Beefy Boxes and Bandwidth Generously Provided by pair Networks
Keep It Simple, Stupid

Basic voting statistics for Node Reputation

by radiantmatrix (Parson)
on Apr 06, 2005 at 18:28 UTC ( #445390=monkdiscuss: print w/replies, xml ) Need Help??

Every once in a while, I find that I end up with a node rated 0 or -1. This happens -- reputation, like XP, is just a number -- and I don't stress about it. However, I've always wondered about why I get a -1. Is it because very few people saw my node, and the one person who voted on it voted it down? Or, did many people vote on it with a net result of -1?

If I got one vote, and it was --, I would simply write it off as a fluke. On the other hand, if 90 people voted and 46 of them thought the node was poor, I can take some time to consider what I might have done to improve the node quality.

To this end, would it be possible to include a number-of-votes statistic next to the node reputaion? In other words, something like:

Reputation: 1 after 12 votes.

I think such a thing would be particularly useful to new posters, so that they get a better idea of what the community considers good, bad, and mediocre. There would also be some benefit to those of us who have been around a bit and want to improve the quality of our contributions. What do the rest of you think?

require General::Disclaimer;
s//2fde04abe76c036c9074586c1/; while(m/(.)/g){print substr(' ,JPacehklnorstu',hex($1),1)}

  • Comment on Basic voting statistics for Node Reputation

Replies are listed 'Best First'.
Re: Basic voting statistics for Node Reputation
by Limbic~Region (Chancellor) on Apr 06, 2005 at 19:44 UTC
    Variations of this theme have been raised several times in the past (see Node Tension for example). In my opinion, the best way to request features that have been discussed countless times before but not implemented is to become a devil and write the patch yourself.

    Cheers - L~R

      I've thought of this question before, and thought it would be a cool statistic, too: to know ++'s vs --'s of a node, although I would only think it a useful breakdown for one's own nodes, not ones written by others. And I've seen this answer ("join pmdev, write the patch!) before, too. After a brief Super Search, I found this node, pmdev user group, and am wondering - is this still the way to join pmdev?

      In some ways, it would be nice if we could see the code prior to trying to join pmdev. That way, we'd know (or have an inkling of an idea) that we'd be overwhelmed, and be much happier in the areas of the site that way may think need improving. But, it is neither my site nor my code. Seeing that it is what it is, this is just an observation, not a complaint ;-)

        Speaking completely non-authoratatively, I would suggest /msg'ing the pmdev group. There is a group inbox - several monks have become devils that way. WRT your comment about being able to see the source first - that is another issue that has been raised several times in the past (Super Search if you are interested).

        Cheers - L~R

        As far as I am aware, the Everything engine as used elsewhere for example, keeps and displays a separate count of ++ and --. So, there's probably not much dev work, just some detective work and merging.

        I'm Not Just Another Perl Hacker

      Except when its not implemented because the powers that be don't think its a good idea.

        Yes, making the presense of downvotes easier to notice and thus increasing the frequency of whines about downvotes would suck. The remaining code isn't hard to implement (most of the harder parts were done quite a while ago -- I think the updating of old nodes still needs to be done, and I'll do that after the new node cache is put into production).

        A tricky part is designing how/when to present the information to provide some benefit while avoiding most of the potential problems (as I discussed in more detail previously).

        - tye        

        Too bad. I think we shouldn't only see how many people voted for our nodes, but also who voted, and which vote they cast. That way, if a node gets downvoted, we can downvote one of theirs.

        Posted as AM because I care about my XP.

Re: Basic voting statistics for Node Reputation
by Joost (Canon) on Apr 06, 2005 at 18:44 UTC

      It isn't necessarily worse, that's my point. If only one person bothered to vote, and it was a downvote, that's not significant. So, it could be just that person misclicked, for all I know.

      However, if 44 people think the node is good, but 45 think it's bad, I can go "hm. there were some good things, and some bad things in my post", and consider what I said that led people to downvote.

      Think of it as "how much weight does this reputation carry". It's much like reading movie reviews. If you read just one review at random for a movie, and the critic says it's good, that's OK and all, but you have to decide "should I agree with that critic"? If 5 say it's great and 4 say it's bad, it's not about whether you trust one person or not. You can start to think "well, there must be some good and some bad things about it -- let's see what's good and bad" and form an opinion based on that.

      Likewise, with the additional data here. It's the difference between "wow, I pissed someone off, I wonder if my node sucked or she's just having a bad day" and "well, I obviously said some good things, but it looks like I said some bad things too" and try to reduce the latter.

      require General::Disclaimer;
      s//2fde04abe76c036c9074586c1/; while(m/(.)/g){print substr(' ,JPacehklnorstu',hex($1),1)}

        If only one person bothered to vote, and it was a downvote, that's not significant

        Not to add to your paranoia ... but that's not necessarily true. There's some monks I respect greatly and if they were the only downvote, I would go back and rethink.

        So my advice, as many others, is to not worry about your node rep and just worry about your rep. If a fellow monk takes the time to explain a downvote, then rethink. If a fellow monk does not explain his downvote then don't worry about it.

Re: Basic voting statistics for Node Reputation
by ambrus (Abbot) on Apr 06, 2005 at 19:33 UTC

    I don't have a clear opinion on whether such statistics would be good or bad. Anyway, if you look at the node reputation often, you can get an idea about the approximate number of votes from how much the reputation changes. It is easy to do this, because you can see all your nodereps for your recent nodes in Newest Nodes or its xml variant user nodes info xml generator.

Re: Basic voting statistics for Node Reputation
by husker (Chaplain) on Apr 07, 2005 at 13:58 UTC
    Would it be helpful to see the number of "views" for each node? Coupled with a vote tally, this would give you an idea of how many people liked your node (++ votes), disliked your node (-- votes), and how many were neutral about it (views but no votes).
Re: Basic voting statistics for Node Reputation
by artist (Parson) on Apr 07, 2005 at 15:40 UTC
    Just have stats for 'total votes' and problem solved

    It's an online community and be happy about the 'real' response about your question that might be useful in real world. (edited:)

Re: Basic voting statistics for Node Reputation
by doom (Deacon) on Apr 08, 2005 at 01:54 UTC
    I sympathize with your feeling that a negative score is a bit bewildering, but I don't know that additional stats would help that much. (Though, on the other hand, recently I got pounded down to -2 and then up to +3, and I probably wouldn't have a clue why I was up at +3 if I didn't understand it was an over-reaction to the original down votes...)

    Anyway, I think the real trouble is that you just don't know the reason people are voting the way they are. Does that -1 vote mean they objected to my techincal content? Or maybe it was my tone? Did they think I was drifting off topic? Maybe my writing was unclear and they just didn't understand what I was saying?

    The obvious (though not necessarily right) solution would be some sort of slashdot-like selection of canned explanations ("+1 insightful", "-1 overrated", etc.).

    The less obvious solution: just get rid of negative voting. Put the emphasis on the positive, and if someone disagrees make 'em post a reply explaining their disagreement.

Re: Basic voting statistics for Node Reputation
by tlm (Prior) on Apr 08, 2005 at 06:19 UTC

    I would like to offer a dissenting view: adding more rep-related info will have the unintended, and undesirable, effect of encouraging users to focus excessively on their nodes' reps. I think the system is already brilliantly tuned as it is in the amount of information it provides and to whom. I wouldn't mess with it.

    What I would like to see is the ability to cast a 0 vote that would enable the person casting the vote to see how a particular node is being received by the Monastery (i.e. the node's rep). It often happens that I think a node is mediocre (though not so bad as to deserve a downvote), but I am still curious to know whether other monks share my assessment. Then I'm faced with the unpleasant choice of either down-voting a node that I don't think is so bad that it deserves such treatment, or, alternatively, up-voting the node, which is even less deserved. Being able to cast a 0 vote would solve this problem. I'd be happy to forgo any probability of voting-dependent XP gain from voting neutral, which may be a desirable way to discourage such votes without outright ruling them out (as is the case now).

    the lowliest monk

Log In?

What's my password?
Create A New User
Node Status?
node history
Node Type: monkdiscuss [id://445390]
Approved by Joost
Front-paged by grinder
[Discipulus]: all is well.. flying home
[Eily]: if I had a flying home I would be rather concerned I think

How do I use this? | Other CB clients
Other Users?
Others meditating upon the Monastery: (8)
As of 2018-02-21 17:33 GMT
Find Nodes?
    Voting Booth?
    When it is dark outside I am happiest to see ...

    Results (285 votes). Check out past polls.