Beefy Boxes and Bandwidth Generously Provided by pair Networks
Your skill will accomplish
what the force of many cannot
 
PerlMonks  

Why do nodes with minimal value get upvoted most?

by Eyck (Priest)
on Dec 23, 2004 at 14:48 UTC ( #417101=monkdiscuss: print w/replies, xml ) Need Help??

I noticed, that nodes with informational value approaching 0 get upvoted surprisingly often.

This seems to punish, and thus ultimately remove from perlmonks data of any value, when someone asks question, and then the most popular answer is just rehash of some popular myth, or maybe copy-paste of perldoc perlfaq entry gets the most votes.

This seems a little like 'first-post' phenomenon, because those elaborate and extremely popular answers usually have nothing to do with question at hand.

Sometimes it looks like they were created by some simple bot - simple pattern matching catching phrases like 'XP','Java','Thread' in question, and then pasting some elaborate answer roughly related to those key-words ( but not necesserily to the question itself ).

UPDATE: Interestingly, this node got downvoted pretty bad (-17 at the moment), which is consistent with my previous observations.

This is sad, because you can't progress when you can't see your mistakes. And any time anyone notices some problem and tries to bring attention to it - bang, community comes.

Conversly, lesson learned here is that one should be secret and just exploit the weekneses.

  • Comment on Why do nodes with minimal value get upvoted most?

Replies are listed 'Best First'.
Re: Why do nodes with minimal value get upvoted most?
by gaal (Parson) on Dec 23, 2004 at 14:58 UTC
      This seems to punish, and thus ultimately remove from perlmonks data of any value

    Only if you assume that XP is what people come here for. And that XP not reflecting value is punishment.

    Let's put the question on its head: in addition to and apart from those nodes with "informational value approaching 0", are you content with the "value" of PerlMonks in general? Ignore XP for a moment: I'm asking for *your* subjective judgement. To me it seems that PM is a long way from having its value removed; and I haven't seen any evidence that it is lessening.

      PM is great.

      That's why I care.

      Flooding it with junk data lowers value of PM.

      For me, the value of PM lies in both questions and the answers.

      But when you have to sort through kilobytes of useless rehashing of old themes to find one or two lines of actual original thought... well...sorting through spam is painfull.

      When you've got nothing to say, you'd better provide a link instead of flooding PM with unoriginal thoughts.

        Something you seem not be taking into account is that it's not the people just posting short replies to anything; there's a constant influx of new users who are asking:

        • the same questions which have been answered here numerous times (and of course they didn't bother using the search box first)
        • questions that are as I mentioned below answered quite succinctly by the standard documentation (or another well known site / FAQ / text)
        • questions which indicate a lack of understanding of programming fundamentals (in as much as there are quite a number of non-programmers who stumble into (or are pushed :) using Perl) or Perl's limitations ("Why does it take 30 hours to search through this 2G file which I parsed into a hash on my computer with 64M or RAM?")

        And that's to say nothing of the hordes:

        • asking HTML / Javascript / PHP / miscellaneous web questions which have nothing to do with Perl (" . . . but I figured you guys would know anyhow")
        • attempting to get someone to write their homework assignment for them ("I have this friend who's trying to . . .")
        • expecting free support for some commercial product or service rather than bothering the people they bought it from ("I found this on a haxx0r site^W^W^W^W^W^Wbought this and was wondering . . .")
        • mistaking this for freshmeat.net / versiontracker.com / Google ("I need a PERL script to do . . .")
        • . . .
        A reply falls below the community's threshold of quality. You may see it by logging in.
        The signal to noise ratio of PM isn't bad; better than in most communities that I've seen. Not to say that it can't bear improving, of course.

        Since you distinguish questions and answers: most, but not all, of the noise (I wouldn't call it spam) is from people asking about simple things that they should have read in the docs first. But the de facto community response to this is accomodating, whether by substantive answers or by pointers to the docs. On the question side of things, it seems that if you don't like the noise, you're better off not fighting it. I'd suggest that you either thicken your sunglasses or adopt as your favorite forum (say) c.l.p.m, which is less tolerant to FAQs. Take it as a psychological given that you are unlikely to get people to change their behavior just by complaining about it.

        On the answer side of things, in general you're asking why people reply with myths; the solution to that riddle lies in the riddle itself. They reply with myths presumably because they believe they are true. This is the case in all communities. The question is how receptive the community is to people who call on the myth. In my humble opinion, PM isn't bad in that respect either. We've had dogma-challenging posts here, we've had contoversies here. Nothing is stopping you from writing a substantive Meditation about a something you think monks hold a wide misconception about. Don't jump to the meta-level yet, please: unless you think there's a *Perl* myth going on about, this whole discussion is academic. I am pretty sure that if you write up a well reasoned noncomformist post, it would get all the attention it deserves.

        Is Perlmonks a search engine that you passively ask questions and get answers in return? Is it not rather a forum for discourse? What is the worth of PM, if not as a facility to allow many people to benefit collectively from the learning experiences of each individual? Do people post misconceptions knowing ahead of time that they are wrong? If a popular misconception goes unposted and in turn unrefuted, how will awareness that it is a misconception spread?

        Makeshifts last the longest.

Re: Why do nodes with minimal value get upvoted most?
by Fletch (Bishop) on Dec 23, 2004 at 15:08 UTC

    Most of the time I paste a one line "perldoc -f foo" reply it's because the entirety of the answer is in that part of TFM. A large part of getting really productive in Perl is to learn how to use perldoc effectively and where to look for what kinds of information (so that eventually they too can completely answer questions with a succinct two word perldoc perlzorch).

    In fact I almost commented as much on the "Hey, let's rewrite perldoc" Discussion the other day. It's not that the documentation is lacking, it's that people don't know where to come in because it's so complete. If you're dissatisfied with Perl's documentation, take a look at what comes with Ruby (disclaimer: I \N{HEART} Ruby; I just think that the current state of the documentation is miles behind Perl's (due in part to it's relative immaturity I'm sure)).

    Update: Just to clarify: this is why I specifically said relative immaturity, and that statement was in regards to the Ruby documentation. The second edition Camel from which most of the core of perldoc is descended came out in 1996; the first major English Ruby documentation (in the form of the Pickaxe book) wasn't out until 2001. And while I also \N{HEART} Smalltalk and its class browser that Rdoc is modeled after, there are times and workflow patterns when perldoc -f foo beats the pants off start-browser-clickety-clickety-bring-up-find-pane-clickety-clickety.

    Another Update: In the process of mucking with Ruby On Rails (which is pretty spiffy, by the way) I've become aware of the ri command which comes with recent (1.8+, I believe) Ruby which functions like perldoc -f for Ruby classes and methods. They're definitely catching up (and Ruby's really growing on me, which maybe means I need to see a dermatologist (or a geologist)).

      The second edition Camel from which most of the core of perldoc is descended came out in 1996.
      That sounds like most of the core docs were copied from the Camel. That's not true. First of all, the authors of the various editions of the Camel also wrote large portions of the documentation. And a lot of what's in the Camel, be it the Camel II and the Camel III, appeared in the docs first.

        *feh*

        Yes Tom Christensen was largely responsible for both huge chunks of the Perl 5 perldoc as well as the various Camels. Much of what got expanded upon and written for the 2nd edition Camel got merged back into the perldoc after its publication (and has subsequently been further honed and improved by countless others). I was using the copyright dates of the two reference tomes five years apart to clarify what I meant by relative immaturity, not to produce further quibbles about the genealogy of perldoc.

      Immature? Ruby? From the Ruby website we can learn that Ruby was born in early 1993, and first released to the internet in 1995. To be more precise, Ruby was born 20 days after the release of perl4.036. That's right. Before the first alpha release of perl5. Ruby hit the internet around the time the first betas of perl5.002 were released. 1995 was also the year Java was announced.

      How long do languages stay "immature" and "new"?

        Ruby was born 20 days after the release of perl4.036.

        Yes, I think you've just proved Fletch's point quite nicely. The point was Ruby's relative immaturity - relative to Perl, of course. The fact that Ruby was born when Perl was already in it's 4th incarnation means that Ruby is immature relative to Perl.

        --
        edan

        I have no personal opinion on the maturity of Ruby but maturity is not only relative to age. There are many potential indicators of maturity; some may matter to you and some may not.
        • Stability
        • Number of bugs in the stable release
        • Number of installations
        • Size of its community
        • Number of releases
        • Does it follow a formalized release process?
        • Does it follow a formalized change control process?
        • Does it make backwards compatability a priority?
        • Age
        • Insert whatever matters to you here...
        • etc...
Re: Why do nodes with minimal value get upvoted most?
by edan (Curate) on Dec 23, 2004 at 15:00 UTC

    First, how does upvoting "punish" or "remove" anything from Perl Monks? Only downvoting does this, IMHO.

    Next, you seem to be saying that an answer that consists of a link or pointer to RTFM (perldoc/perlfaq) is somehow lacking in informational value. Why do you think this? If the answer to the question can be found there, that's where the poster should be looking, and they should be told so.

    Do you have some examples of posts that you think exhibit the problems you mention here? Please, link them in, so we can see too.

    Update: I went back to read your recent thread Ansi Perl, and I assume that your complaints here are about replies that you received there. I don't see why you have a problem with people upvoting replies that you don't happen to agree with - obviously it means that other people do happen to agree with them. How does this reflect on their "informational value" one way or the other?

    --
    edan

      If Ansi Perl is the thread the OP has the problem with, it was front-paged. From what I've seen, front-paged nodes and their children gain points out of all proportion to their actual value.

      ANSI Perl is just one example, I saw this in many other nodes,

      This is of course very visible in one's own nodes, because then I understand the question best, researched the subject a bit, and know when the answer I got is sincere and informational, or when it's just regurgitated shibboleth formatted nicely and upvoted like crazy.

      As to 'punish' thing - it's easy to notice that nodes usually do get upvoted, thus neutral value for node is a bit closer to +5. Anything lower is 'punishment' in XP system.

      Next, when it comes to links, it's OK, I've got nothing against that. In fact, often it's exactly what someone was looking for. What I'm strongly against is re-writing some popular knowledge as an answer.

        Eyck wrote: What I'm strongly against is re-writing some popular knowledge as an answer.

        1. your "popular knowledge" may well be my blind spot.

        2. re-writing may well provide an instance of a good answer to the request "Please rephrase, so I can be sure I understand your meaning?"

Re: Why do nodes with minimal value get upvoted most?
by Aristotle (Chancellor) on Dec 24, 2004 at 02:33 UTC

    just rehash of some popular myth

    You'll have to show me examples where myths persist unchallenged. I don't believe this one for a second.

    maybe copy-paste of perldoc perlfaq

    And why should it not? First of all, why waste time writing an answer that is already answered? Secondly, maybe the seeker was not aware of the existing Perl documentation; assuming the poster mentioned where he got this copy-pasted answer from, it would help the seeker learn how to (attempt to) answer his next questions for himself.

    those elaborate and extremely popular answers usually have nothing to do with question at hand

    Looking at my own top 100 nodes and those of others paints the opposite picture.

    Sometimes it looks like they were created by some simple bot

    Short answers? Maybe. Elaborate ones? You'll have to show me. I never got the feeling around here, ever.

    And all that aside, you seem to expect of the voting system different things than it is able to accomplish. I wrote about what I think it does according to observations (rather than expectations) before.

    Hopefully my quoting you and addressing points directly will hold as sufficient proof that I actually read your node instead of running a bot that responded to it. And maybe that justifies the fact that I downvoted you for wildly exaggerated unsubstantiated claims presented as fact.

    Makeshifts last the longest.

Re: Why do nodes with minimal value get upvoted most?
by kutsu (Priest) on Dec 23, 2004 at 15:07 UTC

    Of course, threads complaining about the xp system also seem to have very high reps (sorry had to say that ;). I notice you give no possible answer and since a complaint without a possible solution is without "informational value", I'll just -- you so you won't be complaining about yourself.

    "Cogito cogito ergo cogito sum - I think that I think, therefore I think that I am." Ambrose Bierce

Re: Why do nodes with minimal value get upvoted most?
by revdiablo (Prior) on Dec 23, 2004 at 18:47 UTC

    I have some advice for you: stop thinking so much about XP. It is a completely bizarre [fun, but bizarre] subsystem on top of a very reasonable system that we call Perl Monks.

    Look at the questions and answers directly, not at how people voted for them. Sure, bad nodes may get upvoted a lot (not that I necessarily agree with this, but I'll grant you the point for the sake of discussion), but the important thing is that good nodes still get posted. And if you pay attention, I think you'll agree that they do.

Re: Why do nodes with minimal value get upvoted most?
by talexb (Chancellor) on Dec 23, 2004 at 15:39 UTC

    You have suggested a hypothesis, but provided no proof or evidence to support your hypothesis. To me, your node has little or no informational value -- and I downvoted it, thus disproving your hypothesis anyway.

    Alex / talexb / Toronto

    "Groklaw is the open-source mentality applied to legal research" ~ Linus Torvalds

Re: Why do nodes with minimal value get upvoted most?
by bradcathey (Prior) on Dec 24, 2004 at 12:35 UTC

    As a untrained "programmer," and one who can only limited time to coding, I often need the rehashed, rephrased, obvious answers pointed out. I am flattered that the more experienced monks take the time to show me stuff immediately that I maybe would have stumbled upon eventually. I'm not afraid of "learning to fish" but sometimes it's nice to be shown where the fish are in the first place. All that to say, I agree with those posters above who point out that a node of seemingly little value to you may mean a lot to someone else.


    —Brad
    "Don't ever take a fence down until you know the reason it was put up." G. K. Chesterton
Re: Why do nodes with minimal value get upvoted most?
by helgi (Hermit) on Dec 24, 2004 at 09:28 UTC
    Because the most important thing to learn about Perl (and any other programming language (or anything else for that matter) is how to find out for yourself.

    Learning to fish, in other words.

    The single most important thing I learned about Perl was how to use perldoc properly. I don't think I have needed to ask a technical question since. Not many anyway.


    --
    Regards,
    Helgi Briem
    hbriem AT simnet DOT is

      I got the impression that the OP referred to linking to useful information that already exists as a good thing, and reposting "common knowledge" when a link to the appropriate place or information on how to find it was more appropriate. In other words, the OP seems to suggest that teaching people to fish is good, and regurgitating data better linked-to or pointed-at is bad.

      To put it elsewise, I think you misunderstood the OP entirely. You're answering something like "Why do people who just cut-and-paste some vaguely unuseful information based on keyword-matching get more upvotes than attempts to 'teach a man to fish' do?" with something like "Because you should teach people to fish." Huh? He basically suggested that very thing, and asked why more points go to those who don't take that attitude.

      Granted, I think he probably could have phrased things slightly better, but not everyone in the world is as eloquent as I. Heh.

      print substr("Just another Perl hacker", 0, -2);
      - apotheon
      CopyWrite Chad Perrin

Re: Why do nodes with minimal value get upvoted most?
by bibo (Pilgrim) on Dec 27, 2004 at 15:02 UTC
    I'm coming in late, as I was offline during the holidays... It seems to me that posts about XP may end up going negative. I am too lazy to back this up with data. My making this comment seems to reinforce this point, as I felt inclined to downvote you, and then in a fit of perversity, voted you up. Go figure. My posting here won't help my XP.

    That said, I found out about this thread in "Recently Active Threads", not the front page. Your post isn't about PM, it's about XP. You are doomed as long as you focus on your own XP, or other monk's XP, instead of the real focus of this site: a support network for perl people.

    I have asked stupid questions here, and received helpful answers. I have posted unhelpful answers by not thinking clearly, and have regretted them and revised them to indicate my error. I come here often, and learn new things. As my other brothers and sisters have pointed out, a response that seems unhelpful to you may be just what I needed. Don't try to judge the whole site's content according to your own experience. perlmonks works because of the diversity of the members, not because of the authority (or XP) of any one member.

    --bibo

Re: Why do nodes with minimal value get upvoted most?
by aquarium (Curate) on Dec 31, 2004 at 13:36 UTC
    don't worry too much about the voting stuff. there's bound to be some groupies about, and other people, who take it much too seriously. you don't exactly get points for being right, or steering someone in the right direction; although some of us do try. you get points for striking a chord with the people who happen to read it. i'm sure that the person starting a thread does appreciate honest/direct/intellectual replies...when their thread gets hi-jacked into perl geek oblivion.
    the hardest line to type correctly is: stty erase ^H
Re: Why do nodes with minimal value get upvoted most?
by jbrugger (Parson) on Jan 06, 2005 at 07:53 UTC
    Would it not be an idea to make it only possible to downgrade a node if you give a reason for it as well?
    Then the owner learns of it, and if it is a nonsence reason, id might be corrected.
Re: Why do nodes with minimal value get upvoted most?
by fraktalisman (Hermit) on Dec 29, 2004 at 21:43 UTC
    Perlmonks works fine, not only for its members' diversity, but also for its voting system.
    Our existing perlmonks xp system, while often criticised (e.g. How is it possible being a saint without posting?, Newbies and votes, ...), is superior to most of the other existing ones. Default ranking mechanisms of online community systems like YaBB only account for posting anything. Just hang around long enough, reply often, and soon you have the honor to be a "senior member" or something the like.

    Perlmonks xp with its voting system could be made more complex, if the community needed it. As a possible new feature, there could be votes in certain categories. Like upvotes for creativity/novelty, upvotes for factual correctness, downvotes for the lack thereof, downvotes for ignorance/laziness, and still the good, old general votes based on the overall (first) impression one gets after reading. At least that seems to be the way most of us cast our votes. - But changing the voting system in a way like that would go too far IMHO.

    Another important aspect for measuring the worth of a node is already achieved by the existing system: taking relevance into account. Someone (who needs to have a certain level) moves the node to the front page. Voila: more upvotes!

    Update: This node keeps getting downvoted. Go on. Vote it down! Vote it down! LOL
Re: Why do nodes with minimal value get upvoted most?
by xorl (Deacon) on Dec 30, 2004 at 21:03 UTC
    Until we can see who voted, how they voted, and can affect their XP we will continue to have this irresponsible behavior.
      I'm willing to admit that I'm a enough of a petty, vindictive, spiteful person that I would use that to massively down vote any one that down voted any of my nodes. I certainly have some nodes that deserve a few --'s and I might let them slide, but I know I've gotten some on reasonably decent nodes, those I doubt I would hold back against the person(s) that 'slighted' me.
      I'm not saying that's the correct behavior in that situation, I'm just saying that I know that I would at some point do it, and if I would I'm willing to bet others would as well.

        Of course if anyone else feels the same, that would just serve to start a self-perpetuating cycle…

        Makeshifts last the longest.

A reply falls below the community's threshold of quality. You may see it by logging in.

Log In?
Username:
Password:

What's my password?
Create A New User
Domain Nodelet?
Node Status?
node history
Node Type: monkdiscuss [id://417101]
Approved by edan
Front-paged by hsmyers
help
Chatterbox?
and the web crawler heard nothing...

How do I use this? | Other CB clients
Other Users?
Others taking refuge in the Monastery: (3)
As of 2022-06-29 00:51 GMT
Sections?
Information?
Find Nodes?
Leftovers?
    Voting Booth?
    My most frequent journeys are powered by:









    Results (94 votes). Check out past polls.

    Notices?