I do wonder how many of these are legit, though - I kind of wish the results would get delayed longer so there’s more time to go through them. I haven’t even begun to be able to do IDs at all, I’ve been so overwhelmed with copyright stuff, but I know there are tens of thousands of incorrect RG observations just in the ones I’ve personally scrolled through.
7 Likes
Where do you see that, Vireya?
It’s expected! Students using iNaturalist often behave badly. The CNC magnifies the issue because of its extra volume, but any high school or even university assignment has its clueless members and intentional cheaters.
6 Likes
After the experience of the last few days trying to sort out tens of thousands of highly problematic observations for just one project, and feeling like I’ve had negligible impact, I’m ready to give up.
Here are some suggestions for future CNC events and also for iNat functionality we need if we are to handle the volume of observations we’re seeing.
- Casual observations marked with incorrect date or location should be excluded, they are fake
- Observations with a copyright flag should be excluded. They are also fake
- The ability to add mass DQA votes to a subset of filtered observations is sorely needed. I have given up marking research grade observations as cultivated, there are way too many for even a decent team of people to handle
- Ditto for marking RG obs that can still be improved because they are blind agreements
- The functionality to both ban a user and remove all their IDs is needed
I’m sure there are more that others can think of. I personally have given up because my inescapable conclusion is that the quality of the final remaining CNC observations will not be enough to justify the labor required. It is a thankless task and I have better things to do with my iNat time.
17 Likes
IMO CNC organizers should be doing the bulk of this work. If someone wants to host the CNC for their city, they need to be in the trenches with the rest of us. Maybe some of them are, and that’s great, but the onus should be on them to not just rely on everyone else to keep their project clean.
20 Likes
It is on my homepage. It says, “If you’re seeing this, you’re in the top 1% of identifiers! From all of us at iNaturalist: thank you for all that you do – we appreciate you!”
It makes me smile and I have saved a copy of it to look at in future if I need a pick-me-up.
6 Likes
Top 1%? I guess I have some more work to do.
Sigh.
5 Likes
This, please!
Actually, I wonder if it would be possible to make copyright-flagged observations be excluded from projects by default - I can’t really think of a reason someone would want one in any legitimate project.
The region that “won” CNC last year was La Paz, Bolivia - there were a ton of problematic observations there during that one. Unfortunately, it seems that the victory firmly rewarded the poor behavior, because the same area is even worse this year, and appears on track to be rewarded for it again. If no intervention happens, it’ll be unbearable next year, I imagine.
Since their project does not disqualify observations for being copyright flagged, medialess, or DQA-marked, there’s absolutely no consequences to bad uploads - even if they get flagged and the uploader is suspended, unless someone goes through and re-IDs every single copyright flagged observation, they still count. The uploader can just make a new account and carry on until that one gets banned too, and no progress is lost.
9 Likes
I suppose I should specify that it feels like it’s unexpected to some, at least insofar it’s not being dealt with by (for example, in this case) restrictions on how many IDs the CNC-related new registered accounts can make, or something like that. I wonder if there’s really a functional difference between “we didn’t expect this!” and “we expected this, but decided there’s no remedy for it”…
3 Likes
Some amount of care should be taken though… i was recently introduced to a casual post stuck at family that was deliberately made as an educational exercise in how far an unusual family can be identified depending on the images. It’s a really nice post with lovely photos and informative links.
I can’t say I value the photos so dark and blurred that I can’t get them past dicot by any means, though. It’s a shrub. There are shrubs in a lot of orders…
3 Likes
I am trying not to see that as a Health Warning 
(you are spending too much time on iNat) but I did spend most of yesterday at Kirstenbosch. A human, being 
not observing.
4 Likes
It would be good to see ‘audited’ CNC results - say in 3 months time - before we are hit by GSB (which is maybe less problematic, much less volume definitely)
2 Likes
You know, I cannot think of a single thing iNat can do to stem the flood of “bad” observations during the CNC, except to close iNat to ALL observations made during the CNC four-day observation period, but that would penalize the “good” observers..
Could the global organizers of the CNC bar a city or region from future participation if too many (maybe 25%?) observations are cultivated, captive, copyrighted, or the like? Maybe, but identifiers would still bear the burden of discovering such observations. Could the global organizers bar the use of Casual observations? That would be easy, except that “bad” observers could still overwhelm the current DQA system with wrong Wild votes.
Is there a way for identifiers to Mark As Reviewed all of the observations from a single project, like the umbrella CNC project? That would blind identifiers to all the problems of CNC observations, but I suppose even if that were possible, taxon specialists would still be driven crazy in future years by observations way out of range.
I’m usually excited by the CNC and motivated to make lots of both observations and IDs, but even I am getting burnt out. Surely there’s a better way?
3 Likes
I would like to see the CNC results restricted to RG
but then
‘they’ will cheat to RG.
5 Likes
Yes, restricting to RG would encourage even worse behavior during CNC without a doubt. The worst elements of CNC submissions that people have described above are in areas where RG was encouraged as part of school or other projects and there are users with sockpuppet accounts or classmates confirming IDs and creating thousands of bogus RG records.
3 Likes
We have a lot of obs by scouts and guides - but mostly - the issue is missing Not Wild. Not seeing sock puppets …
In the spirit of the thread title, today I have made a donation to iNaturalist in honor of all the CNC identifiers/flaggers. I cannot pay you all individually, but thank you for your hard and lately demoralizing work.
16 Likes
In addition to these:
- Add the ability to flag users, rather than observations.
When dealing with large numbers of problematic users during the recent CNC, it has been difficult to track users that were suspended because they were using sockpuppets. The only way to currently do this is to flag the identification or the observation. There should be a way to flag a user, with reason (e.g. being a sockpuppet, using a sockpuppet).
9 Likes
I think something like this could be helpful. Maybe it could be combined with the “Moderator Notes” function or allow better searching of Moderator Notes which is sort of a flag, but not very searchable/visible unless already going to the user.
2 Likes
From working through these kinds of flags, it’s clear that the extent of the issues addressed above is enormous. I back the suggestion of new provisions to prevent these things from happening in future CNC events. It’s super draining and time-consuming to have to deal with this aftermath.
6 Likes