in fact look at this system and know how it will function. 

    People (like you, I think) who want hard upstream controls on what the writers write will doubtless like it less than the current model of soft downstream controls on what the readers read.

  • Clay Shirky

    Nick, I think this is exactly right about Slashdot — the karma/mod/meta-mod system (in addition to being quite baroque) created a success crisis, where the improved signal drew in more users, which degraded the signal.

    Also, Slashdot is considerably more biased towards ‘annotative’ comments, as each comment is given explicit value on its own.

    What I should have emphasized more in the initial piece is that Gawker only cares that it find a few good comments to show the reader — unlike /., they have almost completely abandoned the fact of writing as anything other than a lottery ticket for attention.

  • Raskol

    The wide range of opinions expressed in the comments have always been the best part of the Gawker blogs, especially on Jalopnik. Sure, there are always some trolling or distasteful comments, but I would still like to read the comment which have the most value to me, not the ones which Nick Denton thinks I should care about. Comment systems should elevate discourse by rewarding civility and educated responses (as the star and featured systems do). Allowing a limited number of comments to be “easily” (ha!) accessed only encourages people not to comment, because they know their comments will never be seen. I have always disregarded people’s threats of leaving Gawker because of design changes and whatnot, but if anything threatens open discourse and the readability of the comment sections, I will no longer be reading anything from Gawker websites.

  • Dan Vergano

     One real lost opportunity over the last decade has been the chance to incorporate source comments into articles after publication. If comment threads automatically bumped expert comments from sources mentioned or used in an article (or story topic experts in general) to the top of the queue (or a separate expert queue), and gave moderator rights to sources for the resultant thread, we might have had informed discussion become more normal and created better journalism.

    The experts might have called out factual errors in the stories, but that would be all to the good. It would offer sources better recourse beyond a letter to the editor as well. If at the end of an interview I offered a source a slug and a password to comment as an expert, it would add a minute to the interview and might elevate an ordinary piece into a great one once they start interacting with readers and with each other. As things stand now, an expert would have to be demented to drop their comments into the moronic inferno.

  • http://about.me/scott.bryson/ Scott Bryson

    What might be missing in this discussion is what made Gawker unique when it started and what the reader culture was like that added to its flavor. It was originally a smarmy gossip site that elevated itself above some of the others with a little more wit and putting the writer’s twists on stories that were trending on the web.

    The comments were originally a closed club that one had to be admitted to, and that probably created the in-crowd vibe that later evolved to the reader culture you state Denton hates. Ironically the comments were often more funny or entertaining than the articles, and there was a healthy sense of play between the authors and readers.

    That is what made the site more interesting than many others. It was basically TMZ & Perez Hilton with a little more New York intellect, while not being too stuffy & journalistic to troll the gutter once in a while.

    Now it is being discussed here like it is the New York Times, a similar bastion of journalism, and what initially had a flavor not unlike Fark or Reddit with a little more editorial control, is now being talked about like an example of modern Internet journalism worthy of scholarly debate.

    The fact that in the old days the occasional troll would write something so tasteless and unbelievably eye-rolling was actually entertaining as the territorial regulars would put the writer in their place, which added to the humor. There was wit & zingy puns, and often the comments would out-shadow the articles. The open nature and survival-of-the-fittest reader culture seemed to regulate itself in a cooperative way.

    Then the bean counters and SEO types took over and it just devolved into a parody of Time Inc. in the post AOL days, where accountants and MBA types misplaced the salty journalists. Ah corporate culture!

    Well the way the web works, some young turk will resurrect the old Gawker format and probably reap success, while the Denton empire goes the way of the Huffington Post, sells out and makes a lot of money. You might want to check their typical trending stories before adding your scholarly credentials to this debate ;)

  • MissNormaDesmond

     Thanks. I can’t help feeling that this article was written by someone who doesn’t actually read Gawker on anything approaching a regular basis.  If he did, he’d recognize that the “conversation” he’s happy about was most likely highlighted because it’s between Gawker writers and the subject of the post.  In other words, the algorithm hasn’t magically sieved out the best comments from amongst the hoi polloi, it’s simply weighted to favor Gawker writers, who one would hope are more than capable of asking intelligent questions.

  • MissNormaDesmond

     The hilarious part of this article is the pretense that the new commenting system had anything to do with preventing incivility.  In reality, it has made the site prey for every troll on the Internet with a bit of spare time on his or her hands, because the Burner accounts can be set up within seconds, are anonymous, and have full commenting capability.  For the first few days of their existence they were even permitted to post pictures, which resulted in the site’s being awash in photos of aborted fetuses.  This gives you some idea of the Olympian heights to which the level of discourse has been elevated; the same people are still there, commenting away, they just can’t post pictures.  Yet it’s the Commentariat, who in the previous system had helped to keep such comments from being seen, that Denton elects to hate.  Why doesn’t anybody ever question that, I wonder?  Or, for that matter, the gift-horse orthodontolgical examination involved in hating people who are devoted followers of your product, and antagonizing them every chance you get? 

  • MissNormaDesmond

     This man is a Nabokov devotee, and as such, obviously a writer of no small discernment.

  • http://twitter.com/SailRabbits Magister

    I appreciate you taking the time to respond. As you’ve undoubtedly noticed, posts about Gawker do often get more comments. That’s mostly Gawker’s fault because for the past year or so, they’ve rarely reported on themselves and because they have a devoted fanbase, the comments about them need someplace to go.

    Because it’s most likely of little interest to you and your readership, I’m not going to get into a big discussion of the algorithm and how the site currently works. My main points were that the comment to O’Connor wouldn’t have been seen under the “old” Gawker (hierarchical) system, so it’s not an example of something the new system would hide; it wouldn’t have existed, before. And, my point about a downside of “burners” — not the algorithmic system, but the ability of an unregistered users to make a comment — is that eleven days after the Mormon post published, someone came along and put a comment in the right place which has been seen by everyone who clicked your link. The algorithm didn’t hide the questionable comment and because it was put into that particular spot, unless there’s human intervention, it will remain in the default thread.

    And finally, as I said previously, I like the current incarnation, but it obviously needs some tweaks. You’ve mentioned comment search, Nick has talked about personalized views and there have been several indications that the UI may change next week to something which will put even more emphasis onto the default path through the default thread. So, though we think we may have an idea, I don’t think anyone, including  Nick Denton knows where this experiment will end. Which, we’ve agreed, bears watching.

  • Irony Alert

    “Thinks”?

  • http://twitter.com/iamfantastikate Kate T.

    I frequent Gawker sites, and I have a habit of scrolling down to the comments, which, in general, have not been bad with their star-rating system in place. With this system, I don’t even bother scrolling down because I can’t fucking tell who is talking to whom, and so the comments become meaningless to me as a user. I’m sure some believe that’s great, as it returns the focus to the article and the reader to passivity, but I think it’s bad design.

  • http://annatarkov.posterous.com Anna Tarkov

    I refuse to believe that without cursing the trolling, commenting loses its appeal. That is a dystopian woldview that’s simply not true. Look at the comment on NYT which, yes, are moderated. Moderation is key. Now, many sites don’t have the money to hire moderators. And I think even NYT is doing away with them to a large degree, because they’re going to start using Facebook comments. And of course if you start out without moderators and then try to start using them, the Commentariat would rebel. But Nick doesn’t care about pissing off those people off anyway so I’m not sure why he wouldn’t try it. Also, what does Business Insider do? They attract the slime of the Internet so they must deal with it somehow. What does Boing Boing do? What does TechCrunch do? Why is no one asking these sites how they handle comments?

  • http://annatarkov.posterous.com Anna Tarkov

    I just don’t understand this. If the Commentariat feels it has been so wronged, why don’t they just go elsewhere? Why don’t they start a Google Group and post Gawker links and talk amongst themselves? Or if they want it to be public, start a blog and talk over there and everyone will see it. Why does it have to be on Gawker? Why would anyone want to stay somewhere they’re not wanted? I don’t get it.

  • YourLittleBrother

    “Most news sites have come to treat comments as little more than a necessary evil, a kind of padded room where the third estate can vent, largely at will, and tolerated mainly as a way of generating pageviews.”

    Comments at news sites are atrocious specifically because they are completely ignored (except for banning purposes) by the authors, editors, and publishers.

    Implement a decent crowdsourced spam/moderation mechanism, and then demand that journalists spend 30 minutes per day replying to their articles, editors spend 60 minutes per day on their articles, and publishers two hours each week on their publications, and you’ll see much better comments.

    I’ve seen many occasions where the commenters are clearly more knowledgeable, or insightful than the reporters and have lots to offer to improve the story or lead to a follow on.

    But nothing ever happens.

    At which point the best thing to do is just to spray graffiti and otherwise pee on the article.

  • http://www.facebook.com/simondowens Simon Owens

    At the news organization where I work we’ve been using Facebook’s commenting widget and have found it moderately successful at improving the quality of comments. Some reasons include:

    1. Though we allow users to also sign in with Yahoo, AOL, and Hotmail, most use their Facebook accounts, thereby forcing them to use their own names. This significantly cuts down on trolling.

    2. Facebook comments allow threading, which isn’t exactly a new invention but many news organizations still have a simple reverse chronological order system, and I’ve found Facebook comments to look cleaner than many other systems. Also, because Facebook comments are by default published to a user’s own wall, you’ll see a healthy mixture of people replying to a person’s comment to be both people who came to the article independently and those who found it in their news feeds. 

    3.  Because of the ability to “Like” a comment, better, more thoughtful comments often float to the top. Facebook also sorts comments partly based on the number of subscribers or friends a person has, so it adds at least a little bit of authority into the mix, though it doesn’t make this the sole factor used to determine the order of the comments. 

  • http://twitter.com/chrisamccoy Chris McCoy

    We’re building a system at YourSports that sorts comments by:
    - Your comments first, sorted by # of fans (our version of “like)
    - Comments from those you’re friends with, sorted by # of fans then time
    - Comments from those you’re following, sorted by # of fans then time
    - Comments with most # of fans
    - Comments by users with most # of fans across the entire system
    - Comments by date from newest to oldest

    It requires us to have a native social graph and native reputation system–and it has required us to re-architect of how we use counters in our technology stack–but we definitely think it’s the right approach in a strong-tie/weak-tie/no-tie network of passionate sports enthusiasts.

    We modeled our approach after Facebook’s public commenting engine, but they aren’t accounting for reputation or weak-tie relationships.

    Conversations are the atomic unit of content. Any new service or content platform should design from the comments on up. They are core to the system and have really been neglected by premium publishers as a source of revenue.

    There are definitely revenue opportunities within the comment stream.

    We’ll release that next.

  • http://www.andrew-turnbull.com/ Andrew Turnbull

    Forbes does something interesting with ‘called out’ comments.

    http://www.forbes.com/sites/jennagoudreau/2012/06/21/the-best-jobs-that-dont-require-a-bachelors-degree/
    (random article from the front page)

    Unfortunately, while it highlights interesting contributions to the discussion, it obscures the discussion itself. I think there is room to improve commenting on the web and we’re slowly getting there (disqus 2.0 is another good example). It’s not an easy problem to solve though. So in the meantime I guess we have to be patient (and find the few communities worth contributing to)

  • http://www.markpack.org.uk/ Mark Pack

    I wish more of the people who run the sites I tend to read had your attitude Tracy! I think you’re spot on in wanting to see comments as a service to the readers.

  • http://www.markpack.org.uk/ Mark Pack

    I think you confuse “devoted followers of your product” with “people who post comments on your site”. Not only because some commenters seem to love to hate the site they’re posting on…! But also because all the stats I’ve seen for different news and blog sites show there being hugely more regular loyal readers than their are commenters.

    Even if every comment was a devoted follower, they’d still be a pretty small proportion of the total number of devoted followers.

  • Paultatara

    It’s impossible to follow the Gawker comment system intuitively.  You make a comment, then it’s gone into the mist.  Try finding it without an innate sense of something that can’t even be written about clearly!  It’s a ridiculously terrible idea that’s driven me away from commenting on the site’s articles…and I know I’m not the only person who feels that way.  How this is a big, exciting breakthrough is beyond me.

  • http://twitter.com/sjkmcnally Steve McNally

    Slashdot’s default view could be a mess. On those highly-trafficked posts, especially, it was up to members to browse at +5. Then you’d see only the few dozen standouts.

    Doling out small batches of Mod and Metamod points to a large network helped keep the influence of “enterprising cabals” that devalued the community – as it did on Digg – in check. The relative scarcity and short life of mod points meant people used them in a timely  manner. 

    Putting the filtration onus on individual community members is an editorial decision. By default, e.g., we display comments on articles only when it’s been explicitly called out by an author. Removing that filter, rather than applying it, is up to the individual.  It’s still people, with the assistance of technology, that do a much better job at determining relevance and quality – one man’s Insightful is another’s Offtopic.

  • Msimon

    The measure of this system is whether or not the ads can be attracted to a comments page….for the most part comment pages are dead or the land of google pop up,,,,,Denton is not about charity…..

  • http://twitter.com/JohannaScott Johanna Scott

    Can you explain how the technology works? How does Gawker “decide” which comments rise to the surface – is it human judgement or an algorithm?

  • MissNormaDesmond

    Some people have done that precise thing — there are multiple “Gawker refugee” boards and websites.  Choire Sicha’s blog, The Awl, has many commenters that were formerly at Gawker, as one would expect.  Since I am not in fact the loser without a life Denton delights in depicting commenters as, I don’t really have time to run all over the web looking for people, though.

    Speaking for myself, there was, and occasionally still is, a balance of heavy and light there that I enjoyed a lot, and a group of people, including many of the writers, I enjoyed playing with.  Because it wasn’t just a “talk amongst ourselves” for regular commenters,  it was an interaction with the writers and editors as well.  I adore Hamilton Nolan’s writing; we have precisely the same sense of humor a lot of the time, and the same things tend to piss us off, which of course is key in a writer you like commenting on.  I like Caity Weaver, Max Read, and Louis Peitzman very much as well.  I’m getting to like Rich Juzwiak.  In the past, I loved Alex Pareene and Richard Lawson.

    Denton is portraying the commentariat as not being wanted, but the person it’s not wanted by is mainly him, not the people one finds writing at Gawker on a day-to-day basis, from what I can see — although I suppose I may be kidding myself.  Some people just don’t interact, which is fine; others do, and it’s fun.  Under previous EIC’s, the atmosphere was very different, especially with respect to how commenters were treated.

  • MissNormaDesmond

    Yes, they would.  I don’t believe I said anything that would indicate otherwise.  I think you’re perhaps confusing “people who are devoted followers of your product” with “all the people who are devoted followers of your product”, which isn’t what I said or meant.

    You don’t seem to be referring specifically to Gawker in your comment.  I think it may be a mistake to see the situation there as being interchangeable with how commenting works at other sites.  When Denton refers to the despised commentariat, he’s not referring to all commenters, but to those who comment frequently, who were “starred” under the previous system, and acted as hall monitors, so to speak, to keep out trolls and try to keep the tenor of the conversation interesting and fun.  In other words, the precise people previous writers and editors had selected as being the most valuable commenters, Denton hates.  Rather than seeing them as an asset he could work with, he’s chosen not merely to alienate them on the site, but to badmouth them all over the Web.  This from a man who likes to bust out The Cluetrain Manifesto as a reference for how communication with one’s market should be done.  I can’t be the only one who finds that odd.

  • MissNormaDesmond

    Again, the Commentariat isn’t all commenters, the commentariat (I don’t like capitalizing it somehow, it seems pompous) are the regular commenters who previously acted as moderators themselves.  We’ve been begging for more moderation.

  • MissNormaDesmond

    I feel as if you really are my little brother, and I would like to hug you.

  • http://annatarkov.posterous.com Anna Tarkov

    Sure, I know the Awl. I didn’t realize it was a haven for Gawker refugees. That’s interesting. 

    So tell me this… under the old system, were starred commenters welcoming to newbies or did they wield their power ruthlessly? Or was it a mix of both? I had thought that this is something else Denton was trying to eradicate, but I could be mistaken.

  • Snertly

    I think Denton didn’t like people saying how much they enjoyed reading the comments on Gawker, so he fixed it.  The odds of reading enjoyable comments on Gawker has dramatically decreased. 

  • Snertly

    If I may interject, regarding newbie treatment, it seems to be very dependent on how the newbie conducts his or her self.  When I first bumped into Gawker about three years ago, I really liked the emphasis on writing properly.  The base line behavior was described as use caps properly, write in sentences, be reasonably civil, and lastly, any of these rules could be broken in a one off fashion, if you did it well enough. 

    As a whole they were/are a bright, witty, and expressive group, delightful to know and an honor to be accepted by. 

  • Snertly

    If I may interject, regarding newbie treatment, it seems to be very dependent on how the newbie conducts his or her self.  When I first bumped into Gawker about three years ago, I really liked the emphasis on writing properly.  The base line behavior was described as use caps properly, write in sentences, be reasonably civil, and lastly, any of these rules could be broken in a one off fashion, if you did it well enough. 

    As a whole they were/are a bright, witty, and expressive group, delightful to know and an honor to be accepted by. 

  • Snertly

    If I may interject, regarding newbie treatment, it seems to be very dependent on how the newbie conducts his or her self.  When I first bumped into Gawker about three years ago, I really liked the emphasis on writing properly.  The base line behavior was described as use caps properly, write in sentences, be reasonably civil, and lastly, any of these rules could be broken in a one off fashion, if you did it well enough. 

    As a whole they were/are a bright, witty, and expressive group, delightful to know and an honor to be accepted by. 

  • NotExcited

    “but when the replies are sorted under the new system, the three featured
    comments (out of a hundred) add up to a heartfelt thousand-word coda,
    written by other women grappling with the same issues”

    But why should this be the goal? Why should we only seek to promote a few viewpoints at the expense of others? Sure it partly solves the problem of trolling, but you risk hiding sincerely good comments behind a labyrinth of clicks.