I was just thinking: this site is the result of splitting off from overcomingbias.com earlier this year. With its new format and functionality, comments and posts get ratings. But all of Eliezer Yudkowsky’s posts from before the split don’t have ratings comparable to more recent posts, because that would require people to go back through the old posts mod them up.
Some people have done so, but not enough that their ratings accurately compare with more recent top-level posts.
I suggest that everyone take the time to go back to the Eliezer_Yudkowsky top-level posts from before ~February ’09, and vote up the ones you remember as being particularly good. If you wish, also mention what you voted on in this thread and why.
In order to avoid nudging anyone toward my picks, I’ll wait a while and then say them.
If you need some help jogging your memory, here are some threads where people discussed highlights from when LW was overcomingbias.com: one, two, three.
Instead of vaguely asking people to go read some, why couldn’t we do something more concrete? What we know gets traffic, and what everyone knows that everyone knows gets traffic (so you aren’t pouring the water of your comments onto the sands of an abandoned page), is being on the front page. Instant traffic, entree into RSS feeds, etc.
Why not every 2 days without a fresh front page post, for example, automatically post the next old EY article? (I suggested this back when EY’s old articles were being imported, but my suggestion was unjustly neglected, I felt; perhaps the need has become more apparent since.)
Those are good ideas too. But just to clarify, I wasn’t asking people to read the whole archive, hoping to stumble upon something they remember as being good and modding it up. I was asking that if you already remember certain posts as being good (or find them in a quick perusal of the links I gave), take a moment to mod them up so their rating more accurately reflects their quality in comparison to more recent posts.
Unfortunately, all the non-fiction ones I remember seem to have prerequisites to really understand and grapple with them—prerequisites published before them!
That’s not really needed, I think. Since a simple way to do that is to just notice which ones you tend to reference/search for most, and simply upvote as you happen to have need of them, right?
The purpose is to make the ratings of the old posts more comparable to the new ones, which can only happen when they get as much attention as new ones get, which requires a deliberate effort on the part of regular visitors to make sure their appreciation of older posts is reflected in an upvote.
And upvoting shouldn’t have much to do with when you happen to have need of them. If you like a post and want to refer back to it, that’s what the “save” feature is for. Or bookmark it.
Dunno about the −2. I didn’t vote you down. And I meant “those articles that I tend to reference most, that I tend to find most useful to link to to explain a concept are probably the ones most worth voting up”
What’s going to happen, regardless of how people rate, is that people will read the articles that get lots of links. People who come directly from google will read what has lots of links, whether internal or external. People who become regulars here will read what gets linked to from contemporary LW articles. Some small number of new regulars will use the ratings to decide what to read, but most of them won’t do that until after they read a bunch that were linked recently.
I’m not sure there’s a point to modding up old posts. While threaded comments on old posts might benefit from moderation to help drive-by viewers, I don’t see how ratings on old posts are beneficial to anyone (aside from trying to overflow EY’s karma).
Well, just to give an example, if someone comes in and sees Engines of Cognition, regarded by many here as one of the best posts, they may be misled by its rating of 3, compared to EY’s typical 30+ rating for newer posts which, while good, are not the classic must-read that Engines of Cognition is.
(I don’t see much reason to worry about EY’s karma going up; he’s past “karma escape velocity” :-P )
Though I guess there are easier solutions, like picking the ten most popular posts as per the threads I linked, and give them each 30 points but not add them to EY’s karma. Or, put an explanation about the site’s history on anything here before Feb. 09.
I was just thinking: this site is the result of splitting off from overcomingbias.com earlier this year. With its new format and functionality, comments and posts get ratings. But all of Eliezer Yudkowsky’s posts from before the split don’t have ratings comparable to more recent posts, because that would require people to go back through the old posts mod them up.
Some people have done so, but not enough that their ratings accurately compare with more recent top-level posts.
I suggest that everyone take the time to go back to the Eliezer_Yudkowsky top-level posts from before ~February ’09, and vote up the ones you remember as being particularly good. If you wish, also mention what you voted on in this thread and why.
In order to avoid nudging anyone toward my picks, I’ll wait a while and then say them.
If you need some help jogging your memory, here are some threads where people discussed highlights from when LW was overcomingbias.com: one, two, three.
Instead of vaguely asking people to go read some, why couldn’t we do something more concrete? What we know gets traffic, and what everyone knows that everyone knows gets traffic (so you aren’t pouring the water of your comments onto the sands of an abandoned page), is being on the front page. Instant traffic, entree into RSS feeds, etc.
Why not every 2 days without a fresh front page post, for example, automatically post the next old EY article? (I suggested this back when EY’s old articles were being imported, but my suggestion was unjustly neglected, I felt; perhaps the need has become more apparent since.)
Those are good ideas too. But just to clarify, I wasn’t asking people to read the whole archive, hoping to stumble upon something they remember as being good and modding it up. I was asking that if you already remember certain posts as being good (or find them in a quick perusal of the links I gave), take a moment to mod them up so their rating more accurately reflects their quality in comparison to more recent posts.
Unfortunately, all the non-fiction ones I remember seem to have prerequisites to really understand and grapple with them—prerequisites published before them!
That’s not really needed, I think. Since a simple way to do that is to just notice which ones you tend to reference/search for most, and simply upvote as you happen to have need of them, right?
The purpose is to make the ratings of the old posts more comparable to the new ones, which can only happen when they get as much attention as new ones get, which requires a deliberate effort on the part of regular visitors to make sure their appreciation of older posts is reflected in an upvote.
And upvoting shouldn’t have much to do with when you happen to have need of them. If you like a post and want to refer back to it, that’s what the “save” feature is for. Or bookmark it.
Any reason I got −2 for this suggestion btw?
Dunno about the −2. I didn’t vote you down. And I meant “those articles that I tend to reference most, that I tend to find most useful to link to to explain a concept are probably the ones most worth voting up”
But I guess for consistency, you may be right.
What’s going to happen, regardless of how people rate, is that people will read the articles that get lots of links. People who come directly from google will read what has lots of links, whether internal or external. People who become regulars here will read what gets linked to from contemporary LW articles. Some small number of new regulars will use the ratings to decide what to read, but most of them won’t do that until after they read a bunch that were linked recently.
I’m not sure there’s a point to modding up old posts. While threaded comments on old posts might benefit from moderation to help drive-by viewers, I don’t see how ratings on old posts are beneficial to anyone (aside from trying to overflow EY’s karma).
Well, just to give an example, if someone comes in and sees Engines of Cognition, regarded by many here as one of the best posts, they may be misled by its rating of 3, compared to EY’s typical 30+ rating for newer posts which, while good, are not the classic must-read that Engines of Cognition is.
(I don’t see much reason to worry about EY’s karma going up; he’s past “karma escape velocity” :-P )
Though I guess there are easier solutions, like picking the ten most popular posts as per the threads I linked, and give them each 30 points but not add them to EY’s karma. Or, put an explanation about the site’s history on anything here before Feb. 09.