Okay - that is the Seen Nearby ?
For that I go to the distribution map - and follow up the wrong IDs which have been ‘seen nearby’. Unless there are gazillion, then I move on in despair. But the first ID - that is one ID - on one obs - triggers seen nearby. So if you can catch the first few it is worth the effort.
(Missed your previous comment, I was identifying for Africa)
I can’t find the comment now, but didn’t someone say in another thread that no new species was ever described solely based on genetic sequencing? The same standard should apply to splitting a species.
OK, I’m following you now. If I could recommend a better change than your current idea, which would solve the situation much better I think:
A change that I would welcome is the CV not using or factoring in intercontinental observations in it’s training or suggestions, which I think would also fit your goal better than trying to quick fix North America: if you went through in one fell swoop and set the entire NA dataset back to Order, that would only be a temporary fix, because the CV would still borrow from European observations, which the good people of North America would still happily click on when uploading.
In any case, I still stand by my point of keep placing ID’s, just stick to Europe. Europe only has 47K Millipede, keeping that in order is easier than trying to fix the 217K North American observations.
I recognize your frustration at not being able to add a disagreeing ID other than an affirmative ID for some other (maybe much higher) taxon. But I actually think we would all suffer from a community where (a few) identifiers went around saying “not that” to other people’s best guesses.
However, I do believe the disagreement process could be greatly improved by shifting from a “branch disagreement” to a “leading disagreement” model and by improving the notification process so that observers and identifiers can much more easily see those observations where others have disagreed with their IDs.
Isn’t the iNat geomodel already intended to achieve this and greatly prioritize suggestions of those species that are most recorded in an area?
It prioritizes yes, but if there is an area with limited Research Grade options to choose from, it will widen it’s search. Periodically I have to scan Asian Bombus because the CV suggested something like Bombus perplexus in China.
I imagine Millipedes might option be in a similar boat.
That is what triggered the Pre-Maverick project. We were able to rescue a lot of trapped obs.
Quarter of a million grateful thanks to @jeanphilippeb
Is the situation with millipedes the same as with mushrooms, that is, what were once thought to be circumboreal species are now split by continent? Meaning that the European species name used to be applied to the North American populations as well? Or is it a different situation?
People need to be able to disagree with wrong IDs even if they do not know exactly what it is instead. They already disagree regardless of whether or not they can suggest a specific alternative.
If this creates an ancestor disagreement, that disagreement makes it much more difficult for everyone else to move the community ID forward – but it does not have any consequences for the disagreer. In many cases, the disagreer does not even know that they are blocking the community ID. So ancestor disagreement does not have an effect of deterring broad disagrees.
It is unreasonable to expect specialists to not only be able to ID species within their field of expertise, but to also have detailled knowledge of all species with which their taxon might possibly be confused and the taxonomic relations between them. Users also need to be able to clean up out-of-range IDs in order to prevent proliferation of bad CV suggestions. Sometimes this means that they may be checking observations in a region where they don’t know the local organisms in any detail, so a fairly general ID might be the best they can do.
It is very easy to accidentally make an ancestor disagreement even if one selects a fairly specific ID. It happens to me fairly regularly with e.g. Asteraceae, where there are a few species I know fairly well, and a lot more where the most I can do is suggest the family, or sometimes one of the subgroups like Cardueae or Cichorieae. But there are many subdivisions within the family – subfamily, tribe, subtribe – and inevitably sometimes it will turn out that the ID I disagreed with and the later ID have some shared ancestor. For example, if I disagree with Tagetes and suggest Asteraceae, and the observation is later ID’d as Bidens, my ID counts as an ancestor disagreement because both are in the same subfamily (Asteroideae). I’ve seen this happen with bees, too: e.g. there are users who specialize in Bombus and don’t work with other bee taxa. If they push an observation back to Anthophila because they recognize that it is a bee but not Bombus, and the bee is later ID’d as Anthophora or Eucera, this will be treated as an ancestor agreement because these genera are all in family Apidae and subfamily Apinae. Or I will disagree with a species ID and push it back to genus because that’s the best I can do, and then someone else suggests an ID that is in the same subgenus as the original ID.
Keeping up with notifications is difficult enough; it is an extra burden for IDers to expect them to also examine every notification they get to determine whether they might be holding back the community ID because they inadvertently created an ancestor disagreement.
Some of us understand the how of Ancestor Disagreement.
iNat staff has the why of A D.
But I haven’t met an identifier who deliberately intended to disagree with ALL YOUR IDs. (So rude …)
I’m confused about ancestor disagreement. The sequence below is an example of what you mean, isn’t it?
Mallard
Anatidae (waterfowl, not Mallard, I don’t know what)
Common Eider
It looks to me like if nobody withdraws their ID, it will take an 4 more eider ID’s to achieve more than 2/3 Common Eider votes and make it RG as the eider.
Is that different from this?
Mallard
Gadwall
Common Eider
Seems to me that also takes 4 more eider votes, if nobody withdraws the first two?
Or does the Anatidae vote in the first example prevent the observation from ever reaching RG as eider?
In your example there is no ancestor disagreement, because Anatidae is a shared parent of both mallards and eiders and there are no additional shared taxa between family and species.
So a total of 3 IDs of common eider are needed for research grade (2/3 + 1 consensus; Anatidae is neutral because it was only entered as a disagreement with Anas)
An ancestor agreement would be:
initial ID: Mallard
disagree: Anseriformes
subsequent ID: Common eider
Anseriformes is treated as a disagreement not just with mallards, but also with Anatidae, including all taxa within Anatidae. So the disagree of Anseriformes also counts as a vote against common eider. In this case you would need a total of 5 IDs of common eider to get to research grade; this is much the same as in your second example (mallard/gadwall/common eider) – except that Anseriformes was most likely not meant as a conflict with eider.
Note that it will also require at least 2 eider IDs before the community taxon will be displayed as Anatidae instead of Anseriformes. This may not matter too much with birds, but with many other taxa a high level ancestor disagreement can make a big difference in whether the observation is likely to be seen by the relevant IDers or not.
Note that the disagreeing ID of Anseriformes is still considered a disagreement with Anatidae even if the user who entered the ID of mallards withdraws their ID.
If you’re confused, you’re not alone. This is really non-intuitive.
Even if the original wrong ID is deleted — CID algorithm never forgets.
If you politely choose a ‘soft’ disagreement to avoid triggering unwanted Ancestor Disagreement, then the ‘wrong’ ID is the display ID - you are forced to hard disagree. And then follow notifications …
My honest suggestion is that you only take on as much volunteer work as you feel comfortable and happy to do. If it begins to feel like a burden, take a break.
My point is that ancestor disagreement makes more work for IDers (requiring more IDs, requiring IDers to be aware of when it has happened and withdraw if necessary). Even if one is happy to make IDs, these are things that ask for extra time and effort from IDers – and it is completely unnecessary and could be avoided if disagreement worked the way most people intuitively expect it to.
This is about wanting processes to run more smoothly and reduce the number of obstacles – something that everyone benefits from – not about suggesting that IDing is a burden.
I mentioned notifications because I know that there are not a few users who have given up on following notifications because they find it too difficult to find the ones that are important or relevant to them amongst all the others. Ancestor disagreement interacts with notifications in a way that exacerbates the existing challenges of both.
The original topic here has been marked as solved, and the conversation is starting to focus on another topic entirely (ancestor disagreements). If folks want to continue this conversation, they can start a new topic (and we can move these posts there) or we can reopen one of the existing threads that deal with ancestor disagreements as well.