The community taxon should be “Lestes virens RG”, but it appears as “Lestes virens spp virens needs ID”. In contrast, the Algo summary correctly indicates “Lestes virens” with 2 cumulative counts and score 1. After reading some other posts about species vs. subspecies issues, I unssuccesully attempted to change the title by fav/unfav and playing a bit with DQA to force the system to refresh. The only way I found to get the correct title (in agreement with the algo summay) was to opt-out the community taxon
I don’t think so. It is not a matter of a single ID, but a matter of discrepency between the title the page (wrong) and the algo summary (correct). This does not ever happen fortunately. For example, in this obs
there are several ID at the species level and a single ID at the subspecies level, and both the title and the algo summary give the community ID at the species level with RG.
For both obs, the sub-species ID is labelled “leading” and the title should be at the species level with RG, but this is the case only the second one (and becomes true for the first one only by rejecting the community ID). A difference between the two obs is that in the former, the subspecies ID comes first while in the latter the subspecies ID comes last, maybe this is the cause of the labelling mistake of the first obs?
There has been a recent change (discussed at length in the thread linked above - you’ll need to scroll down to get to the ‘we fixed this’, followed by ‘but hey, that breaks everything else’) which means that theoretically an observation will only become RG if the CID and observation ID are the same (and at species level or lower - ignoring use of the DQA). In your example, the CID (lowest level two IDs agree with) is at species level, whereas the observation ID is at subspecies level.
It’s actually a little more complicated in that if the species-level ID comes first, adding a subspecies ID will still make the observation RG at species level - but I’m appreciating the fact rather than complaining about the lack of logic. :-)
Summary: like it or not, this is a deliberate design choice, not a bug. You can get the ID to RG at species level by using a hard disagreement (orange button) rather than a soft disagreement (green button). That makes both CID and observation taxon species level.
The aim was to fix the bug in the linked bug report, that observations could get to RG with only a single species ID using the ‘as good as it can be’ tickbox. Unfortunately the fix has various consequences that may or may not have been thought through - but the behaviour you’re observing is definitely what they have chosen to go with, at least at this stage.
I would encourage you to read through the discussion under that bug report, if you haven’t already - it may help you get a feel for the reasoning and how the system now works, as well as the fact that there’s really no easy fix, because the old system had problems too (they were just different). The treatment of subspecies has always been a little odd, it’s now just differently odd.
Side note: In the example you gave, I suspect you could actually circumvent the system by removing your latest ID and restoring your original ID (since they’re the same). Because of the difference in ordering, that should remove the need for either opting out or leaving it in Needs ID.
You are right, thanks: indeed restoring my first ID works!
Hope that the Inat staff will find a solution to get a title in agreement with the algo summary whatever the order of ID, as the current behaviour is hard to understand and could be misleading
For what it’s worth, the title has always been the observation ID (level of lowest ID if no disagreements) rather than the algorithm summary (which is the CID). Otherwise there would be no title at all until a second ID was added, and it would always lag behind if an observation started at a higher level (example: initial ID animals, no CID; second ID insects, CID animals, third ID moths/butterflies CID insects etc.). Since the title is what people search by, using the algorithm summary would cause major problems.
As an example, see your own observation https://www.inaturalist.org/observations/139888609, where the title does not match the algorithm result. The one case where I think they will always have to agree is if there is some disagreement.
Thanks a lot for your explanations, this makes things far clearer. In fact, this was not an issue for me to have a title that disagrees with the CID when considering taxonomic levels higer than species level and the last ID is at the lowest level, as in the crab example. What I found confusing (but now clearer thanks to your explanations) is when this deals only with species and subspecies levels, as this adds the question of the label RG vs. needs ID into play. But I guess that it should be quite hard to have a system able to consitently manage all the possibilities of successive agreeing, disagreeing, withdrawing, etc. My example of obs on Lestes virens is probably a rather special case. In the future I will think about the “restore” function if I have to face a similar case (ideally the system may perform an automatic “restore” if somebody (in particular the obs’s author) re-suggests the same ID after having withdrawn it, but I guess this may also introduce some unwanted side effects depending on the context)
