Personally, I delete errors fairly often since I think it’s confusing to leave them up, especially when I’m convinced there’s a ~99% chance that I was totally wrong. Retaining errors doesn’t help me, and I don’t think it usually helps others either, although it is likely to distract-- I know that I personally experience a sudden lack of confidence in identification when I see signs that other knowledgeable people were confused and debating something (as evidenced by withdrawn IDs).
In my view, withdrawing vs. deleting pretty much comes down to “is there a chance that I will need to hit ‘restore ID’ in the future, or can I junk the whole identification”. If I know I was wrong, I don’t want people to get confused by my original suggestion, so I will remove it entirely. If I’m not sure, then I’m far more likely to just withdraw it, and later discuss it with the person disagreeing.
Which is all to say, although it may look a little weird and there certainly is some chance someone may do it just to save their ego, there are other reasons why it’s sensible to use that function as well. Or at least, in my opinion there are. I’m sure someone will disagree, but it’s up to individual preference.
I suppose the absolutely ideal thing would be to withdraw and leave a comment explaining why it isn’t the thing you thought it was, but as someone who is also a a bit of power user, I don’t have time to individually address each thing like this; I would assume that is the case for the person you’re interacting with as well.
They’re probably just scooting along trying to do the most efficient and sensible thing for them, even if it’s not readily apparent to others. Most moments where someone behaves in a rather impersonable way are a result of constraints like that. It’s hard not to take things personally when you’re on the receiving end, but the person doing it likely has no idea that it looks that way on the outside.
I personally don’t see much bad action here, they could have lots of notifications and missing your comments, and then, when they finally see an id, they delete theirs, which isn’t any worse than withdrawing. I’m not sure you can really tell somebody did it “for the leaderboard” or because they wished to do that, I know some users who add wrong ids because they don’t care that they don’t know enough to make those ids, and likely not because it gives them a certain place on the leaderboard. Trying to contact them via direct messages could be a way to see their response.
I’d argue that reducing people’s confidence is usually a good thing, in the context of IDs. If they really know why they’re sure, no problem, but if they’re leaning heavily on “this person is an expert, so they must be right” then agreeing adds little value, and showing that even knowledgeable people make mistakes can be eye-opening.
If I realize I made a mistake but nobody helse has interacted with the obs in the meantime, i.e. there is no history, I delete.
If there is history of somebody else adding their IDs, I withdraw, so the history keeps making sense.
If you want to make sure they don’t get away with erasing the evidence of their mistakes, consider commenting along with your corrective IDs - something like “@ username, this is Species Y not species X - you can tell the difference because it has (distinguishing feature).” It’s not a perfect solution, but it at least leaves some indicators for other users that that user may be a problematic identifier of those species.
I confess I have deleted some of the more embarrassing IDs I’ve made in the past - the ones where I look at it later and can only assume I must have been identifying in my sleep. I’ll also delete mis-clicks and very fresh IDs when I have a sudden attack of uncertainty just after posting them.
My general rule though is that if someone has already interacted with one of my IDs in some way (agreeing/disagreeing, tagging me, comment, etc) I’ll withdraw rather than delete so it still makes sense.
Same!
I don’t ID for the sake of the numbers, but I find it motivating to be able to see my own work and the work of others laid out like that. As with most of my generation, I have the instinctive “number go up = good!” reaction, and my brain needs all the dopamine it can scrape up. I’d probably go back to video games for making numbers go up, and that would be much less useful to others.
I’d support this! And expand it slightly, to include non-ID comments made after an ID as well as subsequent IDs - both should prevent deletion.
I may be wrong, but my impression was that only maverick IDs affect this.
There are unfortunately a couple of very problematic plant identifiers with 100,000+ identifications. I have no idea if they’re motivated by “winning” or actually think they’re helpful, but they’ll agree with every identification for certain species no matter how ludicrously wrong it is. You’ll usually find them in the observations of common, widespread “weedy” species that don’t get a lot of interest from more competent identifiers.
A couple of them also do a lot of damage in the cultivated observations, because those almost never get reviewed by anyone.
This has been a massive problem for a certain user who has gone inactive now seemingly. He was the top identifier of a taxon and still is of most of its descendants mostly by spamming agrees, even on observations clearly not of the taxon, blank photos, etc. and has been very aggressive when called out numerous times. Eventually he began deleting IDs (even as he continued to make more, in my opinion malicious, erroneous ones), and I began collecting screenshots. Because of the ranking system, many people looking for IDs tag him rather than the actual active identifiers of the taxon who can give a proper ID. It’s been a whole mess honestly, there’s thousands of observations he’s RG’d that still need to be corrected and likely won’t be for a long time – he has so many mass IDs that no one can reasonably sort through them while giving a thoughtful ID without it taking at least many months of continuous daily work from multiple people. I think the delete option is helpful for misclicks and such, but it does make this problem even bigger as people don’t see the mistakes when they’re corrected. Maybe it’d be good for a list of mavericks people have (and have had in the past) to show up somewhere on their profile? If someone has a large amount of mavericks with their IDs, it could be a good hint to people that they may be a bit reckless – of course that doesn’t resolve the problem of needing to fix all of the observations to get the IDs to maverick in the first place.
The problem is that it’s an arbitrary reduction of confidence, which is a bad thing. If I see an observation and immediately think “hey, I know this one” and then I roll in, see some withdrawn IDs by experts, and wonder if there’s some information other people who know more than me are already aware of, but that I don’t know…I may not leave an ID on something where I otherwise would’ve been 100% confident, and 100% correct.
It would be great if people who were confidently mis-identifying things had a loss of confidence, but this isn’t the scenario. There was no mis-IDing going on. They could’ve just typed the wrong thing, or clicked the wrong thing, or gotten two species backwards, and I don’t need nor want to know that that happened unless I want to get confused.
So yes, maybe for a total amateur who is predisposed towards what people like to call the Dunning-Kruger effect, less confidence would be better, but that’s not the case for everyone. It’s harmful in many cases, and should only exist as a consequence to an action.
And of course, there’s a balancing act between acknowledging that experts know many things that you don’t without twisting that into thinking that they know everything and are correct 100% of the time.
I guess my interpretation of iNaturalist’s guidance on identifying is for users to rely on their own experience/knowledge when IDing. To me, this means focusing on the primary evidence presented in the observation (photo, sound, location, date/time), not the inputs of other users unless they present specific evidence like traits to look at, links to resources, etc. Other users ID ideally reflect their experience.
Relying on other users’ identifications to make IDs can cause problems - it’s at the root of the “agree” conundrum where some users agree to IDs based on their overconfidence in other users’ expertise and not their own. We then have observations that have more support for an ID than they should. The scenario described above seems like a kind of flip of this - seeing other users might have been incorrect for some (unknown) reason, reduces confidence, perhaps unnecessarily.
For myself, if I found an observation and was seriously discouraged from adding an ID because I saw others’ withdrawals, I would take that as an indication that I shouldn’t be making an ID myself because I don’t have enough expertise/knowledge.
I also think a good approach could be to add a comment/tag some users and ask why there are withdrawn IDs if a user thinks this means an observation is particularly tricky or difficult to ID - this could help users learn or clarify a difficult to ID observation.
I used to do the speed identification when I first joined Inat 2 years ago. I would ID random bird species and just hit “Agree” with every RG observation to get to the top of the leaderboard.
See https://www.inaturalist.org/people/leaderboard/2020/6 https://www.inaturalist.org/people/leaderboard/2020/7
I realized that this was having worse repercussions for me, and other people that had identified this(people were complaining about having 1000 notifications). I would recommend blocking that user if the action continues, especially if you continue to feel hurt by that treatment. If this also continues to happen, you can flag the identification before the user is able to delete it. However, this behavior is not violating the Community Guidelines, unless the user is purposely adding false ID’s.
I also apologize if I have done this in the past.
The top IDer is extremely useful when you want to find the most knowledgeable person to tag for identification of a certain group. I think most people use it for that purpose and not some competition
First Observation: Let me pay extra attention to this one
Next Ten: I know exactly what that is!
Another Nearby: Saw this 10 times before! click & misidentify
Not the best feeling, but it definitely reinforces open-mindedness to:
Is this a different species or just a variation?
Confident but could be one I don’t know
Most of the identifiers whom I admire and seek for help seem to have this cautious confidence.
Regarding ID/withdraw, I think withdraw is helpful to other users. Your observations are your own, of course, but seeing someone struggling with an observation would encourage me to:
Look more carefully, maybe tricky, a hybrid, etc.
Leave a note if there’s an easy way to tell them apart
With delete, there’s no history of the struggle to ID and the assumption is that the observer just knows/agrees with the ID.
Lastly, the iNat suggestions, books, keys, etc. can exaggerate how “expert” people appear on iNat. I’ve seen this with in-person naturalist events like a bioblitz. Respected book authors and top iNat identifiers don’t just point to every single living thing and name the species, sometimes it’s just genus or family. They also have to look some things up and can make mistakes, just like everybody else.
This is one of the main reasons I mostly limit my identifying to a specific geographic range (except for when there’s a widespread species I know with so many mis-IDs it’s driving me nuts).
It’s pretty easy to learn a species + all its lookalikes in a single county, or even state, but people who identify a species worldwide have to be much more careful.
Especially since there’s so many regions that just aren’t well-documented, so the lookalike species might not even have iNat observations yet.
