I’m at 54%, and part of the reason I am that high is because when I first started there were multiple identifiers who followed me and were checking my observations for accuracy(I was 13 and didn’t understand Inat well.) Since the start of 2023 I am at 47% RG.
Compare that to the 68% RG rate for Long Island plants since April 28(when I have identified nearly every single plant observation there, plus that includes my observations as well. Removing me this increases the RG rate to almost 70% for plants. And that is with every observation being identified!
Consider yourself very lucky for 80% RG. You might be the only power user(25K obs+) that has that high of a RG %. I dream of 75%, let alone 80%.
I dream of 50%. I did get two id’s today (one species, two observations) that brought me up to 40.2%. They’re the only recorded observations of Leurolepas roseola.
I’m at 72% Research Grade for Verifiable observations overall (and I don’t make very many Casual observations at all). I think it’s that high for a few reasons:
I mostly photograph species I already know how to ID and therefore I know what characters to photograph.
Most of my observations are from New England, where there’s a very active iNat community of both observers and identifiers.
I make a lot of IDs myself, so I’ve learned what an identifier wants to see to feel comfortable confirming an ID.
In contrast, I went to Australia for three weeks a couple of years ago and for the 669 Verifiable observations I made there, only 46% are Research Grade. That’s because I mostly didn’t know what I was looking at.
Thank you! And I see @grampianshiker is doing the same thing - thank you, too!
So maybe it’s not that I didn’t know what I was photographing (which is still true), it might very well be that there simply aren’t enough identifiers, even around Melbourne where there seems to be an active iNat community.
I think there’s a fairly active identifying community, but not enough, particularly on some taxa. And once you have a backlog, it takes more than just keeping up with current observations (a growing flood, particularly through the upcoming spring), it also needs extras to dig through old observations. Which I’m not sure anyone has for plants, or maybe anything but birds?
I can definitely sympathize, I can’t keep up here in New England. The way I deal with older observations is simply to filter for a particular species (say, Oxalis incarnata in your case) and work through all of them as best as I can. At least for common species, that moves a lot of older observations to RG fairly quickly.
(And I can’t believe I only made 669 observations in almost 3 weeks in Australia! I need to go back, but not for a birding tour - I can’t photograph those.)
I definitely do things like that at times, but there are so many potential species to work through, and so many higher-level observations I also want to tidy up…
I started loading photos in 2021. I have 10100 observations.
Started identifying shortly after but stopped for a while, wasn’t ready.
Gradually got back. I made 8600 identifications so far.
I spend about an hour a day identifying most weekdays. It takes me between 1 minute and 3 days to identify one observation.
I stopped worrying about numbers, leader boards, research grade and cleaning up data. During CNC I take a break. Can’t deal with it.
Most of the time I go through new observations in my area of interest, sometimes browse through a genus or observations of a person who has good photos or good notes. I will keep doing this as I get a good view of what’s out there but I can’t promise I can keep up with increased number of observations.
I guess if we had custom badges, and you wanted a badge for 100% RG, one strategy might be to only observe common birds in your area . . . Mallard and House Sparrow for the win!
It’s a paradox: the more valuable your contributions are to the scientific world, the less likely they are to get identified . . .
Well, I took a look just now, but with six solid pages (at least) of micro mollusk shells, I couldn’t help. My guess is that for your situation, it is less a matter of idntifications taking longer and more a matter of few people in your niche.
I just took a look, and here in California, there were over 1000 mallard observations, most very easily IDed, waiting for ID. There are clear photos of standard male mallards taken seven years ago in the homeland of iNat that never reached RG. Certainly the problem is worse in almost every other taxon, but even birds can go unIDed.
i don’t see anything on that page that supports this. if i look at the Observations (1 week moving window) graph and expand it to display 1 year, it looks like all the lines have grown by more or less 25% from the beginning to the end.
if this is true over the long term, then it should reflect people working on older observations, since i’m not seeing the any significant changes in the ratio of research grade to verifiable observations when i compare to older snapshots.
if what you’re noticing is just short term, then i think it’s just a reflection of the typical seasonal fluctuations in identifier behavior.
I haven’t done a ton of IDs for others, but do try. I don’t do more for three main reasons. 1) Time, 2) I’m not an expert in anything, 3) I really hate making mistakes. I have a couple of thoughts on the imbalance of obs and IDs. Maybe we need to discourage users from submitting poor quality photos that have almost no chance of being accurately identifed. As others have stated, what’s the point in a cell phone photo of a bird that’s half a mile away? It’s not contributing to science and I can’t imagine what the user gets from it either. Fewer observations would solve the problem as well as more identifiers.
I’ve also thought that a photo rating system like eBird uses could help somewhat so that reviewers could filter by photo quality. I suppose there are some issues with that, and perhaps number of times reviewed would serve the same purpose, but we really need a way to segregate the potentially identifiable observations from those that are hopeless.
And, finally, I would just love the algorithm that displays obs for identification to prioritize those from users who identify. When I am identifying, if I find something from a user who also identifies, sometimes I’ll look through all their “needs IDs” (especially if they have good quality images), but if I see zero identifications, I won’t usually do that. Is this just me? I think that it would be motivating and rewarding if the system facilitated that function.
One in four people who flood iNat with low quality photos will become a seasoned observer with ever improving contributions. I come across a couple. Honest but kind feedback helps.
Some people are just not into photography. I met people who know the scientific name of every plant but do not even have a phone.
That blurry photo could be a record of visiting a place, a way to to keep track or to find the location to revisit.
Science is a good excuse to engage with like-minded people, load photos of interesting (or not) finds and learn new things.
That’s why a hopeless photo is still useful.
