Dicovered a possible exploit in the already issue proned identification system

General
1

To keep this short. I lost the first observation i found this on, accidentally linked the wrong one. I removed the link. Below is another observation showing this possible issue. This observation is research grade at species level despite the only species level ID coming from myself. I would withdraw it but im keeping it up for this discussion. The reason it is research grade at species is that “No, it’s as good as it can be” is checked. Certainly This isn’t intended. Right? Like if i wanted to i could turn basically any observation with multiple Genus level IDs magically into Research grade for a species. Even if people might not be using it as an exploit this raises the issue of how many observations are research grade at species level despite there being only one ID of that species on an observation. Not sure if this is a bug or not.

2

To illustrate this again i chose a random genus level observation with multiple IDs, and turned it into a research grade random ant species. https://www.inaturalist.org/observations/24523854
This should probably be looked into @bouteloua @treichard @jonathan142 @kueda @loarie

It also now appears on the species page https://www.inaturalist.org/taxa/564987-Formica-yessensis

3

Yes I noticed this awhile ago when identifying observations marked as “No, it’s as good as it can be.” I don’t think this would qualify as a bug—more appropriately a feature request. Maybe a workaround would be to require at least 2 checks for it to be labelled Research Grade?

4

Hm that doesn’t work if someone comes along and adds a finer ID though (like I was doing). I suppose it would need 2 identical IDs and 2 checks, but then the checks go away if someone adds a finer ID? If they withdraw do the checks reappear? I could see this getting complicated really quickly. I personally have never noticed anyone abusing this, but sometimes it doesn’t seem right if it’s Research Grade with 1 ID and 1 check, even if it is true that the Community Taxon can’t be improved.

5

Yes, it’s known it works like that, just check that it can be improved.

6

Well i think this is close to outrageous. Every other INat issue is tame in comparison to this because this breaks the fundamental rule that has been followed since the site started 2/3rds agreeing are required for research grade. in the case of that ant in the genus Formica, only 1/4th is saying its Formica yessensis yet its research grade. Are we as a community really just gonna request this? Imagine how many more things are research grade that could be wrong because of this issue. Then include all the wrong IDs that are research grade. It just adds further issue to the ID system. I could literally if i wanted to go to almost every Genus Crematogaster observation iv’e made and i could just go to Wikipedia and type every species over every observation and they would all be research grade. If that is not an exploit then what is?

7

This apparently also works on Tribes and subfamilys opening up the ability for 1000s of different species to be picked become research grade with just a few clicks.
https://www.inaturalist.org/observations/45712549 like this midge.

8

There’re thousands of wrong RG, so this problem not that bad for data, iNat community is of those who won’t expoit it intentionally. But as you’re worried about it you either make a post in bugs or in requests, and please remove the mark from that chiro, it works with anything, the pattern matters.

2 Likes
9

I was noticing this too, and I think it is an issue even without peopel trying to exploit it

I do a lot of wasp IDs, and many images of wasps in the genus Polistes cannot be identified past the subgenus Fuscopolistes, so a typical situation is that someone posts an observation as “genus Polistes”, then I ID it as subgenus Fuscopolistes, and mark it as “best it can be” because there is no way to ID it further, however, I have noticed these observations going to research grade for subgenus Fuscopolistes even though I am the only one who made a subgenus level or finer observation

I’m not trying to exploit a bug, I just thought I should mark once it is at the finest ID it can be, so that it isn’t still listed as “needs ID”, but I don’t think it should be reserch grade just because of my single ID

10

You shouldn’t yet vote “No”, because the community taxon isn’t yet to subgenus. It will require two people to agree on the subgenus before the “No, it’s as good as it can be” box should be checked.

@zoology123, check out some of the existing topics related to this subject here:

4 Likes
11

Closing to focus at existing topics