Meanwhile, Experience Points are important so people answer to questions. If this kind of reward didn't exist most people would just ask, and never answer.

Another interesting thing about Experience Points is that it let you know how good are your answers.

If this kind of reward didn't exist most people would just ask, and never answer.

I think people would still answer questions even if there were no XP. Look at the Perl mailing lists. People asking there still get answers, and the responders don't get any "reward" except for feeling good about helping out.

Another interesting thing about Experience Points is that it let you know how good are your answers.

This isn't necessarily true, either. Plausible sounding answers often get upvoted even if they are wrong.

So far, I've found that if you're submitting a plausible sounding wrong error in the hopes of grubbing for XP, that it's best to have a well formatted, plausible sounding wrong answer.

Um...was that out loud?

Actually, to some degree, I think that there are plenty of people with different personalities out there. There probably are those who are concerned with cheating the system, and think that XP on Perl Monks is their path on the way to ... um... well, something. Anyway, there are those that like helping out, there are those that like showing off how much they know, and there are those that just like being complete asses in public forums. (and I say that after years of various online forums from BBSes/MUDs/Usenet/mailing lists/etc.) Nothing's really changed over the years, just the scope of people contributing, so you end up with more extreme people -- the people who really like to help, or really like being obnoxious.

But, let's face it ... people contribute to CPAN, and they don't get anything in return, other than maybe some noteriety. (see the gift economics discussion in the Suffering from CPAN guilt thread) Sure, there's a chance that you might get a job indirectly from posting to CPAN or Perl Monks, but most people aren't contributing because they're expecting something back, other than maybe feeling good about themselves for helping others.

Personally, I tend to contribute because I'm bored and/or procrastinating from doing other things (eg, I probably should be doing my taxes right now, or my homework for class next week.) Yeah, it'd be nice for someone else to learn from my past mistakes, so they don't have to make them, or to make other people think about new and interesting things, but well, I haven't figured out how to turn that into a viable economic model, or to help me in my plans for world conquest, so I'll just accept that ... um ... I should probably go do my homework now.

I don't see his request as a "I want to win at perlmonks" request. I see his request as a "You know, here is yet another way to learn interesting stuff."

I never look at "Worst Nodes", but I commonly look at "Best Nodes", because quite often "Best Nodes" contains nifty solutions to problems. Having a "most change in 5 days" is yet another pointer to nodes that might contain something interesting. I don't see how this could be a bad thing.

-Scott

For this a node with 10 ++ vote and 10 -- votes would get a score of 10, and a node with 10 ++ votes and no -- votes would get 0.

This code is a start. The $mult is used to keep nodes with only 2 votes from getting a score of 10.

for my $p (0..15, 100) {
    for my $n (0 .. 15, 100) {
        next unless $p+$n;
        my $mult = 10 - ($p+$n) > 0 ? 10 - ($p+$n) : 10;
        print(10 - ($mult * (1 - abs($p-$n)/($p+$n))),
              " $p $n $mult\n");
    }
}
[download]

UPDATE: changed divisive to controversial. And added a note to the code and reduced its range.

See: Node Tension

I kind of agree with the idea of being able to draw attention to the nodes that seem to generate the most "controversy", in the sense of attracting relatively higher numbers of both up- and down-votes. I'm not sure how many nodes would score highly on this sort of metric -- partly because PM has never shown us the numbers of up- and down-votes separately, only the resulting node-rep.

Obviously, Best Nodes and Worst Nodes focus attention on nodes that have already attracted attention -- but only if the nodes have evoked a consistently positive or negative response.

Nodes that get a lot of conflicting votes have also attracted attention, and it would seem natural to have a page that could focus more attention on them, despite the fact that their node-rep is close to zero.

As for the code you presented, I don't get what you're trying to provide there. What's the point of printing out 10,200 lines (101**2-1), and what's the meaning or intended use of the first value on each line?

I actually don't know whether the PM database keeps track of up- and down-votes separately; or whether it keeps track of total count of votes cast along with current node-rep, which amounts to the same thing. If it does neither of those things (i.e. it tracks only current node-rep, which is just incremented or decremented as votes are cast), then the information needed for a "Conflicted Nodes" page is simply not available.

But if two values are being tracked, then ranking nodes top-down by "$total_votes - $node_rep" "$total_votes - abs( $node_rep )" would do the right thing.