How the LW2.0 front page could be better at incentivizing good content
post by cousin_it · 2018-01-21T16:11:17.092Z · LW · GW · 14 comments(This should probably be in meta, but I'm not sure how to post to meta. Feel free to move if needed.)
Imagine tomorrow someone writes the best ever LW2.0 post and everyone upvotes it. How can front page visitors read that best and newest post? They must scroll down (no posts are visible above the fold) and then click (no posts are expanded by default).
Attention is a key scarce resource for incentivizing good content creation. Our current front page is an attention firehose pointed away from the best newest stuff. Where does it point instead? 1) Eliezer's sequences, which are finished, so we incentivize nothing. 2) SSC, which already gets a ton of visitors and where only Scott can post anyway, so we incentivize nothing.
Compare with how other websites do it. Reddit and HN put the best newest stuff at the top, though you have to click. Old LW used to put the best newest stuff at the top and expanded by default, so you could check the site first thing in the morning and get the awesomeness. SSC does the same.
Even worse, we have /daily which is more readable than the front page (you don't have to scroll). And getting a post into /daily doesn't require upvotes. The incentive gradient for writers is nowhere to be seen! On old LW, getting my posts promoted to Main felt special. On LW2.0, several of my posts were front paged and curated and it didn't feel special at all.
The best use of the attention firehose is incentivizing good content creation. We should make the front page work like a blog, with the best newest entries shown at the top and expanded. Move the sequences and SSC into a sidebar, like every other website does for good reason. Then we might finally enjoy some actual competition for the front page, which would be a good thing.
14 comments
Comments sorted by top scores.
comment by Said Achmiz (SaidAchmiz) · 2018-01-21T16:15:35.998Z · LW(p) · GW(p)
This is excellent and correct. I disagree with none of this, but do want to add an additional suggestion:
Also have a “best posts of the last time period” sidebar (where ‘time period’ might be a month). That will keep really good posts visible for longer, without requiring explicit “sequence” creation or any other such formal archiving. (Note that this is distinct from any ‘sticky’ or ‘featured’ post feature, which would not have such an effect.)
Replies from: Raemon↑ comment by Raemon · 2018-01-21T20:52:46.718Z · LW(p) · GW(p)
Also have a “best posts of the last time period” sidebar (where ‘time period’ might be a month). That will keep really good posts visible for longer, without requiring explicit “sequence” creation or any other such formal archiving. (Note that this is distinct from any ‘sticky’ or ‘featured’ post feature, which would not have such an effect.)
I like this quite a bit (not sure about exact implementation but the core idea seems quite solid, and seems good from a few different use-cases and philosophies about the front page)
comment by gwillen · 2018-01-21T16:48:38.760Z · LW(p) · GW(p)
Yes, 100%. I was confused and disheartened when the original LW removed all the posts from the frontpage. That UI decision felt to me like one of the steps along the road to death for LW, and I still don't understand why it was done. The posts are the heart of the site. Put the posts where people will see them.
Replies from: cousin_it↑ comment by cousin_it · 2018-01-21T23:58:12.277Z · LW(p) · GW(p)
Yeah.
I just reread the archives from 2011 and it's surprising that pretty much no one saw the issue the way I see it now. The change can be traced to this feature request by Patrick which got highly upvoted. But it only proposed having a welcome page somewhere, not making it the default! Then it became default and nobody objected, even though other aspects of the redesign got many comments (see e.g. here, or in these five posts by Matt). I guess the community had more momentum back then, and Eliezer had been chronically neglecting to promote enough stuff to Main, so the loss of focus on Main wasn't seen as important. But as you say, that decision probably cost us in the end.
comment by Qiaochu_Yuan · 2018-01-22T02:33:46.843Z · LW(p) · GW(p)
Thanks for writing this. I also quite dislike the current front page. I still don't remember on the 1-second level the different between Frontpage and Curated or why I should care about it, and think the part at the top linking to the Sequences and SSC is way too big and should be half as tall, at most.
comment by habryka (habryka4) · 2018-01-21T20:59:50.584Z · LW(p) · GW(p)
Moved to meta.
Replies from: cousin_itcomment by Chris_Leong · 2018-01-21T22:21:03.327Z · LW(p) · GW(p)
I disagree. The problem is not that we are promoting people starting with the old classics, which tend to contain more core insights than whatever happens to have been posted most recently. The problem is that there is no way to collapse it.
You're right that attention incentivises content creation, but we are also custodians of this attention and we want to direct it towards whatever will most benefit our audience. We shouldn't just blindly follow the corporate model. Corporations want to maximise activity by getting people commenting on the new stuff, rather than maximise utility.
Besides, we already get a steady flow of submissions.
Replies from: Raemon, cousin_it↑ comment by Raemon · 2018-01-21T23:26:32.643Z · LW(p) · GW(p)
This framing of the problem seems right to me.
FYI, we're planning to make it so logged in users see a much shorter section, which is something like "sequences you are currently reading or maybe should read", which is more like 1.5 inches tall rather than a full page. I think this will solve most of the problem.
Our current mindset is something like "be more like youtube", in the sense that new content is very visible, but a lot of attention is directed in various ways to the oldest, best content.
↑ comment by cousin_it · 2018-01-21T23:16:43.883Z · LW(p) · GW(p)
Yes! It's a good thing my proposal is trying to improve quality of posts, not maximize activity.
Replies from: Raemon↑ comment by Raemon · 2018-01-21T23:38:06.076Z · LW(p) · GW(p)
This makes sense and I think I know what you mean now, but it wasn't as obvious in the original post.
The underlying goal Habryka is working towards is incentivizing the building of important intellectual knowledge, which can be used concretely in the external world. This is similar to "improve quality of posts", but subtly different. The four sub-things that need incentivizing are:
- Newcomers reading and learning foundational build blocks.
- Veterans coming to check on important new updates to the frontiers of our knowledge that are relevant to them.
- People contributing posts and sequences that build on that knowledge, that we'll want to refer back to and build on yet-more in the coming years.
- People who either improving at thinking and writing, or refining ideas together, until they produce #2.
I probably have more to say, but for just noting that the frontpage has multiple jobs it's trying to accomplish. It could probably improve at all of them. I read your post as saying (at least when translated-into-my-models) something like "Our front page incentivizes #1 at the expense of #4, and #4 is a necessary step to get to #3, which is necessary to get #2."
If this seems radically different than the underlying goals you were pointing at, apologies, and could you explain differently?
Replies from: cousin_it↑ comment by cousin_it · 2018-01-22T00:39:41.805Z · LW(p) · GW(p)
Yes, that's a great explanation.
There's a small caveat though. You say the current front page incentivizes #1, but actually it just recommends people to do #1, there's no incentivizing. On the other hand, if you had a front page giving tasty attention to the best attempts to build on the sequences, that might incentivize all of #1-#4 better than our current front page. And focusing discussion on a few posts would make comments get more votes on average, increasing incentives for commenters as well. I think pre-2011 LW got all of those right, except the curation for Main wasn't responsive enough.
Replies from: Raemon↑ comment by Raemon · 2018-01-22T00:50:28.160Z · LW(p) · GW(p)
Agreed that incentivizing isn't the right word for what's happening with #1, but there's a related thing (maybe a superset that includes incentivizing, in some frames?) that also includes "nudges", or "making certain things convenient" which seems like an important piece of the puzzle.
comment by Thrasymachus · 2018-01-21T23:26:48.735Z · LW(p) · GW(p)
FWIW, I struggle to navigate the front page to look at good posts (I struggle to explain why - I think I found 'frontpage etc.' easier for earlier versions). What I do instead is look at the comments feed and click through to articles that way, which seems suboptimal, as lots of comments may not be a very precise indicator of quality.