Stats are important as they are somehow reward for a person for high quality question/answer, so they help to keep overall good quality of the site. Other reasoning is that I guess few poeple can get proffesional work because of them, once I had a question from company that they needed temporary help with one issue, and they have found me by tag statistics. The site helps people to solve problems, learn new things, meet new people and helps to find employees/co-workers. So IMHO this should be forethought from the beginning and stated somehow in spec/docs.
As for bounties - few thoughts:
as you have mentioned - slowly gain points, lets say if you are trust level 2/3 you will get 1 point every month (sample number, may be dependent on trust level, however I wouldnât make it big)
when you solve somebodys problem with a bounty and gain it, you gain stat thay you solve it and also its points, those points you can spend for your own bounties,
for every 5 accepted answers you get 1 bounty point,
for every 10 upvotes (needs some time treshold) you get 1 bounty point,
for every 50 edits you get 1 bounty point (however this is somehow covered with 1st point where at certain trust level you get some points over time)
in one of discussions Iâve proposed something like a sponsor badge (get rid of ads), as a sponsor you can get addiotnal bounty point every month,
when you get a certain badge, it can give you some points (is the badge status already decided?)
To avoid gamification and changing bounty points to âreputation systemâ - bounty points should be private and not combined into rankings/statistics. When you visit userâs profile you should see solved bounties, but no available bounty points.
Do we really need bounties at all? Iâm not opposed to them, but I always considered them rather pointless on SE. Before going into detail on this, we should know itâs something enough people actually want.
The bounty system of SO was dysfunctional. It served no purpose, it gave people reputation that they had not earned and it had weird side-effects such as posts turning impossible to close.
In my experience of using both SO and SE sites, bounties have a high chance to result in a better answer appearing than⌠basically anything else. You canât exactly say âbumpâ in the comments or âI have this issue too, wonât someone help me?â as these are basically spam comments and are not allowed for good reason. You canât âbumpâ for free.
Putting up a bounty is a way to attract attention of those who are looking at the list of active bounties, which is kind of like saying âI have already contributed to this site, and here is the reputation cost to prove it. Please post a (better) answer if you can.â
I think this system makes sense, and so far I havenât met with a problem of only low quality answers posted by ârep seekersâ to feel like itâs more problematic than useful.
I donât know how weâre going to be paying for bounties on Codidact if weâre not going to have the same reputation system, maybe with some sort of bounty tokens we could earn for participation separate from the main reputation point system, but I do think there needs to be some cost which reflects that the user has contributed enough to gain that ability.
The SO system could have been improved by making it more accessible while limiting it on the other end. E.g. a user can give 10 rep bonus max. There can still be big bounties if lots of people want to know - but thatâs still fair then, according to the rules of offer and demand.
However, you can probably forget about it.
I mean, you see that your whole number thing, â5 accepted answers you get 1 [âŚ] pointâ etc., is pretty much SOâs reputation system, right?
You can compare 2 users by looking at how many bounty points they have.
Somebody would just write a browser plugin to show that number next to an avatar.
But the current state is to get rid of almost all gamification and have people answer because of the goodness of their hearts. In game design terms, thereâs no resources to spend on something like this.
In my experience of using both SO and SE sites, bounties have a high chance to result in a better answer appearing than⌠basically anything else.
So we might just need to be better at providing a good incentive for âbetter answersâ.
Maybe people can vote on questions they feel are interesting that are un-convincingly answered or not answered at all. If enough people declare a question interesting the system somehow makes it more visible and the best answer (by a measure yet to be defined) gets a badge. If enough good answers are provided there could be gold/silver/bronze medals for providing highly sought after answers.
As for gamification, I think the last sentences of my post should sove this - you cannot change this into reputation system if itâs not visible.
Bounties help to atract attention to certain questions, if there are tons of questions then one of yours may be omited, but if you need somebody to take a closer look at this you can set a bounty and feature it.
Itâs worth to mention that this system has nothing to do with reputation. Couple of more thoughts:
bounty points should be balanced, so that on one way they are easy to gain so that an active user will have some to add a bounty on interesting question, but on the other hand the site shouldnât end up with bounty queue of length 1000.
maybe one bounty point will be needed to feature question for one day, so if you set up bounty for 10 days, you will need to spend 10 points.
The rep gained from winning a bounty on SE can be nice, but arguably the bigger benefit is the increased visibility for the question and all its answers â more views, more votes, with luck new answers, maybe edits. The authors of those posts benefit but the whole community benefits if the Q&A is improved.
Instead of thinking of bounties bringing personal benefit, letâs think about approaches that draw community attention to suitable (whatever that means) questions. Communities can even do this themselves without any software support; SE communities have had various types of question drives, âbest answer of the quarterâ contests, and other ways of drawing attention to particular questions (or lists of questions). Weâre not going to have bounties in MVP, but our communities can still organize drawing attention to deserving posts.
Letâs do that first, and then work with our communities to see what our users really want here, and go from there. Reputation is not central here like on SE. If we do something like bounties, itâll be very different. Letâs not design in the absence of experience with the Codidact platform.
This could work if somehow you could identify with high enough precision, what you personally find interesting, and then weâd have to tune the system well enough to show that user questions which would not turn them away too soon, and I think this is quite a task.
Even so, I think itâs an interesting idea worth investigating, to encourage everyone to vote on what they feel is [not] interesting without any incentive which promotes voting one way or the other, but maybe with an incentive towards some âavid reviewerâ badge to somehow turn this into a useful habit (which shouldnât really give any abusable privileges, just be another badge in the list of earned badges). And weâd also have to find a way to filter out noisy random-voters which will inevitably crop up at some point.
Maybe we could try different experimental options and explain how they work so people could try them and see if they like one or the other, kinda like Steam runs their marketplace filters and suggestion experiments: https://store.steampowered.com/labs/
Also, and I think most commonly this is the case, most users who could answer a question donât get to see it because it already left their page of recent questions tagged as some tags they follow, due to the tremendous flow of questions on SO every day. With no way to make sure your post stays visible other than putting a bounty on it, itâs no wonder good/interesting questions get missed, but once bountied, get answers.
I donât think bounties worked at all on Photo-SE â at least not for what users expected them to do, which is get my question a better answer. Often, it there just wasnât one out there. In the best case, too obscure, and a week of extra visibility didnât make a difference â but sometimes the asker wanted the truth to be different from the (correct) answers they got already.
So to me the primary use of bounties was to give an extra reward to answers which clearly had extra effort and care invested, hoping to encourage more.
In my 5 years of active posting at Photo-SE, I think there were only maybe a dozen bounties posted, about half of which offered by you. In my early activity at the site, you awarded me a bounty for a nice answer, which I have to say, was a nice personal kudos for a low-rep user at the time.
So anecdotally, the bounty system encouraged me on an answer (although I would have still contributed regardless of your bounty at the time). It was nice, but in the grander scheme, with so few bounties in the last 5 years at Photo-SE, I canât say it moved the needle as far as participation or retention incentives go.
But how does that help? The person who offered the bounty isnât happy with the current answers, and after the timeout expires, they still wonât be (and may be annoyed that their bounty was awarded to an answer they didnât find satisfying).