in reply to Re^4: Notification of XP increase has gone wild (maybe)
in thread Notification of XP increase has gone wild (maybe)

Do you see a viable reason why someone should get a gratification for casting hundreds of votes week after week on old nodes?

And we already have - for other reasons - dog-votes in place to punish too many down-votes.

My suggestion is just to reduce the 25% ratio to 0% after a frequency of x votes on y old nodes in a time-span of z days, so not really a punishment. (x,y,z is a matter of debate)

Personally ... if I thought 100 archived nodes deserved to be up-voted, I wouldn't really care about the reward.

And this exploit is only of interest for monks in the Saints region, which have too many votes for contemporary nodes. Saints shouldn't care either.

"Gelegenheit macht Diebe" ( ~ Opportunity makes a thief).

> and there is already a blanket rule against abuse.

A rule only makes sense if it can be enforced. I can think of many ways to disguise such votes as legitimate.

We only learned now about it because someone was sloppy enough to mainly upvote davido

"Wo kein Kläger, da kein Richter" ( ~ "where there's no plaintiff, there's no judge")

I think it's much easier to limit the exploit instead of implementing a surveillance apparatus to hunt down the culprits.

Cheers Rolf
(addicted to the 𐍀𐌴𐍂𐌻 Programming Language :)
Wikisyntax for the Monastery

  • Comment on Re^5: Notification of XP increase has gone wild (maybe)

Replies are listed 'Best First'.
Re^6: Notification of XP increase has gone wild (maybe)
by hv (Prior) on Oct 25, 2022 at 22:34 UTC

    Do you see a viable reason why someone should get a gratification for casting hundreds of votes week after week on old nodes?

    I'm not sure why you would ask me that question. Was there anything in my post to suggest I believe someone should get such gratification?

    A rule only makes sense if it can be enforced.

    That's a separate problem. It's also the real problem, and the one worth focusing on.

    I can think of many ways to disguise such votes as legitimate. ... I think it's much easier to limit the exploit instead of implementing a surveillance apparatus to hunt down the culprits.

    It is not unusual that addressing a symptom is easier than addressing an underlying problem.

    I think there are a variety of patterns of behaviour that could easily be detected and flagged for a closer look. If code were implemented to detect such patterns, I do not think it would be beneficial to publicise the details; I am in two minds about whether it would be beneficial to publicise its existence.

[reply]
      > > Do you see a viable reason why someone should get a gratification for casting hundreds of votes week after week on old nodes?

      > I'm not sure why you would ask me that question.

      Because you said

      > > > Changing the economics will not stop the problem of people abusing the system,

      And I disagree. Limiting the gratification will make the abuse useless.

      Additionally: trying to detect the abuse will not only cost far more, the ensuing discussions will be nerve wracking.

      Hence, if there is no viable reason to allow that gratification ... why keep it?

      Cheers Rolf
      (addicted to the 𐍀𐌴𐍂𐌻 Programming Language :)
      Wikisyntax for the Monastery

[reply]

        Ah, I understand. The intent of my earlier message was: changing the economics to fix one type of abuse will not stop the underlying problem that there will always be someone finding a way to abuse the system.

        I assume also that only a very small proportion of users seek to abuse the system; in that case, every time you tweak the system to close down another possible route of abuse, the collective impact on non-abusers is far greater than that on abusers.

[reply]

In Section Perl Monks Discussion