The only times I have seen Tony mention reputation systems is when it would contribute to a solution to the topic/problem being discussed. I do the same thing when I see “probationary period” as a potential solution. It’s no different to you mentioning range maps and data accuracy every time someone says something that has a connection to them :)
2 Likes
Well fair enough, I do go off topic sometimes too. It’s just that in the google group the rep system thing really subverted a lot of threads.
1 Like
Could that be because it genuinely could be a solution to a lot of problems? I think so, but I also think it would be a difficult thing to implement effectively, and the potential for it to cause more problems than it solves also exists!
Bringing it back to the relevance here, inactive users manifest as a problem largely in the realm of IDs by duress users and those that “try the site and don’t continue for long”. There are other forms of inactive user as well, including but not limited to deceased, disgruntled and “too busy to partake often”. But as far as the bulk of the concerns, I think the duress and “tried but did not buy it” observers are most impactful. Potential improvements/solutions will likely come from a reputation system that weights the IDs appropriately, a probationary period (effectively doing the same thing), more patience and acceptance of the limitations of the data by existing users, or more clearer onboarding and reducing the problem from the source! [or some combination of]
2 Likes
I don’t know about you, but I arrived at the site at a time when NZ observations numbered in the 10s per day, if that. There wre days when there were NO observations. The ratio of inexperienced observers to experts was very low, compared to now where the experts are starting to feel overwhelmed by it all. My path to where I am now involved a considerable amount of 1:1 attention from the experts! I try to pay that forward as much as I can…
If you had arrived at the site as it is now, would it be the slight impact of a probationary period (or reputation system) that would have you leaving, or the impact on range maps that the inexperienced observers are having? And the impact from a probationary period for you would be very slight, as I am sure it would be waived for you by someone…!
1 Like
What if Community ID could just be redefined to include active users only? The inactive users’ IDs and comments could be left alone, but if there are conflicting IDs from active users, they wouldn’t be counted towards the Community ID? It looks like this was brought up as a possible solution in nathantaylor’s original post here. I don’t see the downside. Why shouldn’t the currently-active community decide what the observation is? Inactive opt-out IDs could be still visible, just not included in (active) community ID.
2 Likes
hmm, i don’t know. I’m somewhere around user number 4100 and back then it was almost all California so yeah, same story with me starting. Depending on the probationary period and how it worked, maybe i wouldn’t have cared. I don’t care that much about most inexperienced users messing up the range maps because i can just go around and negate their IDs if they are wrong. But when they turn off community ID, but don’t respond to questions either, it’s really frustrating. Honestly while i understand why they did it, i think id’s a bad idea to allow deactivating community for all of ones observations. I think it makes sense to allow people to do it on a case by case basis.
I don’t know why you all think i am opposed to a reputation system. I’m not necessarily, depending on how it is done. My post above was entirely related to one specific case of one specific person doing one specific thing.
3 Likes
That’s why I’m thinking a redefinition of Community ID. The inactive opt-out ID wouldn’t be in Community ID; it would be stated separately.
3 Likes
I would be greatly in favor of a discussion specifically to iron out details on how and if a reputation system might work. That way, we can assess the utility of the system and know whether we should see it as a solution for solving a bunch of problems or a burden that causes just as many or more problems than it solves. I personally like the idea and since @tonyrebelo is the one most interested in the idea, he should probably be the one to start it. If we can make it work, I would be in favor of the system, but coming up with a workable system seems potentially daunting.
2 Likes
This is getting pretty close to a feature idea that has been mentioned before (the idea of making IDs of specific users downloadable in the data). I don’t want to sidetrack too much, but I feel like being able to download data with identifications from specified users would solve some other problems and allow users to still have some control over the name they use even if the community votes against them.
1 Like
One way to do that would be to word-search all the threads in the google group for “reputation” and sum it up as a start. As I recall, the developers do not want to take on deciding who is an expert and who is not, and I can see their point.
4 Likes
It’s a good point, but I think it would still be useful if for no other reason to have a place to have the information and discuss it. It gets brought up often enough that, even if the developers aren’t interested in incorporating something like that now, there will be something well developed if they ever change their minds. Otherwise, there’s really no point in even bringing up the idea in these discussions at all. Good idea about summing up the areas where it is mentioned.
3 Likes
I’m going to try (and probably do a bad job) of summarizing the highlights here. I think I got at least the main solutions proposed, but please let me know if I missed something. I’m hoping this will give a good idea of what may need more discussion and if any of these would make good feature requests in the future.
As a solution to the problem of inactive observers, the following ideas were discussed:
Recruiting identifiers (as a general approach; currently implementable; probably doesn’t need more discussion)
Drawbacks: essentially none, currently implementable.
Using taxon splitting (to update taxonomy and change IDs; currently implementable; probably doesn’t need more discussion)
Drawbacks: essentially none, currently implementable.
Probation period (to minimize the impact of errant IDs from short-term users; not currently implemented; discussed at length)
Drawbacks: potentially isolating new users and could keep expert identifications from being incorporated. Potential workarounds: ability to waive probation period; not have the probation take effect unless it is a conflicting ID.
Inactive display on identifications for inactive users (not currently implemented; not discussed much)
Drawbacks: none that I can see.
"Please help fix" flag (for targeting observations with too many inactive user IDs; not currently implemented; not discussed much)
Drawbacks: none that I can see.
Option to downvote IDs of inactive users (not currently implemented; discussed somewhat)
Drawbacks: potential to negate good identifications and puts this power in the hands of the users who can downvote. Potential workarounds: only allow this ability to curators; only allow for inactive user IDs which conflicts with that of a current user.
Reputations system (lower ranked IDs don’t count as much towards community ID; not currently implemented; discussed, but probably not enough, create a new discussion about this?)
Drawbacks: difficult to develop and many potential problems if developed badly. Potential workarounds: needs more discussion.
4 Likes
I believe your first point should read recruiting identifiers, otherwise I am unclear how recruiting observers fixes anything related to inactive users, unless the goal is to just make their records a smaller percent of what is on the site.
When I proposed a probation period, it was not “to weed out people”… it is to minimise the impact that errant ID practices has on the community and data as a whole. We want them to stick around. Better wording would be “to minimise the impact of errant IDs from duress and absentee new users…”
1 Like
i think that would worsen the already existing issue of people just agreeing with things because either their friends posted them, or they just want more IDs on the list.
One interesting idea with a reputation could be, if you don’t use the site for a long time, like a year or two, perhaps your reputation starts to decline. You wouldn’t want to calculate the new ids constantly, but maybe if fled poster X is gone and poster Y disagrees with them, it recalculates with a new lower weight for poster X.
3 Likes
Sorry, meant identifiers as you suggest. Fixed.
You’re right. I was trying to put it succinctly, but what you said works better. I modified a little bit to say short-term users instead of focusing in on the two most problematic groups. Does that work for you?
@charlie Good point and good idea. I guess I need to add potential problems to the potential workarounds. 
As for the second idea, this might solve the problem. There is still the risk of potentially losing information from good identifications over time (I’m particularly thinking about IDs from experts who are less known in the community). That said, the community has generally been good at finding and agreeing with experts.
What might be nice is if there were a system based on how often your identification caused unresolved disagreement and your reputation went down depending on this number with the option of overriding this while a person is active. I suppose you could also have this apply only when the person is inactive. I know that experts often disagree with the community identification but, in general, the community accepts the ID in time. How often the person has an “improving” identification could be taken into account as well as a measure of how accepted their identifications are in the community. Sort of a improving vs. disagreement calculation as an estimate of how good the community considers your identifications.
The more complicated situations are if you end up in a mess like E. esula/virgata where the taxonomy of local experts hasn’t quite caught up with the taxonomy of taxon experts, but that complicated situation might be a discussion for another post.
3 Likes
I don’t know whether this is pertinent or not (the last entry): https://groups.google.com/forum/#!searchin/inaturalist/joanne$20russo|sort:date/inaturalist/4l0MjWlArFg/UdDVwGPiAQAJ
1 Like
The one thing I am not sure if I understand about this, is if you are actually correct, and are in disagreement, would that not cause your ‘reputation’ to be negatively impacted ? It assumes the initial ID is the right one, when most of what people want to fix is the opposite.
Yes, I guess you could recruit someomne to override the reputation hit, but you may as well just recruit them to fix the ID as it is.
2 Likes
Thanks for this list @nathantaylor! In terms of a probation period, it would make more sense to me to not have the “Agree” button available to new users until they’ve been on the site for some time (a month?) or made a certain number of IDs (50?). Instead, new users would have to type in a name to add an ID. This could help encourage more thoughtful identifications, without reducing the ease of use of “agree” for more established users. This is under the assumption that the main problem with inexperienced users is over-use of “agree” rather than addition of incorrect maverick IDs.
7 Likes
actually, that is a great idea! Experts coming to the site are likely to be entering names rather than agreeing… novice or duress new users are interpreting the agree button almost as a “Like” button! Given that names are so easy to enter (you only have to enter the first 3 or 4 letters of each part) I reckon get rid of the Agree buttons outright :)
If we eliminate the errant "Agree"s I think we’ll find that 90% of the issues surrounding inactive users will be gone. Tagging/Recruiting/Brigading, whatever one calls it, can mop up the remainder.
3 Likes