Beefy Boxes and Bandwidth Generously Provided by pair Networks
"be consistent"

Categorized Votes

by ivory (Pilgrim)
on Jul 29, 2000 at 02:50 UTC ( [id://25035] : monkdiscuss . print w/replies, xml ) Need Help??

I know that to implement what I am about to suggest would be a major departure from the status quo, but I think it might be helpful. I just wrote a reply concerning voting and I started thinking that perhaps I might be on to something. Consider that currently people vote for or against a post without giving any idea why they are voting for or against. I've read a few different threads where people explain that they all have different reasons for voting for or against a post. Because of this, I am thinking that it might be helpful to have categorized voting.

Each post would have seperate scores for accuracy, helpfulness, etc. It would help newcomers feel more comfortable using certain posts as references, even if they themselves are not sure of the solution's accuracy. If this type of voting were adpoted, we could have accuracy ratings for authors (for example, if some user never wrote accurate posts, their rating would be low, and newcomers would know to stay away from that users posts when looking for information). I would recommend that these accuracy votes be limited to monks of a certain level, as I often read that lower level monks vote randomly to increase their XP/level. Anyway, just a thought...


Replies are listed 'Best First'.
(Ozymandias) RE: Categorized Votes
by Ozymandias (Hermit) on Jul 29, 2000 at 03:16 UTC
    The problem with this suggestion is that it simply formalizes something many of us have been working very hard to avoid; that is, votes being cast for individuals rather than posts. Generally speaking, it quickly becomes clear which monks are the most helpful/accurate/insightful, and I don't believe a system such as the one you are proposing would actually help that very much.

    There are a lot of changes that many different monks have thrown about as being solutions for various problems; the inflation of vote counts, for example, or the apparent ease with which relatively new and inexperienced monks are advancing their level of experience. Before we rush into implementing any of them, I'd like to see a lot more discussion of the possibilities and their expected and hoped for effects. Like the old carpenters' rule of measure twice, cut once, let's be sure of what we're doing and what we hope to accomplish before we make too many changes.

    - email Ozymandias
      I don't really agree that it is easy to figure out which monks are the most accurate, especially for newcomers. You have to think of this from the point of view of someone who is trying to learn the language. If you're really looking for answers to a question you don't know the answer to, it is difficult to weed through the different solutions and find the right one. I understand that voting for a specific person might present a problem, but I think that accuracy votes would help more than they would hurt.


        And this mythical new user will be able to figure out what those votes mean, and what good ranges are, and how they were derived?

        Look around you. Reading only two or three days worth of replies to Seekers of Perl Wisdom will reveal several people who are well above average in their answers. That is already available and apparent for anyone with the foresight to look; I doubt that it would be better to add to the complexity of the information presented about each monk. Where would you disply these ranks? On the home nodes? New users, on average, won't even read the site docs before asking a question, but you want them to find each user's home node and compare their scores on a series of scales that they don't know and won't be familiar with?

        - email Ozymandias
Re: Categorized Votes (athomason)
by athomason (Curate) on Jul 29, 2000 at 05:37 UTC
    I have to agree with Ozymandias that such a Slashdot-style voting system wouldn't benefit the site in any noticable way. My reasoning is different though; with the way the system currently stands, a post's reputation has next to no effect upon who reads it or how someone regards it, since you can't see the reputation until after you've voted for it yourself. The exceptions, naturally, are the Best Nodes and Worst Nodes pages, which let you jump the gun. Newbies especially, the folks without many votes and those most likely to learn from others' evaluation of a post's worth, don't ever get to see the reputation of the vast majority of nodes. Without seeing the reputation for all answers to a question, an unexperienced user would be unable to determine which was the best. Categorizing votes wouldn't help newbies at all, IMO, since it would just be a complication of the current system which doesn't help newbies at all.

    Of course, this is all by design, for better or for worse. From Voting/Experience System, "Now you're probably wondering what are levels and experience good for? Well right now it's just about status, but soon higher level users will be able to approve, cool, and delete writeups." The later powers haven't yet come to the masses (at least, not Abbots and below), so voting is valuable only for rating the poster, not the post.

    I've advocated in the past allowing a "bye" vote, which would allow you to forfeit your voting chance on a node in exchange for seeing its reputation. This would be useful for the (many, in my experience) times when you don't like a node enough to vote it up but certainly don't want to vote it down. I'd often like to know what other Monks think of such middle-of-the-pack nodes, despite a wavering opinion myself. Newbies especially could make good use of a such an extension. A person allowed to view the reputations for nodes about unfamiliar topics would inherently familiarize himself more with the more correct approaches. I know there are many holes in my Perl knowledge which would benefit from such a system.

    Back to the proposal at hand then, until a user can make a fair comparison of the reputations of different replies, requiring categorization of a vote would further clutter the system without any payoff.

      Immediate feedback does suffer, but errors have a way of being pointed out and corrected, after a short period of time. There are a handful of pedantic and extremely knowledgeable folks around these parts who are quick to point out better ways to do things. The rest of us point out other ways to do things, sometimes, in the guise of Perl golf or exploring weird tricks.

      If new posters have the expectation that they'll get a good answer within 12 to 18 hours, I think they'll do okay. (If they grab the first answer, they may be in for a bit of a shock -- but on the whole, most of the answers here are good.)

        chromatic said: on the whole, most of the answers here are good.

        Often (usually?) led by you. Both in speed-of-response and overall quality, chromatic often leads the rest. Though I often find myself a bit frustrated that a question I would love to answer has already been (yet once again) answered before I could even read it, I applaud you and the many others who contribute selflessly to this community.

        Brainbench 'Most Valuable Professional' for Perl

RE: Categorized Votes
by ivory (Pilgrim) on Jul 30, 2000 at 03:48 UTC
    Well, I understand what you all are saying, but consider this from my (or a newbie's) perspective. I sometimes read threads about subjects that have not come up for me yet. Sometimes they are really interesting and I see the potential for just such a situation to come up down the line...I like to make a note of these things, if I get a chance.

    The problem is that sometimes I don't get to check the thread all the time to make sure the information I made a note of is accurate. Sometimes as the thread gets longer, and people get pickier with their advice/solutions, it gets terribly frusterating and difficult to tell what is going to work better, what isn't going to work at all, etc. I just thought that an accuracy rating might help in situations like that. Consider not the person who asked the question, oviously they are working on the problem and can try out all the replies, but rather someone making a note of the solutions for later use.


RE: Categorized Votes
by royalanjr (Chaplain) on Jul 31, 2000 at 19:08 UTC
    I can understand Ivory's (and other newbie's) concerns. I mean hey, we were all newbies once.
    I do have to agree with the concensus that adding the accuracy votes would complicate and confuse matters. There is also the fact that people might vote a node accurate when in truth it is not. With my limit of Perl knowledge at the moment, I must confess that such a vote would be a little dangerous in my hands (as well as many others starting up the learning curve).

    I think the best thing is a small piece of advice I have heard applied to newsgroups: lurk for a few days, learn the "climate" of the group. If you apply this to PerlMonks, you will quickly learn whos posts to give more consideration to. Noone knows everything, and certain people may have their strengths and weaknesses within Perl. This too, will become apparant with a little judicial lurking and reading by a newbie.

    A little patience, reading the FAQ's etc, and a lot of reading of posts will enlighten a newbie more than any accuracy vote or reputation score I think.

    Roy Alan

    "For courting so slowly, you have lost this fair maiden / so begone now, for you will never enjoy her" -- old Irish Folk Song
RE: Categorized Votes
by splinky (Hermit) on Jul 30, 2000 at 20:08 UTC
    First off, I agree with the other experienced monks regarding categorized voting. However, I also sympathize with Ivory and the newbies in their quest for the GOOD information.

    One thing I think would help newbies out is if, until you earn your first few votes, you were able to see the reps of all the nodes and replies. As I understand it, the purpose of not showing the reps of the nodes is to not skew somebody's voting habits. But if they haven't earned any votes yet, that precaution hardly seems necessary.


      Hmmm. I wonder if people who are NOT yet allowed to vote, but ARE logged in, should be allowed to see the rep on articles. This lets the newbies get a feel, and keeps people from coming in as an AM to check article reps before voting.


      e-mail jcwren
        This lets the newbies get a feel, and keeps people from coming in as an AM to check article reps before voting.

        <devil's advocate>Not really: it would be trivial to create a second user and do the same thing. I can't imagine anyone so insecure in their voting practices to need to see reputations before voting, though, so I doubt it would cause any real problems. I would just like some way to see reputations of nodes I didn't want to vote on.</devil's advocate>