Timeline for A/B testing the sticky vote controls
Current License: CC BY-SA 3.0
23 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Mar 20, 2017 at 10:31 | history | edited | CommunityBot |
replaced http://meta.stackexchange.com/ with https://meta.stackexchange.com/
|
|
| Jun 9, 2015 at 13:21 | comment | added | FreeMan | I think it's a nice feature, and the not-scrolling aspect of it doesn't really bother me. Obviously, it bothers others, so I think it should be switchable. It is very convenient for voting on a long answer. I feel that having the action buttons at the bottom is a great addition to any blog/post type of site - don't make me scroll back to the top just to do something I wouldn't normally do until I've reached the bottom. | |
| Jun 8, 2015 at 19:06 | comment | added | Brian M. Scott | @KRyan: Serves no purpose? No, it simply doesn’t work as well as a naïve user might suppose – or, apparently, as well as you think (or perhaps would like to think) that it does. In Math.SE the problem is exacerbated by the fact that a question can have a variety of quite different answers that are good (i.e., useful) to different segments of the potential readership and cannot reasonably be expected to get comparable vote counts. | |
| Jun 8, 2015 at 19:00 | comment | added | KRyan | @BrianM.Scott If relying primarily on vote-counts to judge answers to material you are unfamiliar with is a very bad idea on any particular SE site, that site is broken and has serious underlying problems that need to be addressed. The entire point is that the highest-rated answer (and, potentially, other similarly-highly-rated answers) are the best; if that's not true, then the entire system serves no purpose. | |
| Jun 8, 2015 at 18:58 | comment | added | Brian M. Scott | @KRyan: You mean how the underlying mechanisms of SE are presumed to contribute to its content-quality. That they make some positive contribution is not really in doubt, but the extent of that contribution is not clear and appears to vary across sites. And contrary to what a naïve user might naturally suppose, relying primarily on vote-counts to judge answers is a very bad idea. | |
| Jun 8, 2015 at 15:15 | comment | added | KRyan | @aroth Certainly true, but 56% (current voting) dislike in anonymous polling that may include people who haven't tried it is not the same as confirmation that they are "sacrificing UX quality" in the first place. Ultimately, this feature's worth is a question of how it affects content-quality, but that is very difficult to directly measure since so many other variables affect it. I was mostly responding to knee-jerk reactions and failures to understand how the underlying mechanisms of SE contribute to it's content-quality. | |
| Jun 8, 2015 at 0:52 | comment | added | aroth | @KRyan - The issue isn't that one-dimensional. Sacrificing UX quality for an increase in voting quantity/quality is a risky proposition at best. For the SE system to work it also needs to provide an experience that makes users want to keep using it. Ignoring that aspect simply because a change is seen as having a positive impact on voting quality when viewed in a vacuum largely misses the point. Nothing happens in a vacuum, and a pleasant, user-centric UX is just as important to sustained content-quality as voting quality/quantity. | |
| Jun 7, 2015 at 23:27 | comment | added | Brian M. Scott | @quid: Very mild concern, and not in all quarters. | |
| Jun 7, 2015 at 23:27 | comment | added | Brian M. Scott | @Archonic: There obviously is a downside to it, given the number of people who dislike it intensely. They don’t disappear just because you disagree with them. | |
| Jun 7, 2015 at 21:17 | comment | added | Archonic | While it sticks around, it's taking up room that is otherwise unused and anyone saying "it's distracting" must be very easily distracted. It offers a small benefit, but there's no downside to it. Ship it! | |
| Jun 7, 2015 at 14:06 | comment | added | quid | For the record: a certain lack of votes, relative to earlier times, is rather perceived as a reason for mild concern on Mathematics, as can be seen from several meta-posts there. | |
| Jun 7, 2015 at 0:41 | comment | added | Brian M. Scott | @KRyan: I do indeed have a number of fundamental disagreements with the underlying philosophy. And your final sentence is at best an exaggeration: I don’t think that it works especially well as intended. Rather, I use it faute de mieux, because I enjoy helping people with mathematics, and to the best of my knowledge it’s the only site available that has a decent interface for mathematics. | |
| Jun 6, 2015 at 23:21 | comment | added | KRyan | @BrianM.Scott then you fundamentally disagree with the foundation of the site, and I wonder why you use it. Or I would, were it not obvious: you use it because it works. And whether you think so or not, it works because of voting. | |
| Jun 6, 2015 at 21:42 | comment | added | Brian M. Scott | @KRyan: There is no reason to sort answers by vote count unless one thinks that the count is meaningful. In my experience this is less than clear. In particular, I simply do not agree that voting does all that much to identify the best answers. | |
| Jun 6, 2015 at 21:19 | comment | added | KRyan | @BrianM.Scott The vote counts aren't too important; the voting is. The voting is how SE applies a sorting algorithm on answers, and why SE is a system that allows the quick, reliable, and consistent finding of answers to one's questions. Voting is how SE avoids noise and attempts to ensure that the best answers can always be found immediately. Without voting, that doesn't happen, and SE is no better at Q&A than a discussion forum – and discussion forums are not great places for Q&A. More voting improves the likelihood that the SE system will correctly pick out the best answer. | |
| Jun 6, 2015 at 20:50 | comment | added | Brian M. Scott | @KRyan: I disagree with the underlying premise. On the site that I know best (Math) I find the voting on questions virtually meaningless and the voting on answers less than useful, and I see no reason to think that more votes would improve matters. The answers themselves and the comments under them are vastly more useful than the vote counts. | |
| Jun 6, 2015 at 19:28 | comment | added | Ismael Miguel | @KRyan Thanks a lot. Yes, I didn't read the entire page. I hope more users see that link. | |
| Jun 6, 2015 at 19:21 | comment | added | KRyan | @IsmaelMiguel That one person, or sixty-four people, hate it, would be irrelevant if it increases the quantity and/or quality of voting on this site, which is the purpose of the experiment, and would be the purpose of the feature if it is implemented for real. Improving voting improves the site, for millions of people, so no one's personal opinion one way or the other matters much at all compared to that. There will be userscripts to deactivate it if it works but you hate it, there are already userscripts to activate it if it doesn't work but someone wants it, and so on. | |
| Jun 6, 2015 at 19:15 | comment | added | Ismael Miguel | @KRyan They can be from people who would hate such ad-like feature. Like me. I have it and I wish I didn't. | |
| Jun 6, 2015 at 19:07 | comment | added | KRyan | @IsmaelMiguel Actually, it really doesn't speak for much of anything, because those are not the numbers they are interested in. They are interested in whether or not it increases voting. That is the goal of the experiment. It is those numbers that will speak to the success or failure of the experiment, not voting here. These votes aren't even required to be made by people who have even gotten to try it yet. | |
| Jun 6, 2015 at 18:34 | comment | added | Ismael Miguel | @JarrodDixon Currently, the scroll is 57 to "I don't like. If it goes ahead, I hope I can disable." vs 44 "Nice feature.". I guess this speaks for itself. | |
| Jun 6, 2015 at 6:08 | comment | added | Jarrod Dixon | I agree about the "not scrolling vote controls on your own posts," but I can't change it without invalidating the current experiment. Based on the results, we might run another test where the controls always move. | |
| Jun 6, 2015 at 6:05 | history | answered | Patrick Hofman | CC BY-SA 3.0 |