The finer points of Research Grade logic

General
1

Hi folks,

In this instance
https://www.inaturalist.org/observations/29392012#activity_comment_4633061

We have one genus level ID, two species level IDs and one genus level ID from same family.
I would expect the logic here to be that it would remain at family or that it goes to species, but didn’t expect it to be held at genus.

Reading the RG rules again, I am reminded that it is based on what 2/3rds of people agree.
So in this instance, 4/4 agree it is Muscidae, 2/4 agree it is Eudasyphora cyanella and 3/4 of people agree it is the genus Eudasyphora. So leaving it at the genus of Eudasyphora does indeed fit with the more than 2/3rds rule.

BUT if, hypothetically, Entomokot were to withdraw his ID, would it then go to E.cyanella, as 2/3rds now agree on a species level ID? If so, this doesn’t seem to connect to the logic of the situation…
That is to say, logically, his ID to genus shouldn’t hold this observation back from going to species level…more than him withdrawing his genus ID ( UNLESS, if I understand correctly… that he has specifically used the disagree prompt in relation to the ID of Ophrys - which he cannot have done, as that was posted afterwards)

Does that make sense?
Haha…
Sorry - I am mentioning this at the risk of being accused of nit-picking!
I am just curious because, as I mention in the thread, I am trying to get more diptera observations to research grade overall so the CV model is stronger for diptera.
And interested in addressing barriers to this.
This is admittedly maybe a bit of a niche situation though!.. :)

( this also relates to my comment in another thread
https://forum.inaturalist.org/t/genus-level-id-as-research-grade/12508/2 )

2

It isn’t stated very clearly, but more than 2/3 of people have to agree

2 Likes
3

Ok interesting!

But to add to my confusion.
Someone else has now added an ID so we have 3/5 agreeing to E.cyanella.
Its now shifted to RG.
But 3/5 is less than 2/3.

4

Also, in another situation, if there were 2 suggesting Eudasyphora, and 2 suggesting E.cyanella, my sense is that this would go to RG as well right?

5

Ahh ! I just saw the more in depth score detailing when you hover over the Community Taxon “what’s this” bit. I think this answers my other comments.

4 Likes
6

Yep, the genus Eudasyphora ID doesn’t count as an agreement or disagreement with Eudasyphora cyanella, so there are 3 IDs at species and 1 disagreeing ID in the genus Lucilia -> 3/4 agreement at species

2 Likes