If you've spent any amount of time on Dev.to in the past year, you've certainly noticed the uptick of low-quality "listicle" articles (often referring to the low quality click-bait type articles in a "Top 5" style). Or if you haven't noticed it yourself, you've probably seen other users complaining about the uptick - even making the top 7 posts published last week [Dec 11th, 2023].
While this trend may or may not have been exaggerated with the increasing availability of AI tools, this concern about listicles has been a discussion point on Dev.to for a few years. This past weekend while listening to the Cortex Podcast, Host CGP Grey mentioned an economic phenomenon called "The Missing Middle" and how it related to the landscape of YouTube this past year, and it made me think about the listicle situation on Dev.to and how it might be less of a problem and more of a natural economic occurrence for a growing platform.
Disclaimer for Economic students - I know that this isn't exactly how the missing middle is intended to be used, but I couldn't find better phrasing for it and this article was the closest thing I could find. If you know a more appropriate phrasing for what I'm explaining please let me know in the comments ๐๐ป For readers interested in the specifics of Missing Middle Housing, check the Wikipedia page
So what is "The Missing Middle"?
Originally, this was about the housing market and the loss of options in the middle class range, but applied at a higher level this is the phenomenon where over time, a market goes from having a variety of selection in the low, middle, and high quality range - but over time the middle quality options start to disappear as the supplier trends towards either the low or high quality side of things.
This can be for a variety of reasons - in the housing market partially this was caused by changes in zoning laws/requirements making housing harder to build. For online content on platforms like YouTube or Dev.to it can be because of what the algorithm rewards. CGP Grey calls out that there's been a trend on YouTube where users are favoring either short form content (Shorts) or particularly long form content (4-hour documentary deep diving some obscure interest). I particularly like the 2-axis spectrum that he proposes, one axis for effort and one axis for length, where the points that the algorithm is rewarding most is the extremes of the axis - Short + High Effort, Short + Low Effort, Long + Low Effort, and Long + High Effort.
Note that Effort isn't necessarily the exact right word here, but is a "close enough" word to describe something along the lines of "doing the research and validating that everything is going to be accurate for the reader"
Naturally, the most content is going to be created in the Short + Low Effort extreme since that requires the least amount of time/effort to create - which I think applies fairly directly to the Listicle situation here. But the high-effort content still exists - the Top 7 posts each week often has a few articles that I find interesting and will read through - the question becomes how can interested readers find that content outside of the Top 7 posts each week?
I don't necessarily have any profound new addition to the situation, but thought it was interesting that it aligns so closely to what creators on other platforms are experiencing, as well as the physical space as well.
Top comments (24)
I always liked dev.to for the very welcoming attitude. Low quality articles will simply be ignored, but the decision is not done by a maintainer, but by people that just do not read the low quality stuff.
The only downside for me is the fact, that high quality stuff gets easily lost in the background noise and may be hard to find. It would be nice if there was some kind of "dictionary", that gives better access to the important stuff. Maybe similar to the badges and article could be voted up to be added with a certain number of upvotes.
I also like the welcoming attitude! and the discussion has been going on for some years now. As long as we bookmark, upvote and link good articles, the low quality stuff will hopefully just come and go.
The low-quality stuff is annoying. Marking a post as low-quality is slightly harder than it needs to be. Currently, you have to click the little shield icon, then click the ๐ icon. If the โค๏ธ is always shown at the top-level on the left, why aren't the ๐ & ๐ icons also always shown at the top-level on the left to make them easier to click?
(Aside: it's not clear why the shield icon is even a shield. What is that supposed to connote? I'd guess that many users have no idea what that means whereas a ๐ & ๐ at the top-level on the left will be obvious what they mean.)
(Aside 2: when the "drawer" slides out after you click on the shield, the "How does this work?" link says stuff only about the "Flag to Admins" part and nothing about the ๐ & ๐.)
BTW: does sorting by "Relevance" take ๐ & ๐ ratings into account?
I don't remember the specific values, but the ๐๐ป is a mod multiplier for the normal reactions when it comes to the feed algorithm - I think it's something like +5 or +10? Meanwhile ๐๐ป lowers priority in the feed by a similar multiplier.
I think the ๐๐ป๐๐ป being hidden behind the Mod drawer is to make using them more intentional - from the mod emails I don't believe the intent is that they would be used on every/all post (but someone from Dev would have to confirm this).
As for the shape - I think Shields and Stars a fairly common icon for moderators from what I've seen, kind of like a "Shielding the community from bad faith users" type vibe.
Part of my point is that ๐๐ป & ๐๐ป should be available to everyone to rate articles so the community will help de-prioritize low-quality ones.
Personally I disagree because I think having a discussion in the comments about anything that might be inaccurate is more helpful than dropping a ๐๐ป and leaving the post by creating an opportunity for the author to learn and update their post to be more accurate - which is part of what makes Dev different from forum sites like Reddit.
Like some others have stated, I'd also guess that most low-quality articles are simply ignored, so your hoped-for discussion in the comments doesn't happen in practice. Low-quality articles aren't low-quality only because of inaccuracies. IMHO, what makes an article low-quality even if 100% accurate includes one or more of the following:
Meanwhile, the price that everyone has to pay is seeing the low-quality articles because they're not de-prioritized.
Perhaps to get the best of both worlds, the ๐ icon could, once clicked, pop-up a small dialog with a text box where someone could comment why they're downvoting the article. Perhaps the comment could be mandatory; perhaps it could have an "anonymous" checkbox โ all up for discussion.
Great post
Thanks Ben - I appreciate that this is a difficult thing from the team's side, and that there might not even be a "Real" solution ๐
I feel good about our approaches right now to better deal with this. The biggest problems arising from our team's own actions is mostly around a few years of inaction on the subject โ however I'm not sure we would have clearly landed in a better place here.
I think we're taking some good steps now, and things are improving โ but it's hard to say that definitively and convincingly.
Your approach is logical and clear, it made me a bit less angry at the situation. Thanks for that.
Thanks for mentioning my post! And awesome content, too. I agree. Very interesting.
Fantastic post!
I'm not really sure what you wanted to say through this article. I guess that you feel there is too much "low-effort" content on dev.to?
I mean this is always the case when something starts getting popular. It's like having spam in your application. It's annoying - but it's a sign that things are growing.
I think there is no direct issue with "low-effort" content on dev.to. The issue is that low-effort content gets promoted through the algorithm and makes it to the top of all of our feeds. I wouldn't mind listicles one bit if my feed was filled with other, "higher effort" articles. Unfortunately, I find myself having to comb through 5-10 listicles at a time to find something I actually want to read.
Hmm I didn't notice it being that bad, but I don't follow too many people though. Might be also because it's EOY, and people like making lists at the end of the year ๐คท
Hopefully that is it, lol
Thanks for adding an optimistic view!
I didn't notice content length as a relevant metric here yet, and I also don't that the quality middle was missing, only that there tends to be an increasing number of low-quality content, whether short or long. Do you have any metrics about DEV content in this case? and how would we objectively measure quality or effort unless we knew every author and they were able to provide accurate work times information?
I think we can replace "effort" with "perceived quality", that is what is important in the end. No need for a precise scale, most reviewers (knowing the field at least a bit) can tell if the effort/quality is low, average or high, IMHO.
Yeah - I guess I should clarify a bit more that even in the podcast episode they call out that "Effort" isn't exactly the right word to use, but is a "close enough that you can get what I'm going for" word ๐
I think the difference is that in housing, people buy the cheap stuff because it's all they can afford.
On a forum like DEV, any user has a finite amount of time and attention but can choose to read things that interest them regardless of what they are.
The fact that low-effort churn tends to get a lot of reactions compared to the more substantial stuff implies that either it's all that people are seeing (because it's flooded the feed) or that people actually do get something out of it, every time. It's difficult for me to understand that position, because oftentimes listicles are flat-out wrong - they've chosen to copy-paste their content from existing poor sources, for example - but the fact remains, people do interact with them. Hearts, comments, even bookmarks for the 100th "12 react tools every developer must know" in a row.
In housing, people could see a street with 99 terrible homes and 1 mansion, but the problem with the analogy is that on here, they're all the same price, and if someone buys the mansion... you can too.
The first point here that you called out is why I believe there is not the problem you call out in the second highlight - users either have not enough time to find and read the longer articles that they might want to read or prefer to read a few quicker articles over a single slightly longer medium/middle article.
When it comes to reacting and rewarding the short articles, reading multiple listicles all titled "12 react tools" can be helpful, especially if you're new to the topic. Seeing the same recommendation over multiple lists might be a good indicator that that specific tool is worth looking into.
But of course this is only good when you're in the Short + High Effort intersection. As you mention here, the biggest issue with Short + Low Effort is that the content is made without any fact checking/thought (which I agree with - this just wasn't the point I wanted to target with my post ๐):
If users don't want listicles, then stop clicking on them. They appear because they work. Great click bait. I once suggested, and this is a decade ago now, that all my lesson plans (I was teaching dev bootcamps) would have titles begining with a number: 3 best ways to write a loop; 5 principles of SOLID; 3 key landmark roles.
I was joking, but it would have worked, sadly.
Frankly, I don't care whether articles are short or long. I am saddened that so few people appear capable of reading these days. I read hundreds of pages a day, but then that material is not technical. On technical stuff, I scan like everyone else. There isn't enough time, even given my "just-in-time" approach.
What matters to me is that the information is both correct and useful. What drives me to apoplexy (and almost being banned from Dev.to) is people writing articles just spewing out whatever they "figured out themselves" without checking whether anyone else had ever tried that and, more importantly, whether it worked. Or is, in fact, a terrible idea.
Newbies have a hard enough time already without being sent down the wrong paths, or encouraged to rush down rabbit holes. Articles such as "33 things every great JavaScript developer must know" enrage me. IIRC, I knew all but two by virtue of having been around since Adam, but only maybe a third of them were useful to me. Why would I waste my brain space (pretty limited these days) and time/effort/money to learn crap that I will never need?
Which makes articles like that one terrible advice, and yet some of the biggest names in the business retweet them. I suspect that we're all rushing (to our doom?) so quickly that no one has time to stop and think for a moment about what we're actually doing. We might want to rethink that approach.