Skip to main content
13 events
when toggle format what by license comment
May 23, 2017 at 12:39 history edited CommunityBot
replaced http://stackoverflow.com/ with https://stackoverflow.com/
Jul 22, 2010 at 18:02 history migrated from meta.stackexchange.com (revisions)
Mar 29, 2010 at 10:59 comment added Jonik @George: Hmm, I don't understand your last point at all, especially that you try to screw up polls on purpose to prove something. Over time and with lots of users things will work ok, because most people do not want to game the system or screw it up. It's a form of en.wikipedia.org/wiki/Approval_voting (except not just one winner), and nobody is saying anything about accurate.
Mar 29, 2010 at 10:53 comment added Jonik Found a great example of a well-implemented poll on SF: serverfault.com/questions/3780/… It has a combined list which links to answers with more details. (Also pls note that it is not locked or closed :-P)
Mar 28, 2010 at 16:28 comment added George Stocker The problem with Polls Jonik is that it forces everyone to vote up their favorite. The first time someone votes up or down randomly or votes not according to your preconceptions, it screws up the poll. In fact, I do this from time to time, just because people somehow think that you can have a poll without system implemented rules and have it be accurate.
Mar 28, 2010 at 13:32 comment added Jonik @George: Regarding your example: as useful as the compiled list is, it would be better if it somehow also directly showed which of e.g. the ~30 "Language Agnostic" books the community considers the most readworthy and which perhaps have just an upvote or two. I like separate answers & the voting system because that "automatically" provides a dynamic, up-to-date list.
Mar 28, 2010 at 13:17 comment added Jonik @George: Ok, agreed, the result can be really useful if a good editor or few come by. (Creating a categorisation that actually works is often problematic though.) Still, the notion of popularity is (mostly) lost, and in polls such as the one I'm talking about, it's arguably the most important aspect. And even if you re-arranged such list manually to reflect changing scores (which would be silly), that wouldn't work for new entries, as new answers cannot be added if the question is locked (which is why I posted this meta question!).
Mar 28, 2010 at 12:44 comment added George Stocker @jonik Re: Mess: Not true, take this for example: stackoverflow.com/questions/391523/… Oh, and guess what? You can always link the answer in the compiled thread to its place on the thread, just so people can see the votes if they really want to. But if you're collecting information, it's superfluous.
Mar 27, 2010 at 19:55 comment added Jonik @Arjan, agreed. But about the second point: don't be so sure (that this is a problem). I've often seen recent answers (good ones, that is!) climb relatively quickly high up to the first page in these huge & long-lived SO/SU polls. (I can give examples if needed.)
Mar 27, 2010 at 19:45 comment added Jonik Yes, for many questions that approach would arguably be the best. However, I'm convinced that for "polls" like the one in question (another example: stackoverflow.com/questions/38210/… & all other must-have SU questions) using separate answers and the voting system is a far better way to keep it organised & up-to-date. As Arjan pointed out, with single answer we'd end up it a huge semi-randomly ordered list, i.e. mess.
Mar 27, 2010 at 19:05 comment added Arjan That way one would not be able to tell which is most popular. (But on the other hand: more recent answers probably have less votes than earlier answers, even if they would be more popular.)
Mar 27, 2010 at 18:50 comment added Ether This is what community wiki was originally intended to be, but no one seems to like editing an existing response. I guess they are too wedded to seeing their own name next to a particular answer.
Mar 27, 2010 at 18:06 history answered George Stocker CC BY-SA 2.5