I know I can get a bit discouraged sometimes when the first batch of observations I see in Identify are all “picked over”, either because they’re all species that are difficult to identify, or the photos aren’t high quality or sufficient for showing diagnostic characters, or something else. A few tricks for finding more interesting/IDable/fun observations to identify:
sort by random
jump to a random month in a previous year, e.g. June 2016
Good post
i just realized i hadn’t added you to that list cassi! Mostly because you aren’t usually in areas i do ID for but… i did put in some upper midwest areas. If anyone else is in the northeast US/southeast Canada/California and wants on the list let me know!
another one is to filter by a really small area.
And… i sometimes also look at research grade observations. It’s always good to get a third pair of eyes on things, and it means you don’t get stuck with only the hard stuff.
very helpful post and great ideas! is Charlie’s list created by making a project? I think that’s what it is but I’ve never used half of the features on the site so I’m not familiar with these processes. I’m happy to be on the list (NY) but I may not have much that’s interesting to contribute.
Since I usually search for unknown or birds (sometimes insects) in the northeast, the search yields hundreds of pages…sometimes I randomly choose a page in the middle or go to the end and work backwards until I get bored.
I have a few questions about IDing searches but don’t want to clog this thread…where’s the best place to do that, private message one of you two?
yeah, it is a collection project with a LOT of different things filtered for. I don’t know if at some point it will max out. I will add you too, if you don’t post much, i just won’t see much. And yeah you can send me a message on iNat or on here (i don’t know how you even did that, i couldn’t figure it out).
Feel free to message me anytime. @charlie, you can send private messages by clicking someone’s username, then a blue message button appears in the top right.
I have a problem of zipping past the difficult or unfamiliar species and just making high level IDs or only IDing the same handful of species again and again, even when the photos are high quality. It would be nice if the system was tweaked to make adding uncommon IDs or simply things you haven’t ID’d much before more rewarding, like displaying an animated confetti drop or something. I think I would respond well to shiny baubles.
Another thing that would be great is if the computer vision could be expanded to detect photo quality, such as if there are microscope images or focus stacking was used.
I think it would be really fun to run the algorithm on everything, and post the results… not give them any weight in community ID (or maybe just 0.1 weight or something) but allow them to be searchable. Then i could go through everything the algorithm thinks is Coast Live Oak and verify which ones are not… things like that.
I often see a genus level ID that occurs often, and if I think I can find literature on it, I go looking. Researching into the factors and features that differentiate species is a great way to learn about them, but can sometimes be challenging from a terminology perspective. I forget a lot fo what I learn, but re-learn a lot of it too… and for the occassions where I am able to come to a degree of confidence in what I have learnt, I can then go through and refine the IDs for the genus. We are lucky in that NZ is a fairly distinct and bounded fauna/flora, so I am able to access lists of what are present. This helps narrow down the amount of learning needed, so I have become fairly good at identifying NZ pisauridae, for instance, but outside of NZ I know little about them.
It’s not a ‘tip’ as it does not exist, but I would love to see an optional ‘blind’ ID mode, where you can be presented records where the date and location are presented, along with the photos, but any existing ID’s hidden from view.
Allow you to enter enter your ID, then check to see if you are in agreement with what has been done, and then either simply leave without saving it, or if you find your were right (or apparently right based on other id’s) save it. Avoid the stigma of having to enter it if you dont want to be seen to be learning etc.
there was a blind ID study for a while. Not sure if it still works. I tried it and found it incredibly hard because the comments weren’t included. Many photos had more than one plant and i couldn’t see which one they had ID’ed or described. I had no idea how much i used the original ID and comments and not just in the form of unwanted bias, also for context and to just know which organism was intended to be the focus of the observation! I also accidentally knocked some things back to genus…i forget why but it was annoying the other users. It was still an interesting and valuable exercise though.
It might make a fun little “window into iNat” in the same vein as the global recent obs thing and that “badges” thing recently, where you could select an area and/or a top level taxa, and have it throw obs photos at you and see if you can blind ID them to match CID. Not to put IDs, just as a “practise” thing like those flash cards used on bright and gifted kids… (as a kid, I used to think I was not bright, because my parents never got me any of those!)
I had started making a user script that would hide all IDs on the website, but I never got very far. I think that would be a better way to go about this.
I was thinking something that could track your score etc, and maybe prompt you to practice again a week later on taxa branches that you showed weaker results etc. Really depends on how bling the idea wants to become :)
In order to help with identifications (until such time as there is a journal page on the taxon pages), we have started adding links to comments that deal with identifications (mainly species in a genus, but much more for subspecies within a species (which are often not dealt with in field guides).
On the taxon page on the “about” tab, there is an option to “add a link” - this allows one to insert a link to the comment or journal page in which the identification is discussed.
It is a bit obscure and there is no indication until one opens the tab that the information exists, but it is extremely useful when it does.
@bouteloua, thanks for the suggestions! I sometimes do a variation of your “robot mode” by adding species-level IDs to things that are in monotypic genera (or at least monotypic in North America) where the common name is the same as the genus name and therefore observations often get identified only to genus, e.g., sassafras = Sassafras albidum; galax = Galax urceolata. I’m sure there are other things in this category…
@charlie, I also do narrow geographic areas, e.g., specific counties or NPS units.
(Breaking this into two posts because I am “new” on the new forum platform and therefore can only tag two other users in a post : )
@zookanthos, +1 for shiny baubles. I have a sticker that says “I am easily motivated by the promise of stickers.” I kind of like the lack of gamification on iNat (although I understand their new app is swinging that way), but also sometimes wouldn’t mind an automated pat-on-the-back.
@tonyrebelo, I love the idea of adding links to taxon pages. I’ve seen some great discussion & ID tips in comment threads and journal posts, which easily get lost. I’ll try to remember some of them and add links.
I do a lot of coarse IDs on Unknowns with a filter that I bookmark, and I sort the page by date ascending, so I see all the really old neglected stuff first. There are a few duplicates and bad quality photos, but when something that old is still Unknown I think even Flowering Plants is better than nothing?
Also, when you filter by Unknown and date ascending, quite often someone has put the Latin name in the placeholder, and it takes me two seconds to ID it very specifically, which feels like cheating but also the observation does kind of need to be IDed specifically. It means the system thinks I’m way better at IDing things than I actually am, which feels very underhanded! At least it gives the observation a chance to be IDed by a specialist who can get it to Research Grade status, though.
It’s really satisfying to see a 6-month-old Unknown observation, put it in Flowering Plants or something, and two days later 5 people have helped narrow it down to Research Grade. :)
Also I ID a lot, and doing things date ascending means I don’t have 50 notifications every time I log in. ;)
