Do "this question has not received enough attention" bounties work as well as we'd like them to?

Question

I've answered a couple of bounty questions in the last couple of weeks, the bounty list was not something I'd had a lot to do with before. They seem to attract a lot of answers, which is the point, but a disproportionate amount of them appear sub-optimal, which can't be fun for the OP. The answers either cover the same ground as existing answers or, based on comments from the OP, just don't answer the question. It's like people are rushing to make a contribution without as much consideration to what kind of contribution it is as they might give on a "normal" question. I'm wondering if:

1. since my sample size is small, and I'd like to think I'm wrong, my experience is aberrant?

2. bounty posters are generally satisfied with the overall results of offering such bounties even if what I think I'm seeing is in fact the case?

Answer 1

I think I've offered the most bounties on the site as of today - 35, for 3,000 reputation points total - so I'm working from a bit of a larger sample size.

My experience with the bounty system in general on SE - and this holds for Worldbuilding, by the way - is that it's often an enormous success or a total failure. A lot of it depends on who sees the post. For instance, I placed a bounty on this question about habitability because I was concerned about orbital stability, and now someone's running orbital simulations - because the right person saw it. On other occasions, I've been extremely disappointed, and have gotten sub-par responses. I'm not going to give specific examples, though.

Does adding a bounty to the question increase the odds that the right folks will see it? I'd argue yes. We usually have a couple of bounties maximum at any given time; at the moment, we have half a dozen - a bit of an outlier. Nonetheless I think people are often intrigued by seeing the bounty list, and everyone wants the chance to win 50, or 100, or a couple hundred rep points. So . . . yes, I think they do get you extra visibility. The odds aren't high, but they're non-zero.

There are some things I've found that will make some bounties more successful than others:

• Not already having an accepted answer
• Not having many high-voted answers (i.e. >5 total score)
• Not having a lot of answers, which would bury new ones
• Having a question that's accessible to more people

Yes, these do make a difference, and the questions with successful bounties I've raised usually fulfill most of these criteria.

As an aside, another reason to use bounties is to reward existing awesome answers, which will hopefully encourage more awesome answers. I've had a lot of success getting these answers more attention (e.g. 1, 2, 3). But that's kinda unrelated to your question.

Answer 2

Bounties are marketing - if you got some feedback they do what they are supposed to do

It's a bit weird, but you have to view bounties as money you spend for advertising your question a bit more so that some people may look at it. The primary usecase is if you haven't gotten any feedback and therefore your "question has not received enough attention".

As someone who regularly posts bounties on this and other sites I can say that they often don't quite work the way I wish they'd work. For example I noticed that often a 50-rep bounty will go basically unnoticed. Sometimes a bigger bounty helps because you attract lots of people - a big number as a potential bonus lures people to your question. This, of course, also leads to many people who are trying to get a positive-scored answer so that they are eligible for the automatic bounty distribution and those answers may very well not be the top answers that you expect. But that's basically business as usual: many answers mean that quite a few of those will be okay-ish, but hopefully there will be one or two good or even great ones - and those are what you paid your bounty for.

Bounties work most of the time. They work less good for small bounty values and the more active bounties are currently out there, the lower the chances that people will spend significantly more time with your specific question (that's what I saw during the Holiday Bountapalooza for example). A bounty brings in more attention and therefore accomplishes its goal - more attention means potentially more answer, which means potentially some not-so-great answers, but hopefully a good or great one in between that helps you.

Bounties work as intended and as someone who regularly puts up bounties I am satisfied most of the time. The quality of responses is often similar to what I see on other questions and sometimes there are really good responses if you put a high enough bounty on a question requiring a difficult enough answer - I suspect a bounty makes some people think "Well, it wasn't really worth my time before, but with an extra 100/200/... reputation I might have another look at it".

For reference, with kind regards to Kingledion for his great "Top users by bounties" query on SEDE that he introduced in this trivia quiz:

• I have used 2200 rep for bounties on WorldBuilding.SE, making me the number 3 on that site - see here
• I have used 3650 rep for bounties on Writing.SE, making me the number 1 on that site - see here

Answer 3

I've been answering questions on this site for a while now and in the vast majority of cases, my answers have been to questions on the normal queue. I've put answers to questions that have then been put on bounty, and I've noticed that there are a few cases where some 'interesting' interpretations of the question appear.

That said, recently the extinction equilibrium question got a bounty on it. I'd not provided an answer to it previously, but when I noticed it on bounty I have to admit I thought about it just a little harder and came up with an approach that I thought might work. It's currently the highest ranked answer on the question and for what it's worth, I also got a bounty out of it.

So; I'm not saying I'm the best answerer on this site; far from it in fact. That said, I do consider myself a 'senior answerer' these days and a provider of (at least) good quality answers.

Therefore, I'm of the opinion that the bounty works both ways; it attracts attention to be sure, but not all of that is bad attention. It can attract good attention as well, but you as someone who posts a bounty, you're obligated to sort out which is which.

It's already been stated that bounties are more than any other factor a marketing device and I concur with that assessment. The trick is to only offer a bounty if the answer you select:

1) Is the right answer
2) Is of good quality

I'm of the view that a poor answer to a question doesn't deserve a bounty 'just because it's the best of a bad bunch'. That said, that is a decision that the poster of the bounty has to decide for his or herself. Yes, I've won a few bounties in my time, but I'd like to think I won them on merit, rather than a lack of options.

The point of all this background is that I think the answer to this is quite simple; if you post a bounty on a question to draw attention to it, you're not obligated to give that bounty to a poor answer just because it's the best of the poor answers. Doing so only drives the behaviour described in this question and ultimately bounties that are truly earned will lead to more of the 'right' attention for future bounties.

Never having posted a bounty I can't comment on what bounty posters think, but I can say after tracking a number of bounty questions, the added attention does not skew the overall quality of answers provided either to higher or lower quality according to my own sample set.