last day (14 days later) » 

7:15 PM
-2
A: Is there any mechanism to eliminate the 'oldest answer gets more votes' bias?

user121330There are tons of ways to do this, none of which have been implemented by SE. If I was designing the site, you'd get a different version if you came in from a search engine (including the SE search tool, adorable as it is) (looking for answers) than if you were trawling / pruning SE (looking for ...

 
@Cai I think we're in agreement here - I search for questions all the time as well (though I do need my fix of answers from time to time) ;) Is the downvote because you only read the first half paragraph?
 
This doesn't solve the problems at all. In your site you're still showing the highest voted answers at the top, which means those will be read first, and therefore will get more votes. You've done nothing to solve the problem.
 
@Servy It's normalized to how long the answer has been up. It shows a measure of how many votes per time.
 
@user121330 So first off, if that's your solution, then you should actually focus on that in your answer, rather than just mentioning it offhandedly in a parenthetical of an otherwise unrelated answer. That said, it still doesn't remove the problem, as an old answer with lots of upvotes will still have more votes per time than a new answer with no votes. It would help a newer answer rise a bit faster, but it will struggle to rise at all because it'll struggle to get views because everyone is reading the older answer with tons of upvotes instead.
 
@Cai I'm obviously being unclear here. Seekers get to see the answer with the most votes per time on the site. Trawlers don't get to see any votes.
@Servy No, the trawlers don't get to see the votes and they have more power to change problematic answers.
@Cai The people who are looking for answers (if you come in from a search engine) are seekers, and you and I are examples of trawlers.
 
7:15 PM
@user121330 It wouldn't matter. People coming from google are going to upvote the top answer, which will be the old answer with the most votes, which will give it more votes, keeping it at the top. That's the same in your system as it is in SO now. If you're just going to have people edit other people's answers to change the content when they think it's wrong, unlike how SO handles editing now, then there's no need for multiple answers (or voting) in the first place; at that point you just have a wiki, like wikipedia.
 
@Servy, I'm not sure you see how normalization works. If I put up an answer and 2 people vote for it today, it's going at 2 votes per day which would be more than the 730 votes for a 1 year-old answer. If we compound it per week, that's 104 votes on a 1 year old answer, and if we compound it on a per month, that's 24 votes. Setting the right compounding time would be tricky, but we could easily set it up so that the newest answer was the first.
 
@user121330 You're assuming that 2 people read past the 730 vote answer and upvote it, rather than reading the 730 vote answer, because it's first, upvoting it, and not even reading the 0 vote answer underneath it. That's not even getting into the fact that an answer with 730 votes over a year is much more likely to be a good answer than an answer with 2 votes over the course of 1 day, so you're much more likely to end up showing people worse answers in the rare cases where the two systems actually differ in the first place.
 
@Servy I don't disagree. Perhaps the compounding factor is views rather than time. It really doesn't matter - SE already does this except that they compound over all time. They've chosen to prefer old answers.
@Servy Also, the trawlers don't get to see the vote preferences - those are just for people who are looking for answers, so we would expect that any flat wrong answer, highest voted or not, would be addressed by the trawlers.
 
@user121330 What's your point? That still doesn't result in a new answer posted later actually becoming the top answer. It would potentially make it a bit easier, but it doesn't remove the inherent problem that almost nobody is going to even read it in the first place, even with your changes, so it won't have much of a shot at getting above the other answer. Then there's the fact that even in the cases where it does happen, there's a very good chance that it wasn't appropriate and that the old answer that picked up a few pity votes isn't actually the better answer.
 
@Servy It sounds like you see the problem, and you've just given up. That's an option. If seekers get the highest ranked answer per page view first, and trawlers get answers in random order, you can't argue that there's less inertia for old answers. I'd argue that there's a better balance (while acknowledging the value of precedence).
 
7:15 PM
@user121330 Just because I recognize that your solution doesn't solve the problem doesn't mean nothing can be done, it just means that your solution doesn't solve it.
 
Sounds like you have a solution?
 
There have been numerous proposals over the years, several of which have at least some promise
it's not a new discussion, by any means
 
So you don't have a solution.
 
I don't have anything to add over previously proposed solutions
 
Where are those?
 
7:21 PM
I'm not really particularly interested in re-hashing them all again. You wrote your on proposal, it doesn't solve the problem, and comes with many problems, if you want to see others' solutions, you're more than welcome to do your research.
 
I really don't think you understand the proposal. Fine. Have a beautiful day.
 
Well you haven't actually explained how it results in more attention for a new post. I've explained how the behavior wouldn't change anything, and your response has simply been to demand that I come up with something better. You haven't given any reason as to how your proposal actually solves the problem.
if I'm not understanding something about it, fine, explain how your proposal solves the problem.
 

last day (14 days later) »