141

Here's one to get shot down: How about letting bounties to be set using the reputation gained through the use of another site? So, for instance, I'm a programmer with a high Stack Overflow rep who hates using computers. Wouldn't it be useful for me to be able to exchange some of my excess Stack Overflow credit to incentivise a question on Super User?

Or perhaps I'd like to highlight a question on Meta Stack Overflow, but I don't spend my time there.

6
  • 7
    @AdamRackis: I don't participate in Meta very much. It would be a bit odd if, despite not having demonstrated much interest in helping run the site, I could give up some of my more easily-acquired www.stackoverflow reputation to draw more attention to my requests. Giving up reputation is, in part, a way of telling the more helpful people, "I've given a lot to this site, so I am especially deserving of your help." Mark does provide a justification for that, but I can see my point, the reverse of that justification, as being similarly reasonable. There are possibly engineering issues, too.
    – Brian
    Commented Oct 20, 2011 at 21:57
  • 4
    @Brian I don't think bounties are a way to say "I've given a lot to this site, so I am especially deserving of your help." I think bounties are a way for experienced users to spend some rep to buy extra attention on really hard questions. As Lance said, the main risk would be spending your SO rep on unworthy Cooking questions, but as he said, limiting your ability to do this to technical sites would be a good protection against that. Specifically, if you have a high SO rep, you can probably be trusted to not bounty trivial questions on meta. Commented Oct 21, 2011 at 19:50
  • 5
    I went to meta today to make this same feature request! I have a ton of rep on stack overflow, but I'm almost always able to answer my own questions there by searching. Now sometimes I have a really annoying problem with an iphone, for example, and I know very little about apple products so I can't build up enough rep to set a bounty on that stack exchange site. It would be a nice reward or "thank you from stack exchange" for my many hours helping people on SO if I were able to get a larger amount of attention for issues that are really bugging me on other stack exchange sites.
    – wim
    Commented Apr 23, 2013 at 6:36
  • 1
    possible duplicate of Can I Borrow Points from one StackExchange Account for a Bounty on Another?
    – Pollyanna
    Commented Jan 21, 2014 at 17:31
  • 5
    This would be so good for cross pollination. For me I had my first DataScience question that I used up 100 of my 101 sign on points on a bounty. Now I cannot even comment because I don't have a rep of 50. I have another DataScience question I would give up SE points but I can't.
    – paparazzo
    Commented Oct 26, 2015 at 12:48
  • 2
    I'm so frustrated that this is not a thing. I have close to 10k rep on StackOverflow and an important question of ServerFault needs more attention, but I cannot boost it, because I'm not very active on ServerFault. I contribute on the network and I'm still restricted. Commented Dec 2, 2022 at 12:51

13 Answers 13

113

This makes a lot of sense to me. Most of us are specialized to a large degree, making it hard to build up adequate amounts of rep at all of the sites. The site where you aren't an expert is the site where you'd most likely find a bounty to be useful.

4
  • 9
    ++1. It makes perfect sense that someone who needs help and is willing to offer rewards for it won't have the skills in that area necessary to build up the rep to provide those rewards to others. Being able to exchange rep for bounties between sites is a perfect system for people to earn where they can and spend where they need. A "pay it forward" system.
    – Questioner
    Commented Jul 3, 2012 at 6:07
  • 3
    This makes a lot of sense to me. Most of us are specialized to a large degree, making it hard to build up adequate amounts of experience at all of the sites subjects. The site where you aren't an expert is the site where you'd most likely find an inane question to be useful.
    – Jeff
    Commented Jan 30, 2016 at 21:18
  • 2
    @Jeff: ... which is why the association bonus allows anyone to upvote inane questions anywhere, no cost. Commented Jan 30, 2016 at 21:37
  • 1
    @NathanTuggy an upvote is slightly less intrusive than a bounty...
    – Jeff
    Commented Feb 5, 2016 at 2:56
65
+75

Cross site moderation

If you're a moderator that should send shivers down your spine. Thought finding sock puppets and voting rings was bad enough? Now you have 100 sites on which your voting ring could be centered. In theory it can be easy to find if someone gives bounties in exchange for upvotes on one site. But what if Bob gives Alice a bounty on UX for upvotes on Stack Overflow? As a UX mod, all I see is the UX bounty. Seems legit, I'd never think anything of it.

This would require new, cross site mod tools. Which should be scary for all sorts of reasons; now you're querying 100 databases (we get rate limited enough in the mod tools thanks), potentially having to investigate crap on sites you're not a mod on, it's a mess and all for incredibly little gain.

Bounties meant something in my day

When I see a bounty, someone with rep on that site set that bounty. Rep is a rough measure, but usually that means when someone sets a bounty on UX, they did it because it's a good UX question. I really don't want to start seeing people cast 50 rep bounties on every single "design my interface" question because they have more rep on SO than they could give away in a billion years. The barrier to entry is useful. Bounties are supposed to hurt, if I can give away rep from a site I don't care about or a site I have far too much rep on I can just spam rep on anything for any reason, give it to friends, whatever. I'm not going to use my extra 8k MSO rep anyway

Oh god the rules

The obvious solution is to lock this feature down tight. A rule here, two rules there, some rate limits in the back, some mod tools in the truck. But at what cost? Seriously, if you have to add a dozen rules to a feature before it gets released your feature needs major justification, much more than "wouldn't it be cool". Given the suggestions for rules so far this tool sounds like it would:

  • Be very frustrating to use and have multiple rules you need to know if you use it more than once
  • Not actually prevent much abuse because cross-site moderation would be a nightmare even in the case of a single abusive bounty

So no, I think this is a really bad idea which needs to be propped up by lots of really bad rules for amazingly little gain. Let's just keep rep separate, keep bounties meaningful and relevant, and not give moderators massive cross site bounty rings to deal with all for the sake of bountying "how can I make pasta as a programmer".

7
  • 3
    I think you're losing sight of the potential upsides to this change. Just imagine the hilariously disruptive impact on smaller stacks' reputation economies when Jon Skeet takes an interest in them and suddenly most of the rep on those sites comes from his bounties.
    – blahdiblah
    Commented Mar 22, 2013 at 23:47
  • 19
    Even though I really wish we had this ability, I'm forced to upvote this answer. What if this feature were limited to trusted users? That is, I could award a 200 pt bounty on UX, but only because I'm >20K on SO, and pay for it from there. Commented Aug 15, 2013 at 14:05
  • a simple program that monitors it would work wouldn't it? Something that checks a bounty by going through interactions between the two and alerts a monitor if necessary. That would get rid of the need for cross-site mods, right?
    – tox123
    Commented Jul 4, 2014 at 19:49
  • 8
    Re "bounties are supposed to hurt" - I'd be happy with this feature even if I could cross-spend rep for bounties only at a 10:1 exchange rate say (so I would spend 5000 of my easily earned math.SE reps for a 500 bounty on my urgent serverfault.SE problem, for example) Commented Mar 20, 2015 at 8:25
  • Is this the "official answer" for status declined? I.e., if I still believe this is a good idea, should I try to address the points raised here - or is that fruitless, as the administration may have different reasons not voiced here?
    – FooBar
    Commented Jul 16, 2015 at 8:14
  • 2
    IMO it's not simply a "wouldn't it be cool" - its pretty vital ask even if there are implementation and scale/moderation gotchas. While my interest is to use points from one technical site on another I do not see the objection to using SO points to bounty a cooking question. Points are currency of sorts and one should be able to use US$ to buy something from a Japanese website! Sure maybe extra penalty is in order (higher exchange rate).
    – nhed
    Commented Jun 29, 2016 at 22:14
  • 2
    This seems like it was so obvious to me. If this system came into fruition, it would essentially create a huge market for reputation selling. Scrape up a bunch of rep on an active site, and then sell it to people (or give it to your own sockpuppets) by awarding them a bounty on whatever site they want.
    – JMac
    Commented Feb 7, 2019 at 16:37
37

On the first glance, this looks like a good idea: If I earned reputation hard by answering questions in my specialty field, why shouldn't I be allowed to spend some of this for my questions in the fields where I'm not so good, and really need the help?

For example, I've answered lots of questions on Stack Overflow, but now I have a cooking question (and I have about no chance to earn significant reputation on Seasoned Advice, this is why I need the question answered), so why shouldn't I be allowed to use some of my SO reputation to put a bounty on my cooking question?

This would make the cross-site reputation a kind of global SE currency (but usable only for bounty purposes).

The problem with this is that reputation on different sites has actually different value. On some sites it is a lot easier to earn upvotes with an answer (or a question) than on other ones. (The same problem is existent within Stack Overflow, though: Answers in more frequented tags get a lot more votes than answers of similar difficulty in small tags. But it is somehow deemed okay.)

Additionally, the idea is that each site can develop its own community, and the site-reputation should somehow measure the standing in this community. Receiving a bounty from someone outside the community does not relate to this.

7
  • 3
    You raise a good point; however, I think a good compromise would be to just limit how much rep you could use for this purpose. Say, up to 150 per question, 500 per year? 500 per 6 months? I do see how frequently dropping 500pt bounties on cooking might disrupt the "exchange rate" of their rep, but I think there are good ways to limit this, while still getting the benefits this would provide. Commented Oct 20, 2011 at 15:38
  • 9
    Maybe also needing to pay more "foreign reputation" than the bounty size could help (e.g. you pay 100 reputation for a 50 rep bounty or similar). Commented Oct 20, 2011 at 22:36
  • 1
    I agree with this post. It's like trying to convince the US government to let you spend a little Canadian money in the US every once in awhile. They're both called dollars, but they're not compatible. US dollars (site #1 rep) buy US goods (site #1 bounties) and not Canadian goods (site #2 bounties). I don't think we have any appropriate method for currency conversion either; there are no "rep banks" to handle the supply and demand :P
    – user154510
    Commented Nov 18, 2011 at 16:01
  • 1
    And certainly it seems almost abusive to gain rep at site #1 and then go to site #2 and draw attention away from other questions with a bounty you didn't earn the right to offer.
    – user154510
    Commented Nov 18, 2011 at 16:02
  • 1
    What about creating some self-sustaining system for exchanging reputation between sites, just as our bank system can be used for exchanging currencies? It'd be difficult to set up at first, but soon the exchange rates would converge (which would additionally provide very interesting information about the sites).
    – Petr
    Commented Jan 26, 2014 at 14:16
  • Arbitrage scenario? Commented Jan 31, 2016 at 2:05
  • Could smaller stacks be protected from upvote inequality by using the ratio of the average number of upvotes on an accepted answer in each stack as the basis for the conversion rate?
    – cjm
    Commented Jan 31, 2016 at 8:05
12

I think this could lead to some potential abuse (basically using rep on one site to boost up an account on another site).

It's an interesting idea and the way you described it as a way for people who are invested in one site to get a potentially quick answer on another site without having to go through the effort of building up rep there sounds logical. I just don't have a very good feeling that it would ever even get seriously considered.

2
  • 5
    Bounties always have had potential for abuse. Effectively having to have your accounts permanent disassociated would seem a reasonable deterrent given the threat. Commented Jul 20, 2009 at 1:52
  • @TomHawtin-tackline What would you say about setting individual "divide factor" for each site - e.g. "1:5 for TCS.SE might mean - giving 100 bounty on TCS.SE costs 500 from other site" ? Commented Aug 14, 2012 at 18:41
11

I totally agree, though I'm not surprised to notice that it has already been proposed and was just ignored.

I'll copy+paste my duplicate post here:

Wouldn't it be nice to have cross-site bounties?

After all, if I am a developer I am likely to have points on SO, and not many on SF, and vice-versa.

I think probably everyone would benefit from such a feature.

Update: I agree that cross-site reputation, privileges or many other things would be bad. Bounties however, are a completely different topic. You "pay" some earned rep to the person that will answer a question that isn't get an answer otherwise. You are skilled in a field, and you need help in another field: that's the founding principle of the entire economic system in general.

Cross-site privileges would mean awarding powers in a field to people who do not have earned it in that filed, and that is bad.

Cross-site bounties means allowing you to sell your milk to buy some meat.

Edit: one possible addition to that, just in case too many people frown upon such a feature, could be not to add the extra 50 rep for the cross-site bounties.

10
+100

I think that'd work good for the technical, even somewhat related computer sites, like SO, Ubuntu, Apple, etc., but not so sure about the non-technical computer sites like English, Cooking, etc.

Since you're buying the bounty with rep that is supposed to represent some level of knowledge, I think the cross-site bounties would work well for interrelated sites. If you have the rep on Stack Overflow, then that should show you have some level of technical acumen, so that you can be trusted on the other technical sites like Super User to bounty relevant questions.

6
  • 1
    @Brad, that was true at the time you wrote it, but not anymore.
    – Pops
    Commented Oct 4, 2010 at 18:55
  • 1
    @Lance - do you still stand by this answer that it wouldn't work for SE sites? I'd reeeeaaaalllly like to offer a bounty on some apple.SE questions, but have insufficient rep. This feature would be really useful imo. Commented Oct 19, 2011 at 19:28
  • @Adam, modified to be more current, you should put a bounty on this question. Commented Oct 20, 2011 at 4:51
  • That would be a great compromise. Let users open bounties from SO rep to SU, Apple, ServerFault, Programmers, etc. Can you elaborate on why you think it wouldn't work on Cooking, English, etc? I "feel" like I agree, but I can't really figure out why. Commented Oct 20, 2011 at 15:40
  • @Adam, ok, edited. Commented Oct 20, 2011 at 15:45
  • Good point—agree. "trusted to bounty relevant questions" Commented Oct 20, 2011 at 15:50
9

I think this is one example of why fragmentation of the QA sites is not a very good thing (see also What could be done to serve domain specific "communities" better on SO?) and perhaps some tighter integration (perhaps even a reputation shared between all sites) could be better.

I predict this answer can provoke comments about how I cannot compare reputation from Stack Overflow with reputation from Super User, as the fact I understand one does not mean I understand the other. This, however, is in my opinion no different from the fact that while I have quite solid C++ background, I do not understand a bit about AJAX, Ruby, SQL and other things questioned on SO, yet this is not apparent from the reputation I have on SO, and still this does not seem to cause any issues.

6

Even though I support this request, I think there is a problem to this suggestion, which is abuse.

The moment a user register on a site, he gets 101 reputation, those can be used as bounty on any other site, without affecting his reputation on any of the sites he cares about.

I think he later can delete and re-create an account on the same site, and getting 101 reputation again.

The solution Could be to only allow transferring of reputation for bounties, only from sites that has reputation over a specific limit (lets say 400?).

And to put a condition that setting bounties on a remote site should not make the current site reputation fall below a specific limit (lets say 300).

This would also limit the usage of this feature to only users who are somewhat experienced.

1
  • 2
    I think the apparent need for such a complex system to make this feasible is a good indicator that the feature shouldn't happen at all. It's more complicated than it's worth and I still see more options for confusion, rep inflation and abuse than good use.
    – Zelda
    Commented Mar 22, 2013 at 12:00
6

The problem with this is that it's about more than reputation is about more than just privileges, and it's more than just "points" - the reason that it has any value at all is that it is about your reputation.

If I can gain reputation as a cook for my prowess as a programmer, then reputation no longer has any real value at all, and the entire system devolves completely into a game. The reason it's not entirely a game at the moment is that in order to "win" the game, you actually have to be really good at whatever field it is that you're in.

Removing that restriction would entirely ruin SE.

If you think that one SE site is related enough to another site to deserve cross-site reputation bonuses (such as cross-site bounties) a better approach would be to allow some portion of all reputation gained on one site affect your apparent reputation on another.

I personally feel that that is not justified, however - if the two sites are so related that reputation for one can count as reputation for the other, they should probably not be separate sites at all.

1
  • 7
    Being able to give bounties cross-site wouldn't give awarder personally any reputation on the other site (unless fiddling the system). I was looking at bounties/rep as goodwill. However, reputation across different sites doesn't have a 1:1 value. Taking reputation from Stack Overflow and dumping it on a new small site would be unbalanced. Commented Mar 17, 2013 at 16:38
4

I think that in general it is good idea.

Question is about details (whether implement 1:1 credit transfer, or divide with factor different for each site e.g. "bounty 100 on TCS.SE is worth 500 of bounty from another sites", limits... etc).

Idea behind this is about exchange of help, between people keen on different domains.

Let's take specific example:

  • A is able to help people on Unix.SE and earn reputation there.
  • In the same time A would love to get more help on History.SE or other sites. (which are related to areas which are his/her weaknesses)
  • Thanks to such exchange, A could help others on Unix.SE, reward people on other sites, like B
  • Now B can do vice-versa - B can earn reputation points on History.SE and go to Unix.SE and reward there
3

One should be very careful to reject a feature with big potential benefits just because it opens the door to some abuse. One should worry more on the users that do legit use of the site more than the abusers. Also, SE exchange already have systems in place to avoid that, and that it is granting power in proportion of a measure of how trusted a user is (reputation).

I think this feature would be extremely useful! But it should be a privilege earned at a significant level of reputation and requires an exchange rate. There is no need to reinvent the wheel, the bases for such exchange rate could be the same used to weight the "arbitrary hotness points" of questions across different SE sites.

I'm a geoscientist, and I got good reputation on Earth Sciences SE. But now and then, I have an urgent need to solve a programming, biology or English Language question. The benefits to me as a SE user, would be greatly increased if I can use my Earth Sciences reputation to set bounties to questions on topics outside my field of expertise.

I think that's the whole point of SE: sharing knowledge. But this is undermined as the SE network gets fragmented in various specialist sites. Because at the end I will be able to earn reputation only in the very topic I'm specialist on, so that if I can't solve my own question, it is unlikely anyone else will.

7
  • I think meta.stackexchange.com/a/173012 makes an extremely solid case that the danger is not "some" abuse, but "a ridiculously large amount of incredibly difficult to handle" abuse. Commented Feb 17, 2019 at 8:02
  • @NathanTuggy But I think the same question heavily underestimate the benefits that this feature could have. And for that matter bounties and and the very reputation generate the danger of "a ridiculously large amount of incredibly difficult to handle" abuse. It is know that some people have agreements to upvote each other and stuff like that. Commented Feb 17, 2019 at 15:46
  • a) If there is a really major benefit to the sites, or to a lot of users without harming the sites, then you should outline those; what I've seen suggests that the bountied questions would mostly not be all that great, which is less of a major benefit and more of a minor annoyance. b) See the answer again: voting rings are well-known, and there are standard practices and tools to handle them fairly effectively, but all of those rely on moderating a single site at a time. So it's in the context of already knowing how to robustly handle voting rings that the answer calls the problem "scary". Commented Feb 17, 2019 at 22:56
  • @NathanTuggy I get the extra difficulties, but I don't think planning should be done around abusers, but around legit users. Also I think you are erroneously equating good questions with difficult questions. If I offer a bounty on a question within my area of expertise it will be probably a difficult and extremely specific question, not necessarily a good question for the community. People with little reputation are often the authors of very good questions. Commented Feb 17, 2019 at 23:04
  • It's true that "[p]eople with little reputation are often the authors of very good questions". However, most such good questions don't need bounties, because they're good enough to rapidly attract good answers. So we can throw out a solid 95% or more of good questions, right off the bat. Those are just not relevant. If you look at the remaining questions I don't think the picture is so rosy. Remember, we're dealing with people who a) are experienced on one site (well, at least 75 rep, or 400 rep, or whatever the chosen threshold is for this new privilege) [1/3] Commented Feb 17, 2019 at 23:18
  • b) have put little or no effort into engaging with the community on the site they are now asking on (and therefore have only half-decent guesswork on what makes a good question), c) did not ask a sufficiently tractable question to get an answer within 48 hours. These questions are unlikely to be trash. But you admit yourself that they wouldn't necessarily attract much good attention to the site on their own. And the site has little reason to suppose the asker will spend any more time there in future, so what's the value of allowing the asker to hijack answerer attention? [2/3] Commented Feb 17, 2019 at 23:25
  • As far as designing for abuse… it's negligent to ignore easily-foreseeable and severe security holes that are highly likely to require extensive mitigation. That only guarantees that they will need to later be patched, after much damage has already been done, and with much greater difficulty, as the system is already in place and less flexible. Designing solely for the happy path is bad engineering practice, simple as that. [3/3] Commented Feb 17, 2019 at 23:32
3

I've been contributing to math.SE for a few years now, answering where I could. Today I teach mostly in one field of maths, I teach basic stuff but I really would like to know more, so I'm more likely to have questions than answers.

Problem: this field of maths has its own separate SE.

So I've got 25k rep on math.SE, that I can't use to bring attention to the few questions I have, just because people have decided that these questions belong to another site. Makes no sense to me.

So I totally agree with this idea. If people are afraid that there might be abuse, as shown in this question (which is a duplicate but was received totally differently for some reason), then cross-site bounties could be limited to sites with related topics.

1
  • 1
    Or we shouldn't have made two site that cover the same or a similar topic...
    – Luuklag
    Commented Feb 12, 2020 at 10:48
2

In practical terms this would be very interesting if applied across trifecta sites (META has nothing to do with this), and limited in amount and purpose. Namely bounties.

Observe:

I need my website to do thing X. That's a programming question. Yay, I'm good on StackOverflow. Oh no, wait, it's totally dependant on server configuration. I guess there's ServerFault. I can ask there...

Several days later and my question has only had 4 views. I want my question to get more views, so I could offer a bounty. I'm not too good at server configuration and all the questions I could possibly answer on ServerFault to build any sort of semblance of a reputation are really hard and above my pay grade. The questions I could answer at my level have been answered for the last 6-7 years.

Many Trifecta users are full-stack developers with some DevOps on the side, and that makes us travel from one end of the spectrum to the other, but maybe we're only familiar with one end.

tl;dr: Would I give up some of my StackOverflow rep to help a ServerFault question get some attention? Yes. Would I do it for a 2:1 deal? Yes. 5:1? Still yes. I need them answers.

You must log in to answer this question.

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