No, no, I said prevent new accounts to ID during the first initial days. They can obviously post observations if they want. When I checked some of the projects this year on the second day of the event there were observations with 10 wrong IDs weighting on the community ID. All those 10 IDs were corroborating with the initial wrong CV or deliberately wrong ID from the original poster of the observation.
When the observation stage of the event ends we can unblock ID for everyone. Would it prevent gang-ID issues? No, but we can then see how people behave. Maybe we can block for a longer period. I don’t think we will ever get rid of all the issues related to this, but it is better than no filters and massive chaos.
+++
Thanks @thebeachcomber for the clarification, i understood the taxon block wrong.
Ah, yes, I did, too. You didn’t say prevent them from submitting observations, you said prevent them from submitting IDs. Presumably, their own intial IDs would still be allowed, because we do get exasperated with so many “Unknown” observations.
This isn’t functionally identical to your image-search-like proposal, but there is a @kueda proposal on the iNat Github to check metadata for duplicate observations (at upload, if I’m understanding correctly); the idea could potentially be adapted to the purposes of checking others’ observations. There’s been no activity on that Github issue for twelve years though, so…
I wonder whether more mandatory training for organizers of local CNC events would help – i.e., ensuring that organizers are people who are already familiar with iNat and are involved in checking observations and helping provide feedback to new users, or for larger events that they have a team of other experienced people working with the groups who are being encouraged to participate.
It feels like the problems here are very similar to that of school groups, just on a much larger scale. It is not entirely the fault of new users for flooding iNat with poor or faked observations if the purpose of the CNC has not been communicated to them adequately. It is also a pedagogical issue. Perhaps it would make sense for more attention to given to creating better material for teachers and making this material available more prominently.
I’d also like to see the competitive aspects of the CNC reframed. I am not necessarily against gamification, but the current approach that merely focuses on most observations and most species does not strike me as particularly meaningful and does not encourage quality observations or even that the observers expand their appreciation of nature.
But things like acknowledging new personal or local species records, or breadth of organisms covered (number of different kingdoms/families etc.), or improvement compared to previous years, or new observations in areas/taxa that have been previously underrepresented might produce more valuable data. Or for new users it could include components that would recognize progress in learning to use iNat and understand its relevance as a citizen science project (for example, badges for correctly recognizing the difference between a wild and cultivated plant, or using an external resource to make an ID, or adding annotations or observation fields, etc.)
Yes.
And MUCH more emphasis by the organisers on IDs and ‘our own’ identifiers.
The winner by species - in the playground bully, thug sense - is a local project at THREE % RG. Quantity sure, quality not at all.
(I bolded a couple of lines that were NOT bolded in the original post)
Gamification is okay, and trying to cover different kingdoms is okay . . . CNC is creating headaches for identifiers . . . CNC is polluting iNat data and GBIF data . . . It has been mentioned before, but should CNC be moved to more of a Seek-based approach? Would that one decision solve most of these problems?
In my opinion yes. The scale of the problem is mindboggling. The vast majority of Macao CNC research-grade obs are misidentified. In the vascular plants, I’d estimate 90% before I got started. After two days I feel like I’ve hardly made a dent. Some of these have 3 or 4 concurring IDs.
If you haven’t already been doing so, I suggest folks mark these (when obviously wrong) as “Community ID can still be improved” to keep them in Needs ID status and out of GBIF. Doesn’t make the task any easier or less thankless, but maybe helps the reputation of the iNat dataset.
Most of the groups during CNC appear to be school groups anyway. They go during their biology classes or perhaps during some after school groups and make as many observations aas possible. Often many different observations of the same plant.
Isn’t that phenomenon basically Gerald’s origin story? He was misidentified by a large class of students so lots of identifiers joined up to bump his observation to the right species, leading to the huge number of ID’s?
i’ve mostly only been identifying in my area during CNC, but it seems relatively clean here, other than the usual cultivated observations not marked as such. (besides the cultivateds, i’ve come across only a couple of observations where i thought that the subject seemed suspiciously out of season.) there don’t seem to be a lot of students making observations here though, or at least i haven’t come across many that were obviously such.
the main thing i’ve noticed here is that that there are folks who put different photos of the same individual organism in separate observations. it’s annoying mostly because some of these are among the top CNC observers by observation count in my area – so there are a ton of these observations. (i’ve ended up excluding a few people while identifying for my own sanity.)
As an active participant and promoter of City nature challenge for the last three years, there are few things I have noticed. In order to post maximum number of observations (Since the event is mostly projected as some sort of a contest by many local organizations), many users resort to plant observations or abundant species such as pigeons posting multiple encounters of the same species (10-25 times). I personally wrote a user who had posted 125 coconut palm tree observations with the same picture in all of them and had 500+ such observations. Some one wanted to post a gathering of 10k flamingos as 10k observations..this is ridiculous!
There is also another loophole, one just has to create an observation during the day of the event but can later on add photographs or replace the media later on with backdated or captured on a different date (As media and observations are considered and licensed separately). Most CNC users who upload thousands of observation on iNat, completely disappear never bothering to follow-up on their observations or add annotations.
I’ve began to see this too. Like account filled with, no joke, literally copied taxon photos from Inaturalist and ebird. I mean, obviously there are Cassowaries and American hummingbirds found wild on India’s streets, that look and pose exactly like how they do in ebird, right? I haven’t seen enough of these to become a nuisance, but that’s mostly because I haven’t been keeping up with the CNC id stage.
First I better say I have never taken any notice of CNC before so I may have the wrong idea of its role. But if it is basically a competition between cities and there is large scale cheating going on, can’t the cities be penalised for the number of cheats? What if the city’s final score had two points deducted for every false post detected? Or maybe that would just lead to people putting false posts in their rival cities.
Will be interesting to see how they present the CNC results today.
I look at species, not observation or observer numbers. San Antonio led from the start - and to my mind they remain the leader of CNC 25.
And then RG - how many identifiers are working on the quality not quantity for this location? For Africa I had volume I couldn’t hope to clear, but saw few problems - often from the scouts and guides - and there I am glad to see new and young people being encouraged to consider nature. At Kirstenbosch during CNC I was delighted to bump into a group of 9 BotSoc interns!
I agree that the observers seem have not been properly trained, especially to know what a captive/cultivated observation is. It seems that they are observing living things that are right there with easy access, such as landscape plantings, gardens, streetscape plantings, etc.
