20
\$\begingroup\$

Recently, three answers gained 500 rep bounties at about the same time, with the following bounty explanation:

One or more of the answers is exemplary and worthy of an additional bounty.

ton ****** is cool and i want to give him internet points

These three questions are:
Generate lazy values
Compute the Hausdorff Distance
Repair the ranges

(The bounties end on the 19th).

For reference, a user who had 3593 rep is giving 1500 rep to a user with 8643 rep.

My question is: is this considered inappropriate use of a bounty, and if so, what action can be taken against it?

I can provide image proof should it becomes necessary after the 19th.

\$\endgroup\$
  • 2
    \$\begingroup\$ Just for the record, the user giving away the bounties had 3593 rep before starting the bounties. \$\endgroup\$ – Martin Ender Jan 12 '17 at 15:54
  • 1
    \$\begingroup\$ Related discussion on Mother Meta: Is mass bountying considered vote fraud? \$\endgroup\$ – Dennis Jan 12 '17 at 15:54
  • \$\begingroup\$ I can provide image proof too. \$\endgroup\$ – Mitch Schwartz Jan 12 '17 at 15:59
  • \$\begingroup\$ Everyone seems to be basing their argument on the premise that bounties are used to reward good answers. Everything I've found in the help section states that bounties are for bringing attention to a question. (link and link) Is there another page that says bounties can be used to reward good answers? \$\endgroup\$ – Mike Bufardeci Jan 12 '17 at 18:25
  • 5
    \$\begingroup\$ @MikeBufardeci One of the bounty reasons is "One or more of the answers is exemplary and worthy of an additional bounty." It might actually be the most commonly used on PPCG. \$\endgroup\$ – Martin Ender Jan 12 '17 at 18:42
  • \$\begingroup\$ @MartinEnder: I think "this question has not received enough attention" is probably a little more commonly used on PPCG, but "one or more of the answers is exemplary" is definitely top two. \$\endgroup\$ – user62131 Jan 12 '17 at 20:58
  • \$\begingroup\$ @MartinEnder I only really learned about bounties today, thank you for filling me in. \$\endgroup\$ – Mike Bufardeci Jan 12 '17 at 21:06
  • 1
    \$\begingroup\$ I don't think that you should accept an answer when there's no real consensus. Right now there's an accepted answer at +5 that basically says that this is not okay, and there are two answers both at +2 saying that this is okay. \$\endgroup\$ – Mike Bufardeci Jan 12 '17 at 21:08
  • \$\begingroup\$ @MikeBufardeci Okay. For now, I'm taking off the accept until we can get a general consensus, but I would strongly agree with Dennis's points. \$\endgroup\$ – user42649 Jan 12 '17 at 21:27
  • \$\begingroup\$ @AlexL. I also agree with Denis's points, I just think that it's not far enough ahead of the other posts to merit being accepted. I'm not super familiar with how things work around here though so I could be mistaken. \$\endgroup\$ – Mike Bufardeci Jan 12 '17 at 21:41
  • \$\begingroup\$ A similar but more clear-cut situation came up on Puzzling. It was eventually resolved by the recipient giving the rep back to the donor via bounties. \$\endgroup\$ – xnor Jan 12 '17 at 22:17
15
\$\begingroup\$

In general: no

You can only award reputation as bounties if you've already earned that reputation, and you are free to award as much of it as you like to as many posts as you like. If several posts you consider deserving of your hard-earned rep happen to belong to the same user, that's not a problem; posting great answers rarely happens by accident, so it's quite natural that a user that shares your interests with regard to challenge types, topics, and programming languages might post more than one answer that catches your eye.

In this specific case: yes

Unfortunately, moderators do not have magic wands to read the users' minds, but I have yet to find a shred of evidence that these bounties were placed with the intention to reward great answers. There's a lot that suggests quite the opposite though.

  • The post notice templates say

    One or more of the answers is exemplary and worthy of an additional bounty.

    but the user-supplied description

    ton hospel is cool and i want to give him internet points.

    states quite clearly that these bounties will be awarded to a person rather than their answers.

  • The three answers of @TonHospel that were picked for the bounties were simply their three oldest answers. I'm not saying the answers aren't good enough to get bounties, but they were certainly picked arbitrarily.

    • One of the answers is a rather straightforward port of a JavaScript answer by another user to Perl. That doesn't make the answer any worse, but a +500 bounty – the ultimate award an answer can receive – seems like a stretch.

    • Ton Hospel has posted amazing answers that are, in my opinion, more deserving of a bounty that the three ones that were picked. For example, there's Calculate the permanent as quickly as possible which is certainly exemplary.

  • The (ongoing) chat discussion (starts here) and, in particular, the message

    i thought it was really cool that ton hospel visited the site and participated quite a lot

    speak for themselves.

Is this a problem?

I'd say so. Even the simple act of transferring reputation from one account to another is an inappropriate use of the bounty system. If that was intended to be possible, there would be a tip jar.

But that's not all bounties do! The three bountied questions will appear in the featured tab until they're awarded or expired (which might take up to 7 days). A bounty also pretty much guarantees a permanent slot on the front page. That's OK if the bounty draws attention to an exemplary answer, but not so much when they're used as "rep cheques".

What can be done about it?

I'm not saying that we will (or should) in this particular case, but moderators can refund bounties under exceptional circumstances.

\$\endgroup\$
  • 3
    \$\begingroup\$ As long as it's not actual rep fraud, you shouldn't have to justify your reasons for assigning a bounty any more than you would justify your reasons for upvoting an answer. Nobody should be able to tell you that your vote is "wrong", and these kinds of bounties are really just an extended form of upvote. \$\endgroup\$ – James Holderness Jan 12 '17 at 17:31
  • 6
    \$\begingroup\$ Targeted upvoting will get reversed by the system and will result in suspensions if repeated. \$\endgroup\$ – Dennis Jan 12 '17 at 17:34
  • \$\begingroup\$ That kind of vote reversal is an automated process based on X votes in N hours or something of the sort. Nobody (AFAIK) is making a value judgement on the votes and deciding whether they agree with the reason for voting before deciding whether to reverse it or not. And bounties aren't covered by that process because they are already self limiting. \$\endgroup\$ – James Holderness Jan 12 '17 at 18:55
  • 6
    \$\begingroup\$ That is not true. While moderators cannot selectively reverse votes, community managers can and do so whenever needed. While I'm not aware of an automatic process that monitors bounties, they show up in the moderator tools for a reason. \$\endgroup\$ – Dennis Jan 12 '17 at 19:06
  • \$\begingroup\$ Okay. Thank you; this clears up a few things. Also, I have no intention to suggest that Mitch was intentionally trying to commit any sort of rep fraud or anything related. \$\endgroup\$ – user42649 Jan 12 '17 at 20:39
  • \$\begingroup\$ @Dennis Are you saying that community managers will reverse votes, not because of voting fraud, but because they think the answers weren't worthy of that rep? Because that's what we're talking about here. Nobody is claiming voting fraud - it's just that some people think the rewarded answers aren't the most deserving. \$\endgroup\$ – James Holderness Jan 12 '17 at 23:38
  • \$\begingroup\$ @JamesHolderness Not being a community manager myself, I don't even know if CMs can see on which individual posts votes were cast, they certainly don't go through all of them individually. I was responding to your claim that vote reversal is an automated process based on X votes in N hours, which is not true. Targeted upvoting (i.e., upvoting because of the poster rather than the post) can remain undetected by the system, so a CM may have to intervene. I'm not sure if targeted upvoting falls under your definition of vote fraud, but it can get reversed and it can result in a suspension. \$\endgroup\$ – Dennis Jan 12 '17 at 23:51
  • \$\begingroup\$ @Dennis In your efforts to prove me wrong, you seem to be deliberately ignoring the key part of my argument. CMs aren't making a value judgement on the worthiness of the answer when deciding whether to reverse votes. Whatever the rules are for voting fraud, the CM or moderator shouldn't be bending the rules based on whether they like the answers being voted on. If you want to argue that targeted bounties are fraud that's fine. But you don't get to qualify that with "unless I happen to like the answers that were awarded the bounty". \$\endgroup\$ – James Holderness Jan 14 '17 at 1:02
  • \$\begingroup\$ @JamesHolderness I have no idea how CMs determine vote fraud. At least with regard to the process, this situation isn't even remotely comparable though; we're talking about three bounties, not hundreds of votes. \$\endgroup\$ – Dennis Jan 14 '17 at 1:39
  • 1
    \$\begingroup\$ Imagine if Jon Skeet from StackOverflow (who has nearly 1 million rep there) decided that another user was cool and to transfer 100K of his rep to that user via bounties. That would take a long time to do (max 3 bounties at a time), and it is certainly not how bounties are intended to be used. It really is "fraud" of a kind, because using rep in that way can give users access to privileges they didn't have before, and rep is a means of ranking/standing. It would also stick his bountied questions to the "featured" tab. \$\endgroup\$ – mbomb007 Jan 20 '17 at 20:26
  • \$\begingroup\$ @mbomb007 And yet the SE community managers have already said that this kind of rep transfer is not considered fraud. Also nobody seems to have a problem with people gaining thousands of rep through zero or one byte answers. I guess that's a totally reasonable way to earn privileges and ranking. \$\endgroup\$ – James Holderness Jan 23 '17 at 16:44
  • \$\begingroup\$ The first half of their fraud definition is "The sole purpose was transferring reputation (the person offering the bounties had no particular appreciation of the answers to which they were awarded, nor any desire to see the questions attract better answers)." So even if it's not fraud, the user really needs to list the bounty reason as something other than "this user is cool". \$\endgroup\$ – mbomb007 Jan 23 '17 at 17:35
2
\$\begingroup\$

Assume the bounties are being awarded in good faith.

I read as much of the chat history as I could bear, and I have not seen a single person ask Mitch if he thinks the answers posted by Ton Hospel are great answers. It's quite possible that Mitch thinks Ton Hospel is cool, and also thinks his answers are great. That's good enough for me until Mitch gives me a reason to believe otherwise.

Mego's comment motivated me to post this answer.

I think we could all benefit by remembering to Assume Good Faith more often

\$\endgroup\$
  • \$\begingroup\$ His "bounty reason" that he listed, as well as his arguments in chat, prove otherwise. Sure, good faith should be assumed more often, but not in this case when there is significant evidence to the contrary. In chat, the reasons he specifically gave were that "ton is cool". He even acknowledged that he simply chose the first answers that ton had posted. \$\endgroup\$ – mbomb007 Jan 20 '17 at 20:29
  • \$\begingroup\$ @mbomb007 No, they do not prove that. Did you bother to read my answer all the way through before commenting? It's quite possible that Mitch thinks Ton Hospel is cool, and also thinks his answers are great. One does not preclude the other. \$\endgroup\$ – Rainbolt Jan 20 '17 at 20:44
  • \$\begingroup\$ I was there for the chat conversation. He didn't argue that the answers were good. He basically agreed to having only selected the first three answers arbitrarily in order to give rep to ton. \$\endgroup\$ – mbomb007 Jan 20 '17 at 20:50
  • \$\begingroup\$ @mbomb007 That's where I think you messed up. Mitch never agreed that he only selected Ton's answers to give rep to Ton. Mitch gave that as a reason, but he never said that it was the only reason. And it's exactly this kind of reading too much into chat messages that led to an unpleasant experience in chat, where for about an hour it was basically Mitch vs the room. \$\endgroup\$ – Rainbolt Jan 21 '17 at 23:25
  • \$\begingroup\$ He never once gave the argument that he thought the answers were worth they bounties. \$\endgroup\$ – mbomb007 Jan 22 '17 at 2:15
  • \$\begingroup\$ Many people seem to be operating under the premise that a 500 point bounty is a big deal. I reject this premise. I think that more valuable than any upvote or bounty is simply reading a post and trying to understand and appreciate it, and looking for solutions/improvements/errors/omissions. I think Ton's answers, including the early ones here, reflect his excellent golf and problem solving skills along with his humble and unassuming attitude. A post does not need to be astonishing to be exemplary. And answers that may seem simple here could seem tricky in a closed source competition setting. \$\endgroup\$ – Mitch Schwartz Jan 22 '17 at 20:07
  • \$\begingroup\$ @mbomb007 "He didn't say it; therefore, he doesn't think it." I don't know how else to explain that your logic is fallacious, so this will be my last comment (at least in response to you). \$\endgroup\$ – Rainbolt Jan 23 '17 at 21:11
  • \$\begingroup\$ View the comments on Dennis's answer. \$\endgroup\$ – mbomb007 Jan 23 '17 at 21:43
1
\$\begingroup\$

As Dennis mentioned in a comment, a similar discussion was brought up on Mother Meta. Really, users are free to use bounties as they please, with a few restrictions (this list is not necessarily exhaustive):

  • Don't engage in rep fraud (using sockpuppet accounts to build up rep and transfer it via bounties to a "main" account, circumventing the rep cap)
  • Don't pass bounties back and forth repeatedly between the same users to abuse the system and keep a question indefinitely featured

If the user who put up the bounties genuinely thought that answers by another user were worthy of the bounties (which appears to be the case here), there's no issue.

That said, there's always the possibility that the moderators would find signs of abuse via their tools. This is just my opinion, as someone without access to diamond tools.

In fact, depending on the winners of the Best of 2016 competition, a single user could end up with a much larger rep influx from those bounties. That's also fine.

\$\endgroup\$
  • 6
    \$\begingroup\$ I have my doubts that "the user who put up the bounties genuinely thought that answers by another user were worthy of the bounties" (although I don't doubt they thought that the author of them is worthy of the bounties). The three bountied answers are simply the first three the author posted, so it doesn't really look like they specifically picked out the most exceptional answers. \$\endgroup\$ – Martin Ender Jan 12 '17 at 16:05
  • \$\begingroup\$ @MartinEnder While that may be true, there's no clear abuse going on here (at least, that's visible without diamond tools). I'm not going to try to dissect someone's intentions as a distant observer. \$\endgroup\$ – Mego Jan 12 '17 at 16:09
  • 3
    \$\begingroup\$ And one of them is a relatively straightforward JavaScript-to-Perl port of an answer by someone else. I'm not saying it's not a good answer, but I quite frankly fail to see why someone would award a 500 rep bounty to it. \$\endgroup\$ – Dennis Jan 12 '17 at 16:09
  • 1
    \$\begingroup\$ @Mego The linked meta discussion does start with "Bounties are like votes; you're supposed to be awarding the points to the exceptional post, not the user. If it appears someone is repeatedly using bounties to give reputation to the same user on posts that do not appear to deserve it, that's probably an abuse." I do agree though that it's hard to argue the case either way. That said, even if you want to reward a user for their continually great content, it would still be a better course of action to go find their most exceptional posts to bounty those. \$\endgroup\$ – Martin Ender Jan 12 '17 at 16:11
  • \$\begingroup\$ @MartinEnder The line between "great post" and "not-great post" can be fuzzy if you're not familiar with the language. I barely know Perl, so all of Ton's Perl answers look like awesome voodoo to me, and I wouldn't have any reservations about bountying any of them, if I wanted to award him a bounty for clever golfing. \$\endgroup\$ – Mego Jan 12 '17 at 16:13
  • 2
    \$\begingroup\$ @Mego That's part of what I mean though. Bounties are supposed to reward great content, so it doesn't make sense to bounty an answer you don't understand. You could pick out any of the great answers that also have good explanations instead. \$\endgroup\$ – Martin Ender Jan 12 '17 at 16:15
  • \$\begingroup\$ @MartinEnder And I agree that the answers that are apparently intended to receive the bounties aren't the best choice for awarding great content, but the question isn't "is this the best use of bounties?" - it's "is this an inappropriate use of bounties?", to which my answer is "no". \$\endgroup\$ – Mego Jan 12 '17 at 16:17
  • 2
    \$\begingroup\$ @Mego Then you probably shouldn't cite a meta discussion that contradicts this position. :P \$\endgroup\$ – Martin Ender Jan 12 '17 at 16:18
  • \$\begingroup\$ @MartinEnder The answer on the post doesn't directly contradict my position. It's clear that potential issues need to be addressed on a case-by-case basis. As someone without diamond tools, it doesn't look like abuse (Ton did use some clever golfing tricks, even if those answers aren't the most impressive ones he's posted), so there's no problem in my opinion. \$\endgroup\$ – Mego Jan 12 '17 at 16:22
  • 2
    \$\begingroup\$ @Mego I think this answer by Shog9 may be a better reference. From my reading of the post, I think it's clear that this situation would not be considered fraud, even if the bounty was for the sole purpose of transferring reputation, as long as there was no bounty trading or expectation of quid pro quo (which nobody has suggested in this case). \$\endgroup\$ – James Holderness Jan 14 '17 at 1:15
  • \$\begingroup\$ @JamesHolderness That one somehow didn't turn up in my searching. I agree that it applies here - there doesn't seem to be any intent of dishonesty or abuse, so a case like this that straddles the line of acceptableness falls on the good side. Like I said in chat - assume good faith. \$\endgroup\$ – Mego Jan 14 '17 at 1:34

You must log in to answer this question.