I don’t think it’s malicious in terms of trying to screw things up, but he’s pretty clearly trying to run up his ID numbers on things he’s not familiar with rather than be helpful.
3 Likes
If the leaderboard is a potential problem, why can’t the leaderboard only count identifications that are the first ID at that level and not include confirming IDs (excluding CV)? That’s a better reflection of some kind of expertise than counting any ID, confirming or CV.
8 Likes
I would agree. I’ve ended up first on leaderboards just for confirming an ID. It was not my intent, and it gives the wrong impression of my expertise and involvement.
HOWEVER,…
I can imagine some experts who are tagged to confirm being left off the leaderboards if that happened.
E.g. say you’ve got an expert in a certain genus, and that it’s one that’s hard to ID to species without extra evidence that can’t be gleaned from a photo (DNA, audio, etc). The expert has favorited the taxon and are filtering their searches on it (because they like to help clean it up, and because they know it so well). Because of their filters, they are therefore never going to see the observations for that genus until someone else has IDed them to that genus first. Their IDs will always be second. If we change leaderboards to only count first IDs at a level, this expert’s confirming IDs will not be counted towards the leaderboard, which could leave them off the list when someone is searching for an expert to tag.
6 Likes
I’m not sure how valuable that is. For many species, given the number of users on the site, it is pure luck if a record comes up when you are identifying stuff and no one else sees it first.
I’m not sure we should be incentivizing speed identifying if that is the only way to get ‘credit’. It would make it much harder to use the leaderboards to see who is truly a top identifier (yes I know there is currently a balance between legit identifiers and number chasers)
6 Likes
This has been an immense problem with bird IDs too; there’s several users with 200k+ observations that are clearly just hitting “agree” from thumbnails. They do not respond to mentions nor PMs. Considering iNat’s record on site issues, I see no solution in the future to this long-standing problem.
2 Likes
Yes, maybe just over-enthusiasm.
1 Like
the leaderboard issue? I scan the names offered for those I recognise. Either for local or taxon expertise. If I recognise a name, it is probably from helpful comments and interesting discussions. I don’t care who the unknown Actual Leader is.
5 Likes
I know of programmers who make a game out of beating computer games and getting the highest scores during weekend tournaments. I can only imagine with that many IDs they are feeding into some identification program and spewing back the results-- which are so often inaccurate. Maybe we need the algorithm to reject an ID made without opening the description, or made in only a few seconds; and a quota per hour.
It takes me time on every ID to crosscheck the species details. Simply can’t fathom how that many can be generated so quickly-- what good is the site if it’s just a game of numbers, and accuracy be darned…
3 Likes
it depends highly on what you’re identifying and at what level. some common plants can be identified to species in a mere two seconds. some organisms need much time even to get to Family, forget species.
for such a system not to be a burden on some identifers (such as myself) it would have to be selective in what sorts of IDs, at which rank, in which taxa… I cannot imagine that the problem would save more time and effort than it would cost.
10 Likes
I am utterly new to iNaturalist, having just discovered it yesterday. I posted a couple of photos with my identification, and received a couple of “confirmations” very quickly. The photos were of common species locally (a great blue heron and a pied-billed grebe). One thing that could be done is to actually post periodically (perhaps for a short time) some obviously incorrect i.d.’s, and tag anyone “confirming” the bogus identification. If errors of this sort accumulate with a particular person, then it would be clear that their behavior was inappropriate. I know there is an ethical problem with posting anything known to be inaccurate, but if it vastly improves the overall quality of observations, then it seems a reasonable price to pay. As a new user, and a definite amateur, it’s important to me that my identifications be confirmed by someone whose engagement is sincere.
10 Likes
Welcome to iNaturalist!
It sounds like you’ve already done a lot of thinking about iNat and concerns around ID and education of users. It’s true there’s a lot to think about!
For my part, I would guess there are not enough people with the expertise to ID to divert their efforts to intentionally wrong IDs. Also, there are users who will hit the Agree button on most any reasonable looking ID, right or wrong. Once that happens, the observation gets classed as Research Grade, and becomes more difficult to correct.
7 Likes
If you are concerned about the status of a specific observation, there is an option to keep the observations in the Needs ID queue. I see you’re using the iPhone app, which I don’t use, so the option is hidden away in menus and I don’t know where.
On the website, on the bottom of the observation page you’ll see the Data Quality Assessment. If you check the box that says, Based on the evidence, can the Community Taxon still be confirmed or improved? --> Yes, your observation will have its Research Grade status suspended so it can go through further review. Later, if you are satisfied with the ID, you can remove your vote (or, someone else can counter it with a No).
Specialists certainly do check RG observations from time to time for inaccuracies. For better and for worse, there are a lot of birders and birds get IDs at blazing speed.
4 Likes
Kind of like a “bait car” strategy as used in law enforcement to catch car thieves. The “neighborhood” in this case might be a taxonomic group where there are lots of bogus agreements.
I like the idea but not sure how effective it would be. Some “good samaritan” might come along first and fix the ID.
5 Likes
As it is, we have the reverse problem - an expert who posts observations that they ID themselves, sometimes with information that’s not available in the actual observation, is left off the leaderboard completely, while someone who comes along and agrees with all of them can become the leader despite knowing nothing about the group.
8 Likes
I expect this is the kind of user that kind of honeypot is aimed at; actively malicious IDers seem to be rare. And as the poster, you can correct the latter issue by simply deleting the observation and reposting.
2 Likes
While this is an interesting theory, I’m not sure it’s the sort of behaviour I want a newbie iNatter subjected to. If you’re consistently courteous and patient, good may come out of it – but people don’t like feeling manipulated and I’m not sure conscious manipulation is justifiable even with good intentions.
6 Likes
I am new here. I have joined iNaturalist to record observations in my local reserve. I work from a Local government survey of plants in the reserve completed in 1993. I have found (I think) most of those plants, and identified them accordingly to that which they were originally identified. I do get some wrong. It is somewhat frustrating when I identify a plant and then have the ID changed to ID the Genus.
Surely, if you can’t ID it correctly then let it sit as is.
I initially started identifying as “Plant” for Taxa that weren’t on my list, but the response I got was flowering plant. I think that is just a joke. Since then I have started having a guess, mostly guided by iNaturalist.
What is the best way to get the ID to research grade?
3 Likes
Hello and welcome!
It’s great that you’re trying to identify your observations. While getting corrected is a bit frustrating, it’s all part of the process. I get corrected nearly every day and find it a good learning experience (if somewhat embarrassing).
Identifications are often iterative – they will be moved to more specific IDs one step at a time. By adding Plantae you took the first step; whoever added Angiospermae was narrowing the category down as far as they felt comfortable. And it is a significant step, too, excluding eg mosses, ferns, and conifers. If that person had known more they probably would have identified further.
The best way to get RG is to be patient, and to engage with the community. The more people know you the more people are likely to look at your observations or even to start following you. Also, different taxa have very different rates of identification – plants are on the slow end.
Don’t worry too much about getting RG though. You’ve already contributed by adding your observation and we appreciate it :)
11 Likes
Flowering Plants does help to narrow it down a little better and might get it in front of a botanist quicker, but maybe not. I’m not a botanist and Flowering Plants is my default ID for my own records when I don’t know what it is I photographed … but it had flowers!
7 Likes
I’m constantly posting IDs on things for the first time. Tonight, it’s bumblebees. Within 5 minutes, I was the first to ID about 6 of them to species (they were sitting at genus).
2 Likes