Top Identifiers on an Observation page showing the wrong identifiers for the taxon

[agoranomos]

Platform: Website

Browser: Vivaldi 3.4.2066.86 (uses Chromium 86 though)

URLs of any relevant observations or pages:
https://inaturalist.org/observations/27740711
https://inaturalist.org/observations?place_id=any&subview=grid&taxon_id=61856&view=identifiers

Screenshots of what you are seeing:

Top Identifiers section on the observation page:

The actual top identifiers of the White-tailed Bumble Bee (Bombus lucorum):

Description of problem

Step 1: Head to https://inaturalist.org/observations/27740711

Step 2: Notice I’m on the top identifiers list, along with tony_wills

Step 3: Head to the full list of top identifiers for the species: https://inaturalist.org/observations?place_id=any&subview=grid&taxon_id=61856&view=identifiers

Step 4: Notice I’m not on there, and nor is tony_wills. Neither of us have given an ID for this taxon before.

Extra info:

• I noticed that we’ve both ID’d Bombus terrestris quite a lot, and that was the initial ID on the observation, but it has been withdrawn
• The observation has some other issues - it says it doesn’t have 2 IDs yet - when it clearly already has, and also states it still Needs an ID when the 2 agreeing IDs should suffice.

[agoranomos]

After a little more investigation this appears to be an issue with the Top Identifier index for the Bombus lucorum taxon - it’s not specific to the referenced observation.

I can see the same problem on all Bombus lucorum observations. For example: https://www.inaturalist.org/observations/64018711

Additionally tony_wills is marked as the Top Identifier on the taxon page, but can’t be seen when clicking the Leaderboard button: https://www.inaturalist.org/taxa/61856-Bombus-lucorum

[jwidness]

There are a couple of things going on here. First, note that the Explore view does not actually show the users who have identified that taxon the most, rather it shows the users who have added the most identifications to observations that currently have that taxon (or descendants) as the observation taxon. This is intentional:

@tiwane

OK, after consulting with Ken-ichi, it’s not a bug but intended behavior.

“The Identifiers tab in Obs Search shows the people who have added identifications to observations in the current search, not people who have added identifications of the taxon filtering the current obs search. I am fairly sure that was intentional, as obs search could have multiple filters applied to it that may or may not have anything to do with taxa”

Related feature request here.

Second, I do think there is an actual bug, stemming from an issue with Complex Bombus lucorum. At one point Bombus terrestris 57516 was listed as a child of Bombus lucorum 61856. @loarie reverted a taxon change, and I think something ended up a little wonky, because some B. terrestris IDs still think they are underneath 61856 even though B. terrestris knows that it’s under 1092510.

As a result, when you query for IDs of 61856 or its descendants (the list shown on an observation page), you get 118 for tony_wills, who has IDs of B. terrestris that think they are under 61856.

Probably the same or a similar bug is causing at least some of the issues report here.

[agoranomos]

Thanks for the reply @jwidness

I don’t have any issue with how the Explore view is displaying identifiers - I was using that view to try to demonstrate that I have never made an id for Bombus lucorum. I see now that using that screen is not a perfect way to demonstrate what ids I have or haven’t made.

My problem is that I’m being shown as a Top Identifier for a species I didn’t even know existed until yesterday.
It looks like you’ve found the data issue with that incorrect ancestry list though. Thank you for digging it out and explaining it so clearly.

[tiwane]

Will add to my weekly report, thanks for the sleuthing, @jwidness