Default to showing observations on identify in random order

I added some id’s, searching in random order and also using the default of most recent first. For a few days I did random order id-ing with no notifications (besides observer agreement), but for a day I also did recent-first and have gotten notifications on those, within hours. I suspect the older ones just aren’t being seen often because the default sort is recent first. The recent ones, meanwhile, come up for everybody clicking either identify or explore. That, though, will lead to the older ones just staying there, while the recent ones get peeled off the top constantly.

I suggest making the default sort random to get older observations a chance to be reviewed. It won’t change the fact that pretty, charismatic things can still come up often, because some older obs have fine ids and have good pictures, but don’t have confirmation.

It was discussed here:, and it seems like two new sort options were added. I want to open a discussion on changing the default.

While a good idea in theory, I think in practice this will have more drawbacks than benefits.

  1. new users get discouraged quickly, and swift IDs of the newest observations retain new users. Defaulting to random will lead to decreased user retention.

  2. most IDers will simply switch back to Date Added, adding extra steps to the process every time they want to ID.

  3. top identifiers are often (and understandably) at the verge of quitting the site and deleting all their hundreds of thousands of IDs because they feel that iNat is doing things that complicate identifying. I would not be surprised if we lost several experts over this, if it is implemented.

  4. identifiers who want to ID in random sort, such as myself, already use the feature.


If the biodiversity you ID shows seasonal phenology it is more effective to pick a date, rather than random.
I sort by Date Observed, because of duplicates, and one subject split across a few obs (I need to see consecutive obs) - then I start with the oldest first. Still get a good response, especially if I @mention thoughtfully.

African obs have exploded in the last few days (perhaps people got their internet rerouted due to the cable breaks on East and West coasts) Or preparing for the upcoming CNC ? Down to the latest 6 weeks. I have done my best with the previous old ones.

I have bookmarked a string of ID URLs - set the way I want to use them.


That’s a tad dramatic if they do that, and if you’re worried for it.


Just wondering here, might there be value in offering a “newbie encouragement” queue that gives priority to the very first observations by any user and has a weighting factor that tapers off gradually as more of their observations get IDed?

1 Like

i think it would be a mistake to change the existing default. i bet most people are looking for the most recent observations. so having default as submitted descending is optimal. if you want an option to be able to set the default order by account (not for everyone), i think that’s fine. or even better, creating an feature to save some user-defined searches, and pre-defining one of those as a random search, i think, is fine, too.

a similar option already exists:


I can understand if top identifiers get frustrated and quit identifying when they have a workflow that they think is being impeded. But why would they delete all their IDs? They above anyone know how valuable they are.

1 Like

I don’t think I’d change the default either, for the reasons other people have given. Sometimes I feel like looking at the old observations and deliberately choose them, but they can be slow to work through because many have already been set aside for being too small, out of focus or whatever. You get more IDs per hour with the fresher ones.


I think user default is nice. I also think it would be a nice setting to be able to ask that your own observations never be marked as reviewed, so you can see them in the ID pile, but that’s off-topic.

1 Like

An easy way to ensure that you’re not overlapping with other identifiers is to simply click through to page 5 (or 10, or 20, etc) then work from there. I would hate it if random was the default, because seasonality is important for a lot of taxa, and it’s nice to have the observations (mostly) in chronological order.


Sure, if you’re an identifier. But that isn’t true if you’re an observer. This is the one big reason why I adopted the workflow of picking a different profile each day to work on IDs – I would be so thrilled to get a dozen verifications in one day.

Would you happen to know if something similar is true for top observers? Especially the ones who make efforts to provide as narrow an initial ID as they can?

1 Like

I get a bit frustrated with the - give someone else a chance observers. But - we had one person the other day. Downloaded hundreds of photos - all with no ID. Good sharp photos. And all with IDs already.

I like the current default of ‘last upload first’, possibly more because it’s what I’ve been used to, but I think there are good arguments for it. For identifiers who keep up with their preferred group, newly added is what we want to see at the top of the heap.

I do agree that unless one uses the ‘random’ setting, stuff in the middle of large piles may be a very long time getting another look. I may well start using ‘random’ more for groups with many ‘needs ID’ pages, now that’s been pointed out. I do often start with the oldest posts, just to see what there is that might be identifiable now.

For the taxa I know best, I often copy/paste pre-written taxonomic information and links. I usually also check if the observer has anything else in that group that I could identify at the same time. That saves me time, doesn’t deluge the observer with identical comments, and conserves memory for iNat. It also catches a fair number of older observations, though not one-offs. It probably works better for smaller groups, but it’s another way to catch a lot of middling-old stuff.


A consequence would be to get observations already viewed (but not marked as “Reviewed”) mixed randomly with observations not yet viewed.

I wouldn’t like it at all, and will never use any “random order” option.


I think the best option is to make there be a sticky filter setting so people can have whatever default they choose.

1 Like

and I have that with my bookmarked URLs

I don’t use bookmarks like that; it’d clutter my bookmark list.

But what if there was a page in ‘Account Settings’ with a screenshot or example of the filters popup, and whatever you click on there is the default for your filter settings across the site. Of course you can change them at any time, during any specific search.

I use the random option fairly often, but I don’t think it should be the default. As others have said, getting IDs to the most recently added observations probably has the most impact and I think it’s a good default. It’s also the default for the Explore page and one’s Edit Observations page, so there’s consistency there.

When I’ve used random for large sets of observaitons (eg observations in all of California) I get a lot of observations that have been seen many times but aren’t really identifiable, which is frustrating. I think it’s good when intentionally used, but not a great default.


I have iNat folders (many bookmarks each with a good use, for me)

Fair enough, I just wanted to be able to make it default for a while for myself. I would turn this setting on and off as desired - maybe for a few days on, and then off, and then on again for a few.
I no longer want it to be default for the whole site (i.e. all users).