Context note: this is coming off of questions Why can't I just offer a bounty for a question right off the bat? (where an earlier version of it was offered as an answer but seemed unlikely to be seen much) and Mandatory delay in offering a bounty (marked as a dupe of the previous question, where I was advised in comments to present it as a feature request)

Proposal: Let immediate bounties happen, but give them a cost of twice what they reward, minimum bounty 100. Possibly crank up the minimum rep necessary to place an immediate bounty (500? 1000?). Those who are willing to wait the two days, you can go under standard bounty rules. This has the following effects.

  • Immediate bounties are inherently costly. Being able to afford on in the first place means that the user has gained some rep already, and it's not something that can be done lightly or frequently. This should minimize any difficulties with moderation load.

  • It won't break the rep economy, since swapping immediate bounties inherently takes rep out of the system. Anyone who wants to do it and keep doing it is going to have to generate a fair bit more rep than they're handing out.

  • It's true, this is nonegalitarian, to a degree. It's essentially another way to take advantage of high rep, added to the long list of such ways (some explicit, some implied). I would argue that this is not inherently a bad thing, and not really at odds (at least inherently) with the way we do things anyway. In particular, I look at this answer to bounty system discriminates against new users .

  • It will allow a case where a moderately high-rep user who needs an answer to a high-complexity problem immediately can pay for one. It gives a certain class of experts a serious motivation to build rep - because they might want or need an immediate answer (or two, or three) some day.

  • In general, the sorts of questions that will get this treatment will be difficult questions that require rapid answers and that are not available elsewhere. Giving people a way to "buy" answers to such questions here makes it more likely that they'll show up here, which would seem to be a good thing, overall.

  • 2
    Wasn't "It's true, this is nonegalitarian..." the main point in disagreement with previous requests of this nature? You've not managed to address it here. – ben is uǝq backwards Apr 15 '13 at 13:40
  • @benisuǝqbackwards My understanding was that the main point of disagreement was "it will break the answer economy. No one will bother to answer unbountied questions." - which I believe this addresses. The second point of disagreement was "it would be a moderation nightmare" which I believe this also addresses reasonably well. I have seen some complaints about nonegalitarianism, but I haven't yet seen anyone explain why the nonegalitarian aspect of this is bad. I agree that it's there. Could someone explain what makes it a problem? – Ben Barden Apr 15 '13 at 15:01

Ultimately the problem with your proposal is it doesn't actually address the initial concerns, it just patches or reduces them.

  • Bounty economy instead of rep economy? Fine, make the bounties more expensive so there's not as much too much of a bounty economy.
  • Non-egalitarian? Fine, privilege very high rep users instead of all high rep users.
  • Too much of a moderation load from allowing immediate bounties? Fine, allow not that many immediate bounties.

Now you get half the benefits in exchange for half the rewards. If immediate bounties were a problem before, they're still a problem under your proposal just not as much of a problem because there are fewer of them. So therefore you would need to argue the marginal benefit of the first few immediate bounties outweighs the cost. But the problem is all the reasons they were bad before still apply. There's just fewer of them.

For what it's worth - ultimately this answer is my problem with immediate bounties. I see bounties' function as bumping attention when a question was too thorny to garnish it on the first round, so immediate bounties have bad marginal value on the onset, and throttling doesn't help.

| improve this answer | |
  • I admit, I'd like the answer that you link to be the case. I would. The problem is that some questions are simply hard to answer, and if well-written are clearly so (and if well-titled, are clearly so from the title). This means that relatively few people are going to read them in the first place, limiting the potential for upvotes. Further, even if they do get a lot of exposure, they may take a fair amount of effort to answer (being complicated and difficult to explain). Bounties are a way to convince people to go to that effort. Thorniness isn't necessarily about writer skill. – Ben Barden Apr 15 '13 at 15:27
  • I'd also argue that you're getting a lot more than half the benefits for a lot less than half the reward. Making it a net rep negative rather than a net rep positive means that the bounty economy actually feeds more rep demand into the standard economy than rep supply. Well-written questions should not generate moderation load (anyone can still flag the things, after all) and this "filter" should cut out almost all of the cases of poorly written questions. – Ben Barden Apr 15 '13 at 15:33
  • @BenBarden can you think of a mechanism to ensure that writers put effort into making their question as best as it could be so it gets attention even if complex even while they have a shiny immediate bounty sitting right there? Being high rep doesn't exempt you from writing good questions for the low price of 200 rep. – djechlin Apr 15 '13 at 15:36
  • @BenBarden How does creating rep demand have anything to do with the problem of bountied questions stealing the spotlight? Immediate bountied questions will be low-quality more often than ones with delay, this still exacerbates the moderation load. – djechlin Apr 15 '13 at 15:38
  • @djelchin It's true. Being high-rep doesn't in any way exempt you from writing good questions. I would generally think that high rep users would tend to write good questions anyway - especially as the sorts of questions that won't get quick answers on their own often need good questions to answer in the first place - and if it didn't need the bounty, then the time spent to write it as a good question that would naturally get a quick answer is going to be a lot less than the time you'd have to spend elsewhere to earn the 200 points. – Ben Barden Apr 15 '13 at 16:06
  • @djelchin There seems to be a disconnect here. Is "high-rep users asking bad questions" really a problem on SO? How often does that happen, really? I've been under the impression that the primary sources for low-quality questions were new users who were either malicious or didn't know any better. Am I wrong? For that matter, how many of the questions overall are asked by high-rep users? Yes, a bountied question is going to get a bit more of the spotlight, but my impression had been that such questions wouldn't be all that common. Yours seems to be quite the opposite. Why? – Ben Barden Apr 15 '13 at 16:10
  • @BenBarden This is exactly my point - your defense is bountied questions "wouldn't be all that common." So that's just... less evil. How about even less evil by not allowing immediate bounties. Re. question quality, I think this will lower the quality of questions from some users, and that alone makes it a non-starter. It's not relevant whether more low quality questions come from different users. – djechlin Apr 15 '13 at 17:03

Immediate bounties are inherently costly. They're going to take at least 200 points to throw out there. The effectiveness is much higher if you wait. Also, even being able to afford one in the first place indicates that you've been around for a little while, and have hopefully picked up a bit on how to write a good question. Even if we don't open up standard close procedures, I'd think that a 1000+ point user blowing 200 points inefficiently for an immediate answer to a bad question would be rare enough we could afford a bit of mod time for it.

My bet is everyone will just offer immediate bounties since why wait. Certainly the "high rep" users don't care about an extra 100 points. Plus as an 8k user I generally offer 100-150 points after the two day window. I would much rather just offer 50 immediately given the option (but rather the option not exist at all).

So we will still have a bounty economy, and it's nonegalitarian. That leaves

Adding bounties makes moderation more difficult - currently, a question that has a bounty on it cannot be closed or deleted without mod powers.

which AFAICT is still as much of a problem. Your answer does nothing to allay the concerns you've identified.

| improve this answer | |
  • Two sides of your answer. First, your "everyone will do this all the time". They won't be able to afford it. In your case, you have 139 questions. If you'd somehow sunk 200 points worth of bounty on each of those (the minimum under this suggestion), that would be over 20,000 rep cost to you. At the same time, you've put in well over twice as many answers as you have questions. The average question on the site is not asked by an 8K+ rep user. It won't be everyone doing it because not everyone can afford to do it at all, and very few can afford to do it every time. – Ben Barden Apr 15 '13 at 14:52
  • The moderation is the flip side of the same thing. This is only something that can be done by users who have accumulated a fair bit of rep, and only truly high-rep users will be able to do it frequently. That would seem to keep the numbers of questions of this sort that require such attention relatively few in number. – Ben Barden Apr 15 '13 at 14:55
  • @BenBarden I will say that for myself personally, I would never offer a 2 day bounty. I would always offer a cheaper bounty immediately. – djechlin Apr 15 '13 at 14:58
  • @djelchin yes - but would you do that on every question you ask? Part of the cost in the immediate bounty is that it might have been answered within the first two days anyway. – Ben Barden Apr 15 '13 at 15:04

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .