Less Agreeable Observations, More Agreeable Text Formatting

Hey all, just wanted to let you know about two recent changes. First, we've removed the "Agree?" buttons on the website when 1) the observation is Research Grade and 2) the identification you're agreeing with isn't "leading." We did this to discourage people from adding redundant identifications to observations that don't need them, i.e. observations that no longer "need IDs" because there's already a community consensus at the species level. I suspect most people add IDs like this because they're fixated on increasing their identifications count. To be clear, the point of adding identifications is not to make a little number increase. It's to help people first, and to improve the accuracy and precision of the taxon associated with each observation second. And yes, I'm well aware that the identifications "leaderboards" might be the biggest factor motivating people to behave like this, but fixing that is a bit more challenging (it will require taking the site down for an evening, at least; I'd prefer to just remove them, but I'm guessing that would not go over well). And also yes, I'm aware some people do this "defensively" to prevent people from shifting the Community Taxon in the future, and still other people add IDs like this because they rely on the system recognizing their ID when extracting data from iNat. That's why we didn't make adding these kinds of IDs impossible, but things are a little harder for you now. It's a tradeoff. I'm hoping this change will also reduce the amount of "thank you" IDs people add. It's great to express gratitude, but a nice comment is a better option than an ID. Anyway, if you're not seeing an "Agree?" button where you were expecting one, this is why.

The other change is the new support for Markdown in comments and IDs and the formatting buttons. For those not familiar with Markdown, it's a more convenient formatting scheme than HTML that builds on how you may already express things like emphasis in plain text. You can use the buttons that now display when you add comments and IDs on the website to see how this formatting works, but here's on overview:

Code Output Keyboard Shortcut
*italic* italic CMD-i / CTRL-i
**bold** bold CMD-b / CTRL-b
[link](https://www.inaturalist.org) link CMD-k / CTRL-k
* an
* unordered
* list
  • an
  • unordered
  • list
1. an
1. ordered
1. list
  1. an
  2. unordered
  3. list

> block quoted text is a nice way
> to quote external sources
block quoted text is a nice way

to quote external sources

But wait, there's more!

Code Output
`code` code
this is
a table

The important bit is the row of hyphens below the header row.

We're supporting most of basic Markdown formatting, plus the tables extension, even though we don't have buttons for all those things.

It's also worth noting that we're supporting Markdown on comments, identifications, journal posts, and mostly on user profiles and project descriptions (you may run into trouble if that text is being truncated as it is on the project detail page). We're also supporting Markdown in the mobile apps for comments and IDs right now, even though we don't have the formatting buttons there. Mobile support for Markdown in user profiles, project descriptions, and journal posts is a work in progress (bold, italic, and links work fine, lists and tables not quite). We're still supporting HTML like we used to, but we're parsing it a bit more striclty than we used to. There are also a few weird cases where past text may now be formatted incorrectly, e.g. if you (like me) were in the habit of listing traces through keys like this,

1. Hairy patella
4. Red tail
18. Falcate toes

you'll need to switch to something like

1\. Hairy patella
4\. Red tail
18\. Falcate toes

Finally, thanks to everyone who chimed in on the Forum about this. Also, huge kudos to todtb for contributing the keyboard shortcuts for the text editor and for adding it to Identify (he's also working on making the text editor available when editing comments and identification remarks). He just volunteered to do both and did a great job, so thank you!

Posted on July 18, 2020 01:46 AM by kueda kueda



Posted by pwsouth almost 4 years ago

Very disappointed with this change. I often go through both needs ID and research grade frog observations simultaneously to check for wrong IDs. (Pretty common, especially with Green/Bullfrogs). When doing that, I figure I may as well agree to the correct IDs, as a stamp of approval that this ID is in fact correct. This will now take much longer.

Posted by mws almost 4 years ago

In regards to markdown, I think that I would prefer it if the formatting of block quotes was the same as in the rest of the post; as it is right now the text is larger than the rest of the post, which feels jarring. If that is an easy fix I'd like it. :) Other than that I really like the rest of the markdown, so thanks for that!

Posted by ekmes almost 4 years ago

I also add confirming votes when checking ID's. "Agree" makes that easier and faster (which is why it was there in the first place), but I'll still do it with this. And you probably won't be able to hear whatever I'm muttering as I do it.

Posted by sedgequeen almost 4 years ago

FWIW, if you're just trying to make it so you don't see the same observation again in Identify, you can click the "Reviewed" checkbox. @ekmes, I agree the increased font size for blockquote is a bit annoying. We can shrink it down.

Posted by kueda almost 4 years ago

I'm trying to confirm the identification, for me and for others. I'm really good at some species, and I check them. I'm not as much as an expert at Mark Egger, for example, but I really want his confirming or correcting identifications on all Castilleja and related taxa because I know he knows them. Most of us don't. (I do use the "reviewed" feature to make some photos disappear, mostly ones that I will never be able to identify.)

Posted by sedgequeen almost 4 years ago

Please understand, I'm not saying you shouldn't make this change. I'm just bitching about it. The change will probably reduce an apparent problem. That's good. It will slow something useful and that's bad, but it won't slow it a huge amount. Trade offs.

Posted by sedgequeen almost 4 years ago

I think there should be a max number of 'Agrees' to be honest. I like having the few experts leave their approval on my observations, even after RG has been achieved. It's a confidence issue.

Posted by jhousephotos almost 4 years ago

I use "Agree" as well for taxonomy that I'm cleaning up - many mis-identifications. (EDIT: not thumbnailing, of course! That's bad and should be removed for sure :-) )

I'm a bit saddened about this change because it means I will get drastically less notifications and drastically less IDs on my obs in the future. It is a lonely world on iNat for any observer nowadays, because your observation gets so little priority compared to the tens of thousands of other observations observed in just that hour.

However, on the bright side, I hope it will make identifiers focus on Needs ID observations more. And since I don't do IDing much anymore (for a year and a half now), it doesn't really hurt me (yet). I will watch and see how this change impacts the community.

Posted by calebcam almost 4 years ago

Interesting, the only way to get an ID on one of these observations is literally to type in the name and save. Forget all the nice community 'pile-up' notifications I used to get - they won't exist anymore. :-( Possibly more RG observations though...
I guess we took the 'pile-up' IDs for granted while it lasted, though. Won't do that again.

Posted by calebcam almost 4 years ago

I think the changes to the "Agree" button are good news. Thank you.

Posted by dbocanumenthe almost 4 years ago

I agree with @mws and @sedgequeen.

Why not just take the ID leaderboard down? Folks who ID just to rack up big numbers are not true iNat folk.

I do not ID for recognition but for my own edification and learning process.

Besides, WHO makes the ID is important to me. If truncated or discouraged, I may not get an ID from those folks I really respect.

Posted by connlindajo almost 4 years ago

I am also computer illiterate. On my cheap low end laptop, I cannot find a CMD key. Can you help me?

Posted by connlindajo almost 4 years ago

We fully intend to change the leaderboard at some point, b/c I frankly am interesting in who makes the most improving IDs within a taxon, though maybe that's also an incentive with bad outcomes. But like I said, it's a bigger endeavor.

Linda Jo, if you have a Mac, the CMD keys are the ones to the left and right of the SPACE bar labeled command and/or ⌘. If you don't have a Mac use the CTRL or control key, which is usually on the left side of the row that contains the SPACE bar.

Posted by kueda almost 4 years ago

A much-needed and welcome change! Go iNaturalist 😁

Posted by po-po-pro almost 4 years ago

A thought: for people reviewing and wish to provide ' a seal of approval'/and to know that they ve reviewed the pic. perhaps a thumbs up/thumbs down option can be provided next to species name after someone has suggested. This helps in not affecting leaderboard but helps people debate out an ID by voting when they quickly wish to tell they agree on a ID provided by someone vs another they may not agree with but are not able to provide all details and are going by species elimination(ruling out species they are very familiar with). this wouldn't affect the leaderboard and may provide ''an actual community consensus''. of sorts for debating on IDs - which many might enjoy on iNat. This wouldn't affect the community ID provided either.

Posted by hopeland almost 4 years ago

Good idea, @hopeland.

Posted by connlindajo almost 4 years ago

I think it will hurt identifications , not help . Difficult identifications should have multiple agrees . It is a trade off .

Posted by kevinhintsa almost 4 years ago

Interesting suggestion, @hopeland

Posted by ajamalabad almost 4 years ago

I wondered what happened. I don't do mass-ID reviewing, but there have been a number of cases today where I would have hit an "Agree" button that apparently is no longer present.

Here are two cases with only a few IDs where I could not hit "Agree" to add my case.

I have mixed thoughts on whether "competitive IDing" is a problem, but I'd rather cases like this were not impacted as it takes type to manually type the name out.

Posted by silversea_starsong almost 4 years ago

It's a pity that the negative aspects of agreeing beyond research grade were deemed to outweigh the positive ones.

I welcome the markdown support.

Posted by wynand_uys almost 4 years ago

Hi , @kueda :
Are you guys planning to work on " comments with pictures " and would it be possible to get a PLAY function and something like " Slide presentation " ?
Also, is there any way to see more than 500 items under " species " in order to help with ID's ? In the case of Moths of North America , by example , we have more than 6,000 species identified already and if I go to the photos I can only see the 500 more popular ones !
I'd love to be able of looking at ALL of them and help comparing pictures with actual observations.
Thanks .
Saludos .

Posted by juancarlosgarciam... almost 4 years ago

Very well done for adding markdown support!

Posted by crellow almost 4 years ago

Right change, but a single confirmation is little ... maybe two are better.

Posted by elkvorr almost 4 years ago

Is there a bug in the current implementation of the Agree buttons? When I go to the Identify Page, the thumbnails for RG observations still all have them. But when I open up the observations, there aren't any buttons! Surely it makes no sense to still have the Agree buttons on the thumbnails, but not elsewhere?

Posted by bazwal almost 4 years ago

RE Agree button:
Agreed! It will remove the nuisance value of compulsive unnecessary agreements.
False RG is problematic
Credibility of ID giver bears weight when it comes to identifications
Difficult identifications might require more agreements
Possibly the current system of only one confirmation is is bit flimsy
Current status is that Research Grade is the final ID - which it might not be. How about a 'Sign-off' button? Perhaps as a higher function which could rely on curator's discretion?
Encourage the use of the 'Follow' option
Please add a 'Follow' option on the PhoneApp platform. Adding and ID or comment is the only way to keep tabs on an interesting taxon

Posted by sandraf almost 4 years ago

I'm all in favour of discouraging crude gamification of the leaderboards, so making it slightly harder to add Supporting IDs seems like a positive step in the right direction. I'd also like to see some improvements to the ID stats page so that it provides more detailed information about what kind of Supporting IDs are being made. SIDs can be very important, since every Improving ID requires at least one. But some SIDs are clearly more valuable than others, so it would be nice to have some sort of metric that captures that.

Posted by bazwal almost 4 years ago

this is a pain in the neck for those of us that check the already research IDs of genus's that have lots of mistakes made on them. grrrrrrrrrrrrrr.

Posted by dave_holland almost 4 years ago

are you actively trying to drive off the people who do the most IDing??

Posted by dave_holland almost 4 years ago

That's not really solving the problem: It doesn't prevent people from agreeing with ID's on their own observations, either as a way to say "thanks", or because they believe in superhuman capabilities of the "Expert" wo identified the observation in the first place.
This happens quite often, and may lead to false research grade as experts do mistakes as well.

Posted by bagous almost 4 years ago

I have mixed thougths about removing the "Agree" button in research grade observations. I undersand the issue of competitive IDing, but it seems to me we've lost a good feature overall. In order to tackle competitive IDing (and consequently the problem of "blind" agrees that lead to wrong IDs) I think there could be a remix in ID leaderboards. Perhaps take the focus off the number of IDs and put it on the number of relevant IDs.

Posted by joaolemoslima over 3 years ago

@hopeland , your suggestion reminds this comparison with iSpot indicated by @tonyrebelo , about treating agreements differently from identifications:

Posted by jeanphilippeb over 3 years ago

Removing the Agree button is the wrong way of dealing with the leaderboard problem. As the world expert on a particular group (not blowing my own trumpet, just fact as the only taxonomist who has worked on them recently), I am diligent every now and then of going through all observations and confirming the identity - just like adding a conf/det label in a herbarium. This gives people the confidence that the "research grade" id is correct. This is supposed to be a quick job, but this will make it a hassle.
The case pointed out above on https://www.inaturalist.org/observations/37823792 is another major problem as I can no longer agree with a lower rank when I know the higher one is wrong.

Posted by chris_whitehouse over 3 years ago

I´m am not convinced about this change either, at least in it´s extent. Sure, it might disencourage some people hunting for numbers from easily agreeing to an ID. However, it will also disencourage people that know their stuff to do so. I do think that the agreement of some IDers are more valuable, e.g. if I already know their expertise. In those cases one "agree" can be fine. However, I regularly have the case that I suggest an ID and someone agrees, but I do not really know how valuable that is. I really appreciate another opinion in those cases. I think, it would be could to set the goal higher with at least 2 or 3 possibilities to agree.

Posted by ajott over 3 years ago

It also appears to have removed the option from infraspecific taxa. I can no longer simply agree to someone putting a infraspecific rank that is more accurate:

Posted by chris_whitehouse over 3 years ago

If you are bulk reviewing taxa x then just copy the name of the species into the buffer and paste it in. It's not a case where agreeing is turned off.

The site needs to do something to make it less attractive to folks to blindly bulk agree to thousands if not tens of thousands of records. This seems like the lowest cost approach to do that

Posted by cmcheatle over 3 years ago

"If you are bulk reviewing taxa x then just copy the name of the species into the buffer and paste it in". Agreed, for high volume "purposeful" (not gaming) id'ers, it's just a matter of readjusting workflow and muscle memory. Besides pasting a name from a buffer, there are tools that let you autofill with a few self-defined keystrokes. (On PC, I have recently been using use free/lightweight program "Beeftext" to fill in my common replies, and will just be able to add something in for whatever taxon I'm doing a "checking project".)

Posted by lotteryd over 3 years ago

This is a major retrograde step.
As a regional coordinator trying to get experts in taxonomical groups to participate in iNaturalist, the Identification tool was a major incentive. And the purpose of getting world leaders in groups to participate was for them to agree to identifications, and to have their expert IDs.
They do not ID on a species by species basis (copy copy pasting and a few self-defined keystrokes is not an option), but usually at Family or Generic level, and their agreements are essential for species rich areas.
Coming from a system with a reputation (iSpot) and agreements, to iNaturalists system where the only way to measure certainty of an ID is by number of identifications, this is a major retrograde step that will put off experts and speicalists from participating in helping with identifications.
The solution to the problem of gaming is not to make it harder to make identifications, but to downweight the gamer's "scores"

iNaturalist might not care about the quality of the identifications, but other people - espeically users of the data for conservation planning, red list assessments, distribution modelling and taxonomists to care about identifications and ways of quantifying identifications. All this new development does is dumb down identifications without any real benefits.

My task now of identifying Proteaceae in southern Africa will now takes three to five times longer. Clearly iNaturalist does not want experts identifications!

Posted by tonyrebelo over 3 years ago

@cmcheatle That does not work when you are changing species each time, as one is when working through a genus, only if you are working on a single species. Please also bear in mind not all the world works on fast connections. Waiting for an acceptable taxon to appear in the drop down can take something like 5 seconds for those in remote locations.

Posted by chris_whitehouse over 3 years ago

@tonyrebelo Tough to comprehend your statement that iNaturalist might not care about the quality of identifications.
iNaturalist cares about the quality of the identifications and that's precisely why they seem to be experimenting with this change. Since they've said that they'll look to add better changes and tweak the system going forward, I don't think it's a very big ask for us to try out this system for a few months. People who genuinely care about IDing will still continue to ID.

Posted by po-po-pro over 3 years ago

I'm in agreement with @chris_whitehouse, @tonyrebelo, and others that don't think this is an appropriate solution. I'm one who " "defensively" to prevent people from shifting the Community Taxon in the future". My personal policy is to add a third "confirming" ID because a majority of observations are identified by someone who doesn't really know for sure what the organism is, and then by someone who does. In my mind, that's really just one solid ID. So I add a third "confirming" ID. With only two identifications, someone can come along at a later date and add an incorrect ID and bump it down from Research Grade. I agree that observations probably don't need 5 or more IDs, and I usually never add an ID when there are already 3 IDs.

For my own observations, I am most pleased when someone with greater expertise comes along and adds confirming IDs. If those events will be decreased by removing the ID button, that will be sad.

Is leaderboard gamification a serious enough issue that it endangers the functioning and purpose of iNaturalist? Has it and it's negative effects been quantified?

A problem that I see far, far more often than leaderboard gamification is "thank you confirmations". The folks that don't know what the organism is, but click "agree" to someone they think has greater knowledge as a way of showing trust or bumping up their observation to RG.

This seems like yet another example of a few bad apples spoiling the barrel....and the majority end up suffering the most.

Is it not possible to just prevent redundant agreements from counting on the leaderboard--if that is the source of the issue?

Posted by pfau_tarleton over 3 years ago

This change to Agree may or may not be throwing the baby out with the bathwater... for me at least, I think the idea of acknowledging "improved" IDs could be a more impacting and overall positive shift in this area. I can imagine this will also decrease overall accuracy as others mention. But in any case, I just wanted to note for those troubled by this.... elements of the responses here relate to a range of conversations in the forum which you might not be aware of....and which might be a more practical place for discussion around these issues than within the blog comments (which are a bit limiting for real debate)...

Some posts which relate to the change and some of the responses above include :

Posted by sbushes over 3 years ago

I am really disappointed in this change. It essentially breaks my major use of iNat to review taxa that I care about (anoles) and also pretty much eliminates my ability to do my outreach project on iNat in a good way. The students and teachers that contribute to my project want to see that the project leaders are responding to and valuing their identifications, which we do by agreeing with students and their teachers when they make correct IDs to give them positive feedback and let them know we're seeing all the work that they are putting in. I can't manually add IDs to the thousands of observations that they post...it takes a lot of time with the agree button as is.

As others have noted, this is a really fundamental alteration to the way iNat works for major identifiers and experts, and I don't think the reasoning justifies the change. There are other potential solutions (just get rid of the darn leaderboards, they don't add anything to data quality anyways). My use of iNat is going to decrease in a big way as a result of this change.

Posted by cthawley over 3 years ago

I am completely agreed with @cthawley and @tonyrebelo.

Removing the leaderboards would be an appropriate change. But then I really wouldn't know who to tag when I don't know an ID.

You are going to start to see a decline of IDs. And a decline of RG observations. And a decline of activity from long-time iNatters. I've already withdrawn from IDing after seeing these sorts of changes, so you can count me out. I'll continue observing since nothing is being changed there yet :-)

Posted by calebcam over 3 years ago

I don't see the addition of confirming IDs to be unnecessary or redundant, particularly by anyone with expertise. I don't add many confirming IDs to ones that are already RG but if I were to export data for use then I might. I get that a large proportion of 3rd+ IDs may not be high quality but an expert (however defined) adds value by confirming a page of a species that they know well.

I know the leaderboard has been discussed ad nauseum but I would vote for removing the raw counts entirely. I would replace taxon specific leaderboards with just the number of improving IDs so you can still find the best IDers.

Posted by hanly over 3 years ago

Sounds good to me.
If you come across a research grade observation and you think it's correct, there's no need to add another supportive identification anyway.

Posted by fuerchtegott over 3 years ago

The redundant IDing can be helpful for people to engage with the website, but some people really abused this feature, and the number of bogus notifications I get from them hinders my ability to look at notifications that I actually want to. It can be helpful to have multiple IDs (has been for pigs), but maybe taking the ID leaderboard would be more helpful.

Posted by elliotgreiner over 3 years ago

Others have stated the case against this change very well. I am in complete agreement that this change should be reversed.

Posted by leslieh over 3 years ago

Please add Agree back. Professional botanists in my state use it to confirm they have reviewed and agree with research grade identifications within Arkansas and this is going to possibly kick them off iNaturalist entirely.

Posted by eric_hunt over 3 years ago

Removing the agree button may address one "problem" but it is a serious disruption for identifiers whose input is needed even for "research grade" observations. There are too many incorrect IDs as it is, and this will not help that problem. At a minimum, the agree button should remain available until more than 2 people have agreed. Maybe 4 or 5 or 6.

Posted by milliebasden over 3 years ago

@milliebasden, that would be a much more reasonable thing to do. Agreed!
@eric_hunt Sorry to hear this! Although I can't say it is surprising.

Posted by calebcam over 3 years ago

I'd like to add my two cents worth. I identify a lot of Noctuidae. My main goal is not to check whether research grade observations are correct, but to get the id's that have been sitting for years or months into the system. Sometimes I focus on one genus, in which case I can copy and paste, but often I don't. I also try to ensure that the observation has annotations filled, and is in some sort of data collecting project, for biodiversity reasons. If I find a wrong ID, I write an explanation to help users learn. Unless I know a collector, and know they are thorough, this all takes time. This morning I'm going through a Graphiphora sp., and have so far had to look at every observation rather than just agreeing based on the thumbnail photo (some moths can be identified this way). This change is likely going to increase my work, which means I'll probably cut back on the annotation aspect - It's time consuming as it is.
I understand the concern about gamification, but frankly, I don't care. If someone wants to add a third, fourth confirmation to climb a leaderboard, it doesn't affect me at all. I DO care about incorrect ID's, and can see how the agree button can lead to some of those. So I guess I'm kind of ambivalent about the change. It's an extra layer of friction, but I suppose it will reduce incorrect id's.

Posted by mamestraconfigurata over 3 years ago

@calebcam how would it cause a decline in RG obs? It currently only applies to obs that are already RG. However, I would support removing the Agree button in one or two specific additional places to discourage the "false" RG that @notiophilus described.

Posted by upupa-epops over 3 years ago

@upupa-epops With more identifiers (like @eric_hunt confirmed) leaving iNat, less observations will become RG.

Posted by calebcam over 3 years ago

Y'all need to figure out a way to alert the entire userbase of potential UI changes at this scale before they are implemented. I know interstitials and notice boxes are beyond annoying but this really needed to get more community input.

Posted by eric_hunt over 3 years ago

Both of these changes are welcome, thank you! In case anyone is not aware, if you type the first few letters of genus and species name, that's usually enough to get to the right name. For example, type "sc vu" or "sci vul" to get Eurasian Red Squirrel. You should never have to actually type the whole name.

I hope the removal of Agree buttons is rolled out to the mobile apps, as I suspect that's a common source of people who don't know what an organism is, but agree to the first ID as a thank you.

Posted by deboas over 3 years ago

If your goal is to review species already at RG, even if as an expert who works on a particular genus, group or whatever, why would you not simply go to the explore view, filter to RG and the relevant taxa and then go to the species view and look at the species one at a time?

It has to be faster and easier mentally than changing your mental search image map because you keep going to different species?

Posted by cmcheatle over 3 years ago

I'm VERY GLAD to see these changes, as so many observations which have already been "Research Grade" often get many more 'ID"s...perhaps just so folks can chalk up numbers of IDs??? Not too ethical it seems.

Posted by katharinab over 3 years ago

I've always found pile-on agrees to be a bit bewildering, especially when there is so much "Needs ID" (at least for the taxa I try to ID, or for my own observations), so I'm neutral to positive on this change. I also am seeing a lot of comments here from people who have been using what might look like pile-on agrees for good purposes, in large volumes, so perhaps this is not the last word.

I suppose we could think about whether adding an "made by agree button" icon (similar to the existing "Computer Vision Suggestion" icon) would help. But I'm not sure how informative it would be - as we've seen, people use the agree button a lot of different ways.

Posted by jimkingdon over 3 years ago

This is most disappointing. Research Grade IDs are by no means always the correct IDs. Downloaded data can show you the number of agreeing IDs and you can have more faith in an ID with more agreeing IDs. It is also nice to have knowledgeable experts/enthusiasts to confirm IDs. There is no downside of having many confirming IDs. It is so important to have that agree function - the work around is so stodgy. Please, please please bring it back!

Posted by robert_taylor over 3 years ago

What would be useful is to track the impact of this change in Agree on identification rate and accuracy over some reasonable period (a month?) and then, alongside user feedback (best expressed in the forums, e.g. here: https://forum.inaturalist.org/t/improve-id-function-and-name-of-agree-buttons-a-modest-proposal-not/6589) decide what further refinement, if any, is needed. To me, it seems like this will really help to solve the "Gerald" problem, but the downsides need to be carefully looked at too.

Posted by deboas over 3 years ago

@robert_taylor the downside is well illustrated by the case of Gerald (https://forum.inaturalist.org/t/how-did-the-observation-of-gerald-the-muskrat-go-viral-on-inaturalist/7851/4). He's not alone. Anyone, experts included, can still add additional IDs. I would disagree that the number of IDs on an observation tells you anything about whether it is correct, but I guess it might be possible for iNat to do an analysis of that (although it could be confounded by hard it is to change an ID once many people have added the same incorrect ID to it).

Posted by deboas over 3 years ago

@deboas 'Gerald' and others are very much the exception...

Posted by robert_taylor over 3 years ago

@robert_taylor, absolutely agreed. I love the iSpot South Africa (now on iNat) community (which I would like to say I am part of!) because they pile-up IDs and really welcome new users well. Now ... that's either going to be much harder to do, or won't happen anymore. :-(

Posted by calebcam over 3 years ago

Another notorious example - the Julia Skimmer some people have referred to. Even with 21 correct (I assume) species-level identifications, this observation has still not reached a species-level Community ID: https://www.inaturalist.org/observations/34353827. This kind of scale is an exception, yes, but it's very common to see four or five wrong IDs that make it a lot of work to shift to an accurate community ID (e.g. see this thread: https://forum.inaturalist.org/t/genus-cicada-does-not-occur-in-north-america/14413/)

Posted by deboas over 3 years ago

I'm also disappointed with this change. I understand the issues with 'gaming' the leaderboards, and people blindly clicking 'agree' without doing a bit of their own research is definitely a problem. However, this will make my IDing substantially slower. As I think several people have stated above, I tend to go through a specific group of organisms that I know well, and clicking 'agree' is similar to adding a det. label in a specimen collection - a way of saying that I, with my level of expertise, believe that this ID is correct at this point in time (which of course may change as I gain more experience etc). My daily check of Orthoptera from Florida took nearly twice as long - since many of the recent posts are of Romalea microptera - a species which people and the AI usually ID correctly and I ordinarily just click 'agree' on. I could just go through only Romalea observations, by copying and pasting the name, but that will just add an extra step to my work - and there's already way too many Orthoptera observations streaming in every day for me to deal with (which I might add is a GOOD thing!! I'll get to them in the winter! the more data the better :)

And as for my own observations, I don't particularly care if they reach 'research grade' or have multiple random people agreeing - as long as someone that knows the group has said it's species X, that's good enough for me. If this change will hamper that - not a fan.

Posted by brandonwoo over 3 years ago

While I agree with the problem of ID collectors. I do foresee other issues of which some have been discussed already.

I just got an obs where I had already agreed to species, and it has subsequently been refined to ssp. Now to agree with the ssp I need to copy and paste the revised identification in, or type it in (yes, the first 3 letters of g&s). This because iNat confirms research grade at species and does not see spp as refining the obs and thus re-instating the agree button.

Also, another obs which I had misidentified, was corrected, and subsequently agreed to by another user. Thus, the Agree button was removed. Now, either I just need to leave it as is, or withdraw my original mistaken ID, or else do the same as previous paragraph to cancel out my ID and replace it with the correct one.

Some obs posted get agreed by total unknown users. Now, before anyone gets into a tizz, I am in no way stating that I do not appreciate other people agreeing, or that strangers cannot agree. Yes, I am happy that an obs is agreed to and gets research grade by whoever, but in the same breath if it is confirmed by someone I know to be knowledgeable in a certain field (even after 5 other agrees), it makes me more comfortable on the ID confirmed.

1) Remove all occasions of ID count to discourage users been ID collectors.
2) Recognize specialists in a field so that users know the quality of their agreement. Me having 1000 ID’s on a species does not make me a specialist. Not by a long shot.
3) Return the AGREE button please.

Besides this, iNaturalist remains a great site. I am using it almost daily, either for 5 min, or an hour or 3, and will continue to do so. Hats off to the management side for trying to please everyone.

Posted by shauns over 3 years ago

@shauns and @brandonwoo, very well said.

2) Recognize specialists in a field so that users know the quality of their agreement. Me having 1000 ID’s on a species does not make me a specialist. Not by a long shot.

Gamification would be worst case scenario. I'm against it and apparently iNat staff is too.

1) Remove all occasions of ID count to discourage users been ID collectors.

This is the correct way to get rid of 'gamers'. Not by removing an agree button and making the whole non-gamer community suffer!

Posted by calebcam over 3 years ago

It would be really nice to have a confidence indicator like MushroomObserver has, perhaps set with a conservative default so that those who are just clicking "agree" without serious review don't skew the data. Obviously that goes beyond a UI change, but it seems like the problems here almost all stem (except from those who are afraid of observers feeling ignored if they don't get an identification from one of their desired experts) from the fuzzy threshold around what Research Grade means, and what an identification means.

Posted by caladri over 3 years ago

Terrible decision. it seems Inat is struggling with some sort of identity crisis. Inat is primarily a social media site centered on nature. that's what it is, and its great because experts and novices alike can make use of it. if you change the set-up too drastically, you'll ruin it for a lot of people. i would imagine almost everyone has certain people, whom they've met on inat, that they want to see if they agree with an observation they've made, even if they're the 5th one to do so. to intentionally take away the opportunity for them to agree and comment seems detrimental to me.
for people who don't care about the leaderboards, don't look at them. i have read about this "issue" for a while and i don't get it. one of the great things about inat is the ability to look at observation and identification numbers through many different filters. Again, intentionally removing features that limit the sites usefulness in order to prevent an imagined problem, is just weird.

Posted by lanechaffin over 3 years ago

I do not like this change at all (RE: the "Agree" button). This seems like an issue more to do with leaderboards than multiple agreements. Even then, in my experience the good of leaderboards far outweighs the bad. Edit: Something like making RG require more than two agreements - three for example - is something I could rather get behind here.

Removing the "Agree" button after RG has effectively just made it harder for experts to give their confirming ID's, which I find far more valuable. I doubt that said experts will now be putting as much effort into reviewing already RG observations.

A problem also comes in with agreements to finer ID's, as mentioned by @chris_whitehouse, one can no longer simply agree to ID's of infraspecific taxa. That will now also be much more time consuming in the long run.

Posted by jeremygilmore over 3 years ago

prevent an imagined problem


Posted by calebcam over 3 years ago

On the issue of the "Agree" button, I am really not sure this is a good idea. If someone wants to pump up their IDs without making much effort, they still have an easy way to do so - by agreeing with observations that already have a single species ID. This will increase the numbers of RG observations, but will reduce the quality of the identifications dramatically, even more as the first ID is often just the default one proposed by iNat, or a best attempt by an amateur.

I agree with @deboas up there that you need to monitor what happens here for a while before making these changes definitive.

I agree with several others that this seems like a tortuous way of solving the leaderboard problem - you have enough data in your system to evaluate whether on average people make good or bad IDs on particular taxa, use that to recommend who can help with the IDs for each taxa. In particular: are people making leading IDs that are then validated by others? (use the information on people's own IDs for this purpose - sometimes they are the experts!)

By all means let people still ID easily ("agree") to pump up the individual ID numbers in their profile, it is a harmless way of tracking one's individual contribution to iNat and it will work as an incentive to ID for both experts and non-experts alike. What you need is to avoid non-experts gaming the system to appear as experts in the field when all they do is agree with someone else who did the ID work. Removing the "agree" button for RG observations won't solve that.

Posted by anasacuta over 3 years ago

To me, the problem with the agree button is this: I put on the first ID, and then the owner of the observation clicks "agree," when clearly they didn't actually know the species (if they did they'd have put the ID on in the first place!) and now we have a research grade observation based on one person's opinion (mine.) This change doesn't fix that.

Posted by arboretum_amy over 3 years ago

@arboretum_amy - One way to resolve this aspect might be to have a
second button

Posted by sbushes over 3 years ago

@connlindajo I can't find a CMD key either. However, looks like you could just use CTRL instead.

Posted by fluffyinca over 3 years ago

Yes, thanks @fluffyinca. The CTRL works.

Posted by connlindajo over 3 years ago

@arboretum_amy - I see your point, and I think that observation-owner agreements like the ones you're describing are the perfect argument for encouraging supporting/confirming IDs, rather than making them a bit more difficult by removing the "Agree" button. Also, an observation owner might agree with an ID because the identifier/expert has explained how to ID the species, and the owner now knows how to ID the organism. This is one of the best aspects of iNaturalist - it's a great teaching tool, and if observers were somehow denied the right to add an ID to their own observations (even if it's just agreeing with yours), that would seriously undermine any confidence they might have in their developing ID skills.

Posted by weecorbie over 3 years ago

I don't think it's a good idea to forcefully limit the number of people who support the definition. First, in difficult cases, I would like to see the consensus of specialists under my ID. Secondly, again, it turns out that if a person unknown to me has already agreed with me, this means that people whose opinion is important to me will no longer appear here (well, it would be presumptuous to expect that they have to tinker with manual typing of the name of the species). Thirdly, it interferes with the practice when one of the recognized specialists from time to time runs through a particular community and supports its members with his authority. If he has to manually enter each ID, it will take him much longer, and it is very likely that this specialist will simply not bother with it. From my point of view, the loss of these opportunities is much more sad than the fact that someone will adorn themselves with a couple of unnecessary identifications. I think this solution is a big mistake and would be happy if the button returned.

Posted by prokhozhyj over 3 years ago

@prokhozhyj, well put.

Posted by calebcam over 3 years ago

Aw, I like the leaderboard competition among informed identifiers. I think it's fun.

I also frequently reference leaderboards in order to get the attention of specialists when something unusual pops up on iNat. This is especially useful for insects that mimic one another.

(this coming from someone who IDs thousands of wasps)

Posted by humanbyweight over 3 years ago

As a curator of a large project that wants to have one of the admins/curators provide a confirming ID for inclusion in our data set, this makes my work harder. I am for taking down leaderboards for IDs but could really use the confirming "Agree" button back.

Posted by gtsalmon over 3 years ago

@kueda @loarie I thought this over and honestly, I understand where the iNat team is coming from. "Leaderboard-based mass IDing" is definitely an incentive for many people, and in some cases it is competitively focused. It very often leads to "blind agreements" and "redundant agreements", which are very much unproductive.

What has failed however is the implementation aiming to account for those cases, where it may actually solve this issue. It's now, presumably as an oversight, caused a reasonable inconvenience for a large number of "genuine" identifiers, including myself. If I opened an observation with 6+ agreements, then I would be fine not seeing the agree button. But for an observation that isn't even research grade, and only has 2 or 3 agreements involving different taxa? This change has no benefit and only causes an added burden.

The solution is ensuring this removal of an agree button applies to cases where there is a substantial amount of community ID agreement. It may even be considered only for research grade observations as well? Those cases actually do help avoid redundant ID. In any other instance, it's impeding and discouraging to identifiers that want to vet or add weight to an ID that may only be proposed by one person at the time.

This is probably a case where more IDs overall actually leads to higher and coverage of identification, due to the ease of access for experts who might not want to have to deal with the extra effort.

Posted by silversea_starsong over 3 years ago

Another thing worth mentioning is leaderboard incentives are good encouragement for many people. And the implication that it results in a lot of low-quality IDs is probable, and does happen, but a lot of genuinely careful IDs come out of it as well. Again I feel like the solution is to just remove the agree button from observations with a high number of community-ID-agreement IDs. That would help prune out the unnecessary IDs without having a negative impact on everyone else.

Posted by silversea_starsong over 3 years ago

Tossing in my two cents into the bucket of change here... I'm not bonkers about the removal of the 'agree' button. I tend to do a few ID's (like 90% just regional ones in Dallas/Fort Worth), and when I speak to relatively new observers on iNat, they seem to enjoy getting ID's - even multiple ID's. It's another layer of engagement -- another verification of a verified observation. Granted, we can still do this -- adding an ID by writing out the taxon name, but I suppose it will slow down the process a bit. I won't mind because I'm already "all-in" with iNat, but I hope it doesn't deter the taxon experts from verifying ID's.

I think the removal of the ID leaderboards would be a better solution than this change -- again, as it's been stated before, I think that the leaderboards are already a bit hidden -- you have to search for them. It seemed like something odd to be genuinely pestered by, but we humans love to complain about something! :)

Of the 1 million observers on iNat, I betcha that getting redundant ID's is not a deterrent to their engagement with nature. It's likely a loud minority that get frustrated when a mallard gets ID'ed over and over again or when someone climbs up an imaginary leaderboard.

Not to get too too off topic, but I've had some recent conversations about this, and especially the concept of "research grade." I really wish that these words could be replaced with a "community reviewed" or something along those lines instead. "Research grade" comes with a lot of implications -- that those observations without it aren't 'good enough' or 'applicable to research.' Perhaps a simple change in the verbiage of "research grade" to something less loaded like "community reviewed" could resolve some of the frustration with redundant ID's.

Two cents worth! Oh, and still "all-in."

Posted by sambiology over 3 years ago

I like the term "community reviewed"! A good description of what it is, and with less implied judgement.

Posted by sedgequeen over 3 years ago

Fantastic two cents, @sambiology! Sadly for the past few months I have not had a single observation where I get >4 agreements, AFAIK. (EDIT: Correction! I have had one. 1/2540 observations) In fact, in the last three months (May, June, July), as of now there have been a little more than 7 million observations. Less than half of them are RG! And from a small sample, >98% of these observations have only one or two agreements.

It is an imaginary problem, and the "bugfix" is a deterrent to community interactions on iNat :-(

Posted by calebcam over 3 years ago

I tried to res this message but I don't understand what you are saying?

So let me try to understand.

"First, we've removed the "Agree?" buttons on the website when 1) the observation is Research Grade" "We did this to discourage people from adding redundant identifications"
I really can't see why an identification is redundant? Isn't that the whole reason of iNaturalist, that the whole community could say something about an identification? In my opinion an identification can never be redundant. The more identifications, the more powerfull the ID!

"and 2) the identification you're agreeing with isn't "leading." I don't understand what this means? Does it mean that you can'r agree on an opposing ID once the ID is research grade???? Is that what this means? Or does it means something else? Do you perhaps have an example? I really hope it's not what I think it is... that once an ID got research grade, the ID is fixed... That would be an enormous mistake! I've checked hundreds of observations, also the ones that were reasearch grade and found multiple mistakes.

"because there's already a community consensus at the species level"
So what? People can make mistakes. One should always be open for discussion and arguments. Community concencus could change based on arguments of only one person!

"I suspect most people add IDs like this because they're fixated on increasing their identifications count."
First of all, why would you think this? Have you asked people, that are doing this? And second of all, what does it matter if they would. Who cares.

"To be clear, the point of adding identifications is not to make a little number increase. It's to help people first, and to improve the accuracy and precision of the taxon associated with each observation second".
To be clear, if you want data that can be trusted, you need proper ID's! You could get this by either real experts checking the data, or showing the data to a very large amount of people giving their opinion. iNaturalist chose for the latter. Now if you trust only on 2 or 3 people for the ID, it's no use for me anymore. Data of iNaturalist will be influenced by friends that went into the field together and check their friends observations.

"And yes, I'm well aware that the identifications "leaderboards" might be the biggest factor motivating people to behave like this, but fixing that is a bit more challenging (it will require taking the site down for an evening, at least; I'd prefer to just remove them, but I'm guessing that would not go over well). " This sounds more like your not on the board anymore and are whining about it... I can't see why you would otherwhise bother about it. If I was you I would see it as an enourmous power of iNaturlalist

"And also yes, I'm aware some people do this "defensively" to prevent people from shifting the Community Taxon in the future, and still other people add IDs like this because they rely on the system recognizing their ID when extracting data from iNat. That's why we didn't make adding these kinds of IDs impossible, but things are a little harder for you now. It's a tradeoff. I'm hoping this change will also reduce the amount of "thank you" IDs people add. It's great to express gratitude, but a nice comment is a better option than an ID. "
Don't understand what you are saying here?

Posted by wouterteunissen over 3 years ago

I'm by no means a big IDer, but as an observer, I think comments and faves work better if you're just interested in acknowledging an observation. It always feels nice to see a fave, or a "Neat find!" pop up on one of my observations.
Also, @sambiology , "Community Reviewed" sounds pretty good.

Posted by tgbirdnerd over 3 years ago

If ID's get fixed once they are research grade, than I wouldn't be any longer interested in iNaturalist.

And if ID's are only based on the opinion of 2-3 people (including the observer)... that's also stupid... this would mean that only the first two people who give their ID will decise the ID. Experts that like to help but only once every month can't even give their opinion anymore... For me personally, living in Europe, this would mean I will be always too late giving an ID for the US... since I will see these observations only the next day... and that will be surely after some other people already looked at it. This would mean I wouldn't be bothered after a short time anymore...

Then there's quality. I already think quality is already a big issue for observations on iNaturalist. Many mistakes, lot's of rubbish and sometimes hoaxes. And worse ...there's no good way of adding or retrieving important data on numbers, life stage or whatever... it's crap... I always hoped this was just a beginners thing and things would get better... but reading this, I fear things will only get worse...

A big turnoff...

So my ID's are not important or redundant... This feals like a big slap in the face, since I checked all 78510 of them... all very seriously... all to improve iNaturalist!!! And then you get this... Would you therefore please be so kind to remove all my ID's!

Posted by wouterteunissen over 3 years ago

If I want to identify second, third, fourth ... I just have to write the name of the species instead of clicking a button. Before I write the name of a species, I think about it more than before I just click a button. This avoids rushed clicks and I don't think it's bad. Good identification and ... sorry for my bad english.

Posted by elkvorr over 3 years ago

I decided to check out my own observations for these horrible, intrusive agreements that are obviously taking iNat down, so that we must make this absolutely crucial change. Going on observations from the last three months (May, June, and now July).

I have 2540 observations so far. 421 (less than 20%) are RG, and 3 are Casual.

Of these 421 RG observations, I have:
70 observations with 2 agreements
1 observation with 3 agreements
1 observation with 5 agreements (A Prairie Rattlesnake, for the record.)

I would highly consider removing the Identify page, too. I mean, this is insanity, folks, can you imagine someone agreeing to my already RG observation? This is already 2.8% of all my observations.

I apologize for being sarcastic :-) Please don't take it seriously. I am only emphasizing that the amount of IDs on iNaturalist are frightfully low, not high. And I really appreciate "Community Reviewed" - thank you so much for those 72 observations, whoever helped to review them :-)

Posted by calebcam over 3 years ago

@brandonwoo 's "a way of saying that I, with my level of expertise, believe that this ID is correct at this point in time (which of course may change as I gain more experience etc)." is spot on as an advantage to the Agree button as originally used by many folks. I'd favor preserving it for that reason alone. It doesn't matter what I favor of course...

Is there a Forum thread for this topic? It's a bit awkward to "disqus" as a journal comment section..

Posted by lotteryd over 3 years ago

So my ID's are not important or redundant... This feals like a big slap in the face, since I checked all 78510 of them... all very seriously... all to improve iNaturalist!!! And then you get this... Would you therefore please be so kind to remove all my IDs!

@kueda, I'm sorry to be the bearer of bad news, but identifiers - really useful ones - are being put off by this change! Please consider our arguments.

@wouterteunissen, your IDs here have been extremely useful, please stick with us! You have corrected at least a dozen of my IDs, which were really excellent spots. I hope you feel better soon :-(

Posted by calebcam over 3 years ago

Sometimes I think comments on posts like these should be turned off to force people to live with it for a day or two before responding :) - counting myself among the typical change-grumblers. No one is being prevented from adding IDs. It's also great that some people haven't had mass blind agreements affect their observations, but no, it's not an imagined problem. It's also still pretty dang easy to just type in a couple letters of the genus/species/common name to find the taxon in the cases you're confirming existing research grade observations.

"Forget all the nice community 'pile-up' notifications I used to get - they won't exist anymore. :-( Possibly more RG observations though... I guess we took the 'pile-up' IDs for granted while it lasted, though." I had to reread this a couple times to understand whether it was meant in jest lol, I can't stand the notifications (and have them off in my account settings) and newbie iNatters tell me all the time that it's super annoying when people add numerous confirming IDs. Thankfully I can point them to the relevant account setting.

Hoping, as others have stated, that the buttons are also removed from all of one's own observations rather than just RG ones, as well as on the mobile apps for consistency. It looks like the keyboard shortcut for agreeing on Identify was removed for your own obs, but seems strange to have not removed the actual button in those cases.

I've also encountered several pretty high volume identifiers in the past few months who still hadn't stumbled upon the Identify page and instead only open up observations in individual browser tabs from Explore or their dashboard to add IDs. In the hopes that it'll help at least one person in this comments section out, here's a link to tiwane's quick tutorial on the Identify page, which makes IDing orders of magnitude quicker than that approach: https://www.inaturalist.org/pages/video+tutorials#identify

Whether adding new IDs or agreeing ones, using your keyboard to i add an ID, type in an abbreviated search query, like "bou cur" for "Bouteloua curtipendula", arrow down ↓ to select the taxon, enterx2 to save the ID, and move to the next obs (→) is still very fast once you've practiced the motions a few times (1-2 seconds for me).

Posted by bouteloua over 3 years ago

you can't be serious about this..."Whether adding new IDs or agreeing ones, using your keyboard to i add an ID, type in an abbreviated search query, like "bou cur" for "Bouteloua curtipendula", arrow down ↓ to select the taxon, enterx2 to save the ID, and move to the next obs (→) is still very fast once you've practiced the motions a few times."

Have you any idea of the way how I would ID? I indentify in the hundreds, not in the tens...
I go to the identify page, filter for a specific species, and agree on all the pictures where the ID is correct. Then after checking a couple of hundreds I refresh and start on the ones where the picture is not that good, I open them one by one.
Then after doing that I pick out the ones that are not properly ID'd and comment on them.
the others that I cannot ID, I leave as "viewed", so I won't come accross them again....

Your way of working is nothing for me...
But it doesn't care anymore I am already so much offended about this, that I will no longer in the future ID observations of others anymore. I asked Kueda to remove my identifications for others since they are apperantly redundant.

Posted by wouterteunissen over 3 years ago

Wow! Oh the irony that I agree with so many of the commenters and wouldn't have to type this out if there were just... I don't know... say... an "Agree" button going unused somewhere. (That's light-hearted sarcasm, friends. I love iNat and the community greatly!)

My experience, FWIW, is that I can name 2 people in my nearly 5 years on iNat that have just randomly agreed with IDs that were clearly wrong. I finally just blocked one of them to stop his inaccuracies from polluting my observations. But that's just 2 people in 5 years. The hundreds or thousands of people that may not be experts but have agreed with other IDs have done so with the honest intent of their amateur opinion or as a learning experience. I communicate a lot with the folks that ID more than a handful of my observations. It's my experience that gamifying IDs is not an issue big enough to warrant limiting the experienced folks that I'm most wanting to hear from. If an expert agrees with a hobbyist's ID, it's validating to the observer AND the hobbyist. Now that they can't easily agree, I'll never know that they've seen my observation and considered it vs the hobbyist who just happened to see it first. There has to be a better way. I hope you'll reconsider this change based on the comments. I'm not active on the forum because it's a deep dark hole I don't have time for on top of the iNat hole I spend a LOT of hours in.

Posted by kimberlietx over 3 years ago

@wouterteunissen yes. That is exactly what I do to, checking the research IDs mostly for plant IDs that are plants from New Zealand that other countries people have IDed using the AI and then people who don't know anything have agreed. I'm not impressed with this change either! 🤬 It adds alot more effort to something I'm already not getting paid for! Of all the things wrong with Inat they pick this to "fix"???!!!

Posted by dave_holland over 3 years ago

@dave_holland yes, exactly that!

Posted by wouterteunissen over 3 years ago

The ironic thing about "gameifying" is that making this site more game like would solve many of the bad issues. there is no online game that just opens the door and says come on in you can use every feature. they have a training programme that you must get thru BEFORE you are allowed in, then once you are in you must use your skills and expertise to open the features of the game, this could be applied to Inat by not allowing new people to mass ID, not allowing them to ID on speices that are not in their own country, not allowing them to ID on genera that have a profusion of species etc etc. Make new people prove they have the skills by "levelling them up" just like in a game. No one cares about the "leader boards" they are total nonsense. If real world experts are wanting to join Inat, say an expert in flatworms, then the could send a message to a curator with a little evidence and have that whole phylum unlocked for them.

Posted by dave_holland over 3 years ago

I agree with you about the fact that the leaderboard, is probably not for many a real leaderboard compared to a sportsmatch. It could be however a nice way of showing gratitude to all the efforts someone has put into the site. A way of saying "thank you".

Posted by wouterteunissen over 3 years ago

i do not particularly mind the agree change - i do not find the value in searching for a species and ticking agree when there’s no conflict and there’s already 3 agrees. even if he fifth agree was from an “expert”, i’m not sure those notifications (and they really are not great to get a flood of) are worth as much as that expert handling contested identifications or difficult identifications. given that the notification options are sort of all or nothing for agreements and the paging is wonky, it is hard to track down the ids i really want to know about. i can say that i’ve rarely seen those species verification mass agrees result in much correction of rg observations. might be interesting to see how many of those mass verifications are of fairly common organisms (looks at my notifications flooded with dozens of old mule deer obs, none of which were contested).

i wonder if the “cannot be improved” or the “follow this observation” selections could handle some of the issues brought up here? this current solution may not be quite right but it’s not terrible. but i’m also fine moving to the next observation if i don’t think the ids that got it to rg are bad ids - my name with an agree does not add anything to that. there’s no rg+ and a months-later agree to one of my own observations from an expert is not super compelling. so maybe there’s a way to have “reviewed by” without those surplus agrees?

also wonder if there’s some weirdness today with the agree button and the thumbnail views? like an async timing thing that is not obvious from the id page as one of the users at that moment.

Posted by roomthily over 3 years ago

What Dave Holland said - we have barriers in the (parallel universe) iNat Forum.

But not on iNat obs. Where it is significant.

Posted by dianastuder over 3 years ago

I will add my voice here to being really disappointed at the removal of the "agree" button. This is the single worst feature change I have seen iNat make as a regular user here for many years. If the problem is the leaderboards, why not just remove them and leave the functionality intact?

I likely will reduce the amount of data checking I do (going through taxa to look for misidentifications) due to this change, which is nothing more than intentionally making it more difficult for IDers to add a confirming identification.

Posted by fogartyf over 3 years ago

Not a fan of this change with the agree button. I add my ID to every robber fly observation I can confirm as an expert in the taxon, and this change makes that process more time consuming. When a taxon expert agrees with or confirms an ID, that gives the observer and other users greater confidence in the ID. Don't make our job harder!

Also, I have no issue with the leaderboards - in fact, I find that they are a useful indicator for who to contact regarding the ID of specific taxa.

I think this is yet another example of iNat trying to solve a problem that doesn't really exist, at least not for the majority of users.

Posted by myelaphus over 3 years ago

As the comments continue to pour in, I am in the house during the heat of the Texas day, identifying observations from my state.... and wondering... am I being criticized for racking up numbers on the ID stats??? :-(

Posted by connlindajo over 3 years ago

@connlindajo yes you are being criticized for that....

To all experts, who have spent much time helping with ID-ing on iNaturalist:
Please ask yourself the following questions:
-How much time have you spent on ID-ing?
-In what way, did you personally benefit from this? Either financially, either in status or just by getting thanks from other users or the developpers of this website?
-what other things could you have done during that time? (contributions to conservation projects, spending time with friends and family, doing other cool stuff, or maybe going into the field yourself?)
-did you have fun doing it? More fun than the other things you could have done?

Maybe this decision of iNaturalist is a nice reason of thinking things all over again? And spent more time in other things? We are not welcome at iNaturalist, so why longer bother about the website? Go into nature yourself! Spent some good time outside or with friends! I guarantee you, it will give much more pleasure than spending more time on this unthankfull website!

(and if someone could help me removing my own observations in a controlled way, I would be very thankfull. Don't want to lose the pictures or info)

Posted by wouterteunissen over 3 years ago


Posted by diegoalmendras over 3 years ago

@wouterteunissen @connlindajo -
Rightly or wrongly, this, as I understand it, is to grapple with situations like this and this. I very much doubt the intention is to disencourage identifiers or appear thankless!
Or indeed to be solidly against all use of leaderboards or gamification.

A better solution to the issue though perhaps, is this subtler shift proposed by @bouteloua. Which would, additionally, better acknowledge expertise.

Posted by sbushes over 3 years ago

@sbushes your two examples of situations where this goes wrong are actually perfect examples why you never should do what kueda just did.... because those people with no expertise at all, will continue anyway.. And now, in the new situation, because the observations are research grade, no expert or other user can comment on them anymore... it's just a stupid choice and the wrong people are blamed for it...

But I'm out of this discussion now. Good bye to you all!!

Posted by wouterteunissen over 3 years ago

I have to be honest, personally I struggle with this concept that it is somehow validating when an expert agrees with an ID. If an expert DISAGREES with an ID, that is meaningful to me, and indicates something I should value and pay attention to. An expert AGREES to an already identified record doesn't validate it, the ID is not suddenly more correct because they've signed off on the non-experts work. I guess I will get told that unless an expert agrees and it shows as such I should always assume it is wrong because its likely a guess.

Posted by cmcheatle over 3 years ago

@cmcheatle I think this point is just opinion-based for each person. For some it may be validating to have an expert agree, for some not. There's not a right or wrong answer there. But I think it's fair for folks who feel that is an important and valuable part of their iNat experience to say so here. If this change would then decrease their enjoyment of iNat, that's something worth taking into account.

Posted by cthawley over 3 years ago

@cmcheatle - This is certainly a big incentive for me, and part of the reason I initiated the forum thread around acknowledging expertise this week. Maybe this differs if you are more confident of your identification skills than others? Or if you are focussed on certain taxa? For me, in Diptera, I absolutely want my IDs signed off by someone I trust.... I make mistakes all too often ...and its often difficult to find the literature to know for sure if my IDs are accurate or not....never mind remember the specific direction a leg bristle has to face for it to be species X or Y...

Posted by sbushes over 3 years ago

I wonder how many of you would trash the Operating System on your computer because of an update you didn't like. Fortunately iNat has people who listen to and consider suggestions/complaints. Why not work with them to resolve issues rather than simply walking away?
I am a low volume user but can understand the concerns of the high volume folks and appreciate their contributions.
I have always felt RG should require more than two ID's to avoid the thank yous and erroneous ID's reaching RG.
Most of us say we are open to change (as long as we like it). Things are hard before they are easy.
Why not discuss rather argue?

Posted by henryy1355 over 3 years ago

In my eyes, the changes to the agree function fly in the face of making iNat a welcoming space and hurts the mission of providing platform for helping to connect people to nature. I've been on iNat for 2 years and have learned a massive amount. I'm not an expert in a specific taxon by any means, I'm still finishing my bachelors in wildlife biology, but I enjoy going through posts in my area and checking to see if I agree with the ID's. If I agree, I like to hit agree. The action of doing so makes people feel more connected to the platform and therefore more likely to use and learn from it. Feeling connected to the platform boosts connection to nature.

I think solutions that others have suggested, such as raising the number of ID's needed to reach RG and eliminating leader boards are viable solutions. That all said, I love this platform and certainly will not be leaving it over this.

Posted by antrozousamelia over 3 years ago

@dave_holland - can you please clarify what you mean by this :
'That is exactly what I do to, checking the research IDs mostly for plant IDs that are plants from New Zealand that other countries people have IDed using the AI and then people who don't know anything have agreed'

if they blindly agreed or guessed or used the AI which was correct or whatever and got it right what difference does it make? Right is right. Like I said before, it's not somehow more right because someone else agrees
if the ID is wrong, this change has absolutely no impact because you wouldn't be using the agree button anyways to fix it

I also just want to make sure as it seems to be a common thread in here, like the staff have said, nothing is locked, you are not blocked from commenting on, or adding an agreeing ID to something that is already research grade. Is there something in the wording of the announcement giving this impression, because more than a few commenters here seem to be under the impression this change means once something is at RG it can't be further interacted with.

Posted by cmcheatle over 3 years ago

unless you are a bulk IDer of genus's that have trouble then you will have no idea how bad this change is. it is a very big pain in the neck for us that are checking the research grade obs. just because it is research there is no gaurantee that it is right! we go thru and bring up either a species or a genus and go quickly thru agreeing with the many obviouosly right obs to quickly get rid of them out of the list so we dont have to deal with them again ever (unless we tick the "reviewed' option). then we go back slowly and open each of the ones that may or may not be right. those which are right we agree with, those that are not we either ID them correctly (which also gets rid of them out of the list) or we comment on them with something like "the pics are not clear enough to ID this to sp level it needs to be returned to genus" but we dont ID to genus on those, as we want them to still be in the list when we come back in a few days or a month or whatever it takes for us to get back to checking that genus or species. it was already time consuming and frustrating enough. by taking away the quickly agree option you have made this totally untenable and you will find that the people that check all these research obs will no longer do so.

Posted by dave_holland over 3 years ago

@dave_holland's approach is identical to mine....and for the same reason is why I prefer the Agree button to return until an alternative solution can be identified that doesn't through the baby out with the bathwater

Posted by pfau_tarleton over 3 years ago

I have mixed thoughts about this change. As someone who IDs a lot I find the agree button to be really helpful and it made things easier and faster for sure. But, I also found myself making a lot of errors which is why I think it is probably for the better that it is gone. Having said that, I feel like the agree button on the identify page is the real target here, since that is where most IDing happens and where it is easiest for someone to hit agree without paying much attention to the tiny thumbnail version of the photo. Whereas I don't think the agree button in the actual observation page is that much of an issue. Having said that, I only deal with a small corner of iNat, so the issues may be different across other taxa.

Posted by afid over 3 years ago

This change isn't going to greatly affect me. I've only made around 24,000 IDs so that really pales in comparison to others on this site that use the ID function more often. However, I agree with the majority here that this is a bad idea. I could beat a dead horse and re-hash all the arguments against this change listed above but I'm not going to do that because it's not going to stop me from using iNat regardless. If I was to identify a plant I've never seen before in the Colorado Desert and a person I don't know (with no profile information listed) adds an ID in agreement of mine I wouldn't feel anymore confident that my ID is correct. I know iNat is about the community at large and doesn't weigh any person's ID heavier than another's, and that's fine, but when an expert like @jrebman confirms an ID it gives me confidence that I made the correct ID. So for me personally, I definitely hold the quality of an ID at a different weight depending on who is making it. With this change Dr. Rebman will not be able to add an ID to an observation already deemed research grade and I think that is a problem. This comment is mostly to serve as a no vote, if anyone from the iNat team takes the time to read through the comments from these posts.

Posted by matthew_salkiewicz over 3 years ago

@wouterteunissen I recognize that you are out of the conversation now, but did you bother looking at @connlindajo profile? She has a lot of identifications, but also a lot of observations. I suspect folks like her are not part of the problem. Your 'advice' is both unnecessary, and frankly, insulting.

Posted by mamestraconfigurata over 3 years ago

@matthew_salkiewicz I have word that the iNat Staff is considering each opinion and is not concrete on this change. This may be revised/reverted/kept, depending on what they decide.

On a side note ...
Great to see you still around, buddy! :-)

Posted by calebcam over 3 years ago

@mamestraconfigurata, I'm not supporting everything wouterteunissen said, but I'm pretty sure he was implying that the justification to remove the Agree button was, in effect, criticizing people for adding an Agreement on RG observations. I think many of us feel that it came across this way also, even though I'm sure that wasn't the intent of iNat staff.

Posted by pfau_tarleton over 3 years ago

'With this change Dr. Rebman will not be able to add an ID to an observation already deemed research grade', yes he can, he may choose not to, he just can't click the agree button anymore.

As has been stated several times in the thread, research grade records are not locked by this, no one is prevented from adding another ID, a comment or anything else.

Posted by cmcheatle over 3 years ago

@cmcheatle Ok I misunderstood then. I tested this and you are correct. You can still add an ID to a research grade observation, its just more tedious now. Thanks for the correction.

Posted by matthew_salkiewicz over 3 years ago

I don't like this change. But I can live with it. I do a lot of identifying. This will sometimes be an inconvenience. Probably not a big one. We can still confirm identifications of any observations we want to, (RG or not). The process is just a little slower in those cases where the observation already is RG. I don't see the change as a statement of the value of my identifications or anyone else's. It doesn't change how useful I think the data here can be. I see the change as a solution to a problem that is really frustrating when it happens. I think the problem is rare enough that I wouldn't change because of it, but obviously others see the situation differently, and they may have a better understanding of iNaturalist as a whole. Lets keep our eyes on the value of iNaturalist data and community, keep working, and see how this goes. If it's bad enough, perhaps we can change it again.

By the way, I'd keep the leaderboards. I use them occasionally to find people who may know how to identify the species I just posted. Some people seem to be motivated by them, and that's not all bad.

Posted by sedgequeen over 3 years ago

I don’t like the change to “Agree” buttons. It substantially slows my ID process down and makes it a lot less fun to see CA lichens ( and ID the few I can). I mostly work on a slow, old iPad that I can easily hold in my hand. Typing in a name (with one finger) that is already entered above seems silly to me. Perhaps I don’t understand the depth of the problem you’re trying to solve because I don’t know what “leaderboards “ are.

Posted by metsa over 3 years ago

I have nothing to add but to echo other's responses: removing "Agree" from RG observations was a terrible decision. You didn't remove the ability of people to add further IDs after it was deemed RG, you just make it more difficult. That doesn't solve anything.

"I suspect most people add IDs like this because they're fixated on increasing their identifications count."
That just shows that you're out of touch with why people ID, and the benefits of having additional verification for RG observations.

"...but fixing that is a bit more challenging" You also admit that this is a piss-poor bandaid fix, rather than addressing the core issue (which I don't even believe is an issue).

"...to improve the accuracy and precision of the taxon associated with each observation second. "
This change is the complete antithesis of this....

"That's why we didn't make adding these kinds of IDs impossible, but things are a little harder for you now. "
So you you're intentionally making things difficult for literally no reason. If it's not a big enough issue to prevent it from happening then it's not a big enough issue to intentionally make the process more difficult. People that ID on iNat aren't paid...you intentionally making things harder for users is a kick in the balls for people who have volunteered countless hours to this website. You're seemingly actively trying to dissuade the members who do the heavy lifting for IDs on this website from contributing further.

"It's great to express gratitude, but a nice comment is a better option than an ID."
The two aren't mutually exclusive....

Posted by kevtolan over 3 years ago

I don't like the change. I enter ID's for many unknown observations. Often, the observer is new to iNat and just agrees with what I suggested. Having a backup for these situations only makes sense. I can see the problem - I am listed as a top identifier of the Black Dash butterfly, but all of them are my own observations and I am no expert on this species (or any other species). Having lists of top anything creates the issues that removing the "agree" button is meant to fix. Wouldn't removing the lists make more sense? There are many good arguments for keeping the agree button above so if the only reason to remove it is to prevent people trying to get better numbers for lists, remove the incentive to do so. It is basic behaviour that even an inexperienced animal trainer can understand. Indira Gandhi was meeting with her cabinet to discuss how to deal with the problem of women being assaulted routinely while out, especially at night. Their suggestion was they institute a curfew - for the women. Gandhi said she would not do that since it was the men who were the problem. The lists here are the problem, so get rid of the lists. This is a sideways solution that will likely have unintended consequences even beyond those mentioned above.

Posted by juliereid over 3 years ago

I totally agree with tightening up the IDs. I often get IDs for fungi that just ID my obs as 'Fungi & Lichens'. What use is that? I already KNOW it's a fungus! Any ID that is above e.g. Family level should be blocked from even appearing. Or, could we delete vexatious IDs from our own observations please?

Posted by val_la_may over 3 years ago

@val_la_may Experts may confirm that your pictures are not good enough to go past 'Fungi & Lichens'. Maybe you should consider turning off agreeing identification notifications in your settings, if you haven't already?

Posted by calebcam over 3 years ago

loads of fungi cannot be IDed by photos alone. they often need spore prints or chemical anaylse or even a dna sample...

Posted by dave_holland over 3 years ago

This change makes it more annoying to identify things, and I really value it when an expert I recognize agrees with the others. I think this will make the expert opinions much less frequent.

Posted by alan_rockefeller over 3 years ago

I have been working on a periodic project to catch up with adding IDs for the plant genus Chaenactis, for which I was the Flora of North America author. I review and identify the RG and Needs-ID observations all in one batch, usually for one State or other geographic region at a time.

This change was implemented about half way through my current push, and has provided a perfect opportunity to compare experiences as an identifier post-change versus pre-change. So I've held my tongue for a day to gather some personal experience first.

The sky is not falling.

I estimate that my ID efficiency is at least 90% of what it was before the change. In the cases where an agree button used to be present but is not any more, and in the smaller subset of cases where I would have used that agree button (via the "A" keyboard shortcut), one keystroke has now been replaced by about 10 keystrokes. But given all the other activity that goes into an ID session (mostly reviewing the photographs and adding comments with a disagreeing ID), the extra 3-5 seconds for those 10 keystrokes becomes minor.

It helps, of course, to be thoroughly familiar with the keyboard shortcuts in the Identify Modal. If by chance that sentence means nothing to you (as it once did not to me), then you have a whole world of ID efficiency awaiting your discovery, as @bouteloua elaborated in an earlier comment.

I once expressed some other ideas (https://forum.inaturalist.org/t/improve-id-function-and-name-of-agree-buttons-a-modest-proposal-not/6589) about how the negative aspects of Agree buttons could be addressed. They were not what was implemented here. But I have to trust that the the co-founder and lead developer of the site has a broader view of what the best balance is for the site, and will continue to make adjustments if and when they seem warranted.

Some general thoughts in response to some of the previous posts:

There seems to be an underlying assumption by some that "expert identifiers," more so than other identifiers, will be driven away from the site if they have to do a little extra work to add their IDs. The same amount of work, it turns out, that the current "Leading" identifier(s) for each observation had to do (unless they used a Computer Vision suggestion). I don't think that is too much to ask, and I don't think the assumption is warranted.

If an identifier is seeing identification on iNaturalist primarily as thankless unpaid labor and a time drain, then iNaturalist may not be the place for them. While the occasional thanks I receive always brightens my day, I identify only to contribute what I can to this amazing community, full stop, no strings attached. If that needs to become a little less efficient for the greater good of the site, the motivation remains unchanged. And if it ever stops being fun, then I will certainly move on a more rewarding activity.

Posted by jdmore over 3 years ago

@val_la_may and anyone interested: See my comments in your observation https://www.inaturalist.org/observations/49534380.
Today I must have gone through two to three dozen observations that were entered as "Unknown". The observer had not entered an initial ID. As some of us ID'ers in Texas say, "We are removing the observations from the Unknown Black Hole or Abyss" where they may languish for some time.

Not on the Agree Button subject, but it would be nice to have it so that an observation could not be added to iNat without a recognized ID (and not a placeholder that gets blipped).... at least whether it is the bug or the plant or whatever that is the focus of attention.

Those of us who enter ID's for Unknowns as "Plants" or "Animals" are making very basic ID's, but trying to be helpful to the best of our ability.

Posted by connlindajo over 3 years ago

I'd like to get an additional İD both for my observations and for my identifications from the monograph of the group (Dirk @albach for Veronica species for instance) or the specialist on the local flora (@vvolkotrub for the Russian Far East or somebody else for another region), this İD means that the specialist has seen and agree with me, and now it is much more difficult. Sometimes in the difficult groups it means that the İD is doubtful a bit.

Posted by julia_shner over 3 years ago

I understand where this change is coming from and that there might be people who spam IDs for the sake of climbing the leaderboard, something which is totally not in the spirit of what iNaturalist is meant to be. But I would like to clarify that there are other reasons why some people might make a lot of identifications in a short period of time. For instance, I have been busy going through species of the genus Vespa (hornets), my field of expertise and a group in which there have been a lot of misidentifications, so after going through the entire genus level for species which need ID, I have now moved on to already identified species, and besides correcting wrong IDs, I always use "agree" with correctly identified species to keep them research grade. This one extra identification can sometimes be the buffer to prevent them from going back to "needs ID" because of people who are not familiar with these taxa and yet add their incorrect identifications, something which I have seen all too often. Thus, it might look like I am spamming identifications when I suddenly top the leaderboard for a given species but that isn't the case at all, and I believe this is also the case for many experts in their relevant fields who make many more identifications than the average user. Once again I appreciate the reason behind this change, but simply want to highlight this other angle to avoid misunderstandings and show that not everyone who suddenly tops a leaderboard is doing so for the sake of being on top. Thanks for taking the time to read this!

Posted by bicolor-jxql over 3 years ago

To reiterate what others have said, I don't think I like this change. I'm a hobbyist ID'er, mainly bumble bees, and frequently I will correct an ID or improve one, and the original observer will simply agree with me, viola, it's Research Grade! Woo Hoo! Except the original poster in most cases doesn't know why I put the ID that I put, and in some cases, I'm not 100% confident in my ID, and I breath a sigh of relief when one of the experts comes along and agrees with the already research grade observation. Getting rid of this, I may not know if an expert even saw it. If you want to make this change than I think it would be worth considering either: increasing the amount of ID's to make research grade, put a higher limit than one on research grade observations, or making it so agreeing ID's don't add to your score.

And to speak about the Leader Boards, a lot of people are saying "just get rid of them", I think that's a horrible idea. I use the leader boards to find experts, I am doing a project right now trying to get a very difficult species confirmed in my area, and to find people to help with that I went through the leader boards to find people who were a combination of experts (based on reading their profiles), and that actually used the site, it's going well, gotten several confirmed. Without that, none of these would be confirmed, as I don't think any of the experts even look at that species for this area, I've had observations of this species sit for a year with no activity. Also, the leader boards are fun, what's wrong with me seeing a little number associated with my name rising? If you make changes that make the site less fun to use, I'm certainly going to use it less, not out of malice, but I am slightly competitive, if you remove even the semblance of competition, I lose interest.

Posted by neylon over 3 years ago

First of all I agree that two ID's to reach RG is setting the bar too low on a site like iNaturalist. @pfau_tarleton explains why much better than me: "My personal policy is to add a third "confirming" ID because a majority of observations are identified by someone who doesn't really know for sure what the organism is, and then by someone who does. In my mind, that's really just one solid ID."

Several people have argued very well in my view why removing the Agree button could be very counterproductive. I am not experienced enough to judge so I guess is worth trying it out as long as the effect is somehow evaluated

I don't understand why it is a problem to have several confirming ID's apart from the problem arising from "spammers" who wants many ID's to climb the leader board. I appreciate that not all people topping leader boards are spammers, especially on less popular taxon groups but I would still prefer getting rid of the leader boards or maybe tweak the algorithm so only leading ID's are counted?

Personally I sometimes add a supporting ID to a newbie observer as a "pat on the back" to the observer, or to add confidence to a particular observation especially if there has been disagreement on the ID.

Posted by nicholas_williams over 3 years ago

This is a policy that discourages use, rather than encourage. The more people that are actively engaged in iNat, the more people learn about the environment, and the more people that support conservation. I posted in detail on the forum, but wanted to add this note here.

Posted by williamwisephoto over 3 years ago

@pfau_tarleton I literally woke up in the middle of the night thinking the same thing - if I misunderstood, I apologise. Fools rush in etc.

Posted by mamestraconfigurata over 3 years ago

Unfortunately removing the "Agree" Button doesn't help, instead making it more difficult for someone to write the species instead of pressing one button.

Posted by ctaklis over 3 years ago

Although I understand the reasoning behind removing the "Agree" button, I feel it is a step backwards. I normally only agree with a RG id when I note that two or more novices have Identified and agreed to a taxa that I work on. I consider this a confirmation of the Identification and a means of building others (novices) confidence in their identifications.
By removing the "agree" button on RG effectively only allows us to disagree with an ID. Further If I give an ID and a novice agrees this prohibits an expert from agreeing with me.
All in all, I personally under the new situation, I do not see why I should take the time to Ident at all.

Posted by fubr over 3 years ago

@fubr I don't think this is the case. It should've been mentioned more explicitly in the article above (it is mentioned in the link on the homepage), but removing the agree button does not only allow us to disagree with an ID. We can still add a 3rd agreeing ID, but have to type it in rather than use a button. The same goes for the second example you cited.

Posted by po-po-pro over 3 years ago

@po-po-pro I agree with you, not a problem, I just will not Ident at all.

Posted by fubr over 3 years ago

OK. I tried to learn something, and this may not be so grave. @bouteloua has a link to keyboard shortcuts, and I tried them out this morning. Hitting 'I' goes straight to the identify box, and opens the menu. That almost always goes to the species I am trying to confirm, so one button, one click (most of the time), and it's done. I don't know yet if the 'A' button works, but so far, the buttons work faster than scrolling and clicking. Depending on how you use the site (I like to confirm single id's, not RG), this may or may not speed things up. I've made a hard list of the common buttons - the tutorial is on a video - and plan to practice with them. Keep your chins up!!
Oh, and don't forget to click on the 'submit' button. So three steps, rather than 2.

Posted by mamestraconfigurata over 3 years ago

I have a friend who knows the local plants better than anyone I know. He also does an amazing amount of iNaturalist ID's on observations whether or not they are already deemed "research grade". I don't believe he is offering all of these ID's to increase his "ID numbers". I am always pleased to see his ID on an observation, even after more than one person has already given the same ID. This is especially considering that the second ID on iNaturalist is all too often just the original poster, who didn't know what species he posted, wanting to put his ID onto his observation, and expecting that the person giving the ID after him knew what species it was.

Under the current system, the term "research grade" is given to too many wrong ID's, and wouldn't fit my standards for any research I might do. It will now be harder for my friend to offer the expertise of his ID, to the ID that may have may have 2 correct ID's, but I don't know they are correct until I see my friend's ID on them, only after which it is a very good bet that it is indeed correct! As it will now take more time to get his expert ID, we won't be able to get as many of his expert ID's.

Posted by stewartwechsler over 3 years ago

Terrible idea to remove the Agree after three confirmations in plants. I agree with all the harsh criticism of this. You have forgotten that every ID adds also a new way to connect with naturists and experts that could be very helpful to future or current collaborations. And many plants need far more than three confirmations to shore up their ID, (ferns, lycopphytes, graminoids with not enough photos, for example). You are throwing out hard-earned data! That shouldn't be done in a scientific endeavor. If you have too much, analyze it differently, don't throw it out. Having said all that, I can see limiting it to say 6 agrees, but not three!

Posted by wernerehl over 3 years ago

I agree that two "agrees" just doe snot seem enough in some cases and I appreciate folks I know are experts to add an "agree" on some subjects. RG observations with two agrees are many times incorrect....I find dozens every day. If it is felt it is necessary to lessen the number of agrees, I think that 2 is an arbitrary and too small a number. Just my 2 cents worth.

Posted by greglasley over 3 years ago

Unfortunately, I have encounter many, many IDs that were termed "research grade" that were incorrect. I will usually tag others in my field to "flip" the ID by agreeing with the correct ID. Case in point https://www.inaturalist.org/observations/53286060

Posted by suepemberton over 3 years ago

@wernerehl +100 to that!

Posted by calebcam over 3 years ago

Hi folks - first I'd encourage folks to try experimenting with this change for a few days to before reacting. This change was released on Friday and here Sunday morning there's already well over 100 comments on this thread.
I think everyone on the team is surprised by the amount of controversy around this change. And as a result we want to make sure we understand people's concerns. But this will take time. I've spent a few hours reading through comments and have only made it to anasacuta's comment.

I would ask that you please, as many of you have done, describe the specific use-case with examples that you used to value and now can't do as efficiently. Many of the comments fall more under the 'general outrage' category which is premature until we understand the specific use-cases. Discussions about the philosophy of ID-ing and related feature ideas like reputation or leaderboard changes are useful but please keep them on the forum as they make it harder to find the use cases impact by this change. Also, we've received at least one request to help someone delete all their thousands of IDs to protest to this feature change - again premature. Please give this feature change a try and again focus your feedback at this point to helping us understand specific use-cases this change is impacting that perhaps didn't occur to us.

It might also help to understand that everyone on the staff (myself included) almost exclusively uses the identify tool to make IDs. Since that tool only shows needs_id obs by default we're not seeing these effected use cases. So thanks for your patience in helping us understand them. Here's my attempt to describe specific use-cases raised in the first 50 or so comments that I've read through so far:

1) Correcting wrong Research grade observations (raised by @dave_holland and others)
The simplest case of correcting a wrong research grade obs wasn't effected

But I see the point that you can no longer agree with a maverick ID on RG obs. I can see how leaving agree buttons on maverick IDs would save time in this use-case for correcting wrong IDs

@silversea_starsong @chris_whitehouse, I'm trying to understand the ancestor agreement situation you describe here. In my understanding ronvanderhoff's ID already rolled the ID back to Ceanothus so agreeing with Ceanothus isn't necessary to correct a wrong research grade obs, correct? So in this case the desire to agree with a lower rank when the higher rank is wrong has more to do with the 'redundant ID' describe below (just want to make sure I'm understanding correctly)

Here are two cases with only a few IDs where I could not hit "Agree" to add my case.

@chris_whitehouse I understand and agree with how the leading intraspecific taxa use-case you describe was impacted, e.g.
https://www.inaturalist.org/observations/8818311 - ssp are already a bit of an edge case but seems like we could further hack around this edge case to restore the use case you describe

2) Agreeing at coarse clades
@tonyrebelo "They do not ID on a species by species basis... but usually at Family or Generic level". Is this use-case not adding disagreements at ancestor ranks (e.g. it is not species X but it is genus Y) but rather just reinforcing the ancestor (e.g. no comment on species X but it is genus Y)?

3) 'redundant' ID's on RG obs
many of the comments are about the value of having additional IDs on RG obs.
e.g. @pfau_tarleton mentioned doing this to cement the position of an obs
@cthawley mentioned it in order to have project members see that the project leaders are involved
adding 3rd confirming IDs to RG obs by project leaders to show project members leaders are invested
@eric_hunt mentioned doing it so that individual IDers can keep track of the obs they have personally reviewed e.g. using tools like this https://www.inaturalist.org/observations/identify?ident_user_id=tiwane

I'll be honest these are use cases that hadn't occurred to me until reading the comments here
@brandonwoo. In the Romalea microptera example you mention, are you adding IDs to obs that are already RG? If so, which of the 'redundant ID's on RG obs' case's does this fall under?

I'll try to read through more of these comments as soon as I am able. And thanks for your patience helping me understand these use cases we might not be aware of.

Posted by loarie over 3 years ago

The change itself is a minor inconvenience. What I find more concerning is the apparent disconnect between iNat staff and the identifier community. That this change is 'controversial' (dare I say unpopular) should not have been surprising to anyone. The rationale for the change likewise seems to fundamentally misunderstand why many people make additional IDs.

Posted by myelaphus over 3 years ago

@loarie I cannot speak for him, but since @jrebman is away/off the grid until sometime in August and he is one of the most prolific identifiers of plants on iNat, I hope his opinion of this change is available before you make any final decision. He adds IDs to research grade observations so that he knows he has vetted the observation and it can be included in distributional range data for the geographic areas he covers. Even if the new process adds only 10 seconds and a "few" keystrokes to each ID, that will add up in his case (currently more than 250,000 identifications and it is not because he wants to be at the top of any leaderboard, it is so the data will be useful).

Posted by milliebasden over 3 years ago

@loarie excellent advice and summary of the practical considerations. I can't speak directly for brandonwoo of course, but from reading his earlier description his use case would fall under 3) @eric_hunt perhaps, with a "3a" extension of, "getting the 'good ones' out of the way" by agreement (similar to how someone else might just hit R, but also with the intent of 3) proper. Thereafter being able to go back easily and focus on the "bad ones" that are RG.

In my own case, RG "Brefeldia maxima" cleanup is a hill I've been just starting to climb. I would be using the agree-clearance method there myself, if there were enough of the true RG records to want to "get out of the way" (cough, there are not but that's a different forum thread).

If I remember right, I did use the Agree method 3) eric/brandon use case to learn my way through Ceratiomyxa fruticulosa records (RG and Needs ID as a whole set). It worked- I moved farther toward becoming an expert during that process. My id count on those sessions did move me up the leaderboard for that species. I see that as a plus not for gamification purposes, but because people are now more likely to tag me in on their Ceratiomyxa questions- which means I can both potentially help them and learn more myself. Big win-win, and that just rose naturally from clicking the Agree button a lot.

Posted by lotteryd over 3 years ago

In the case of Ceanothus and the adding of a "redundant" ID, the phrase "cementing the position" feels like it doesn't quite cover it...as its actually fairly cemented through the power of a disagree. But for me, a second agreement that it should be left at genus level is important.... Both for me as an observer (to hear someone second the person who is correcting me, assures me this might be true - and I might then withdraw my species level ID).... as well as for me as an identifier ( to know that another identifier agrees with me ).

Plus, its weird to have an agree button for the original species level ID still there, but not the genus level ID - how does that help achieve the intended goal of the change even? It just feels like it creates a path of least resistance for others to agree with a probably less accurate ID.

Posted by sbushes over 3 years ago

This has tremendously taken away from the social aspect of iNaturalist, which is an important outlet away from typical social media outlets that I don't want much to do with these days. Also, sometimes I don't identify to species when uploading because I want to compare and study for a little bit. This removes the ability for me to agree with someone who ID's my observation in the time it takes for me to hit upload and get back to the observation. I don't see the benefit in changing the agree function, even after reading the whole post.

Posted by baxter-slye over 3 years ago

"Even if the new process adds only 10 seconds and a "few" keystrokes to each ID"
@milliebasden Yes, it will add a couple more seconds but an ID usually only takes me less than 5 seconds each for my usual Needs ID workflow where I already had to type out leading IDs.

"But for me, a second agreement that it should be left at genus level is important.... Both for me as an observer (to hear someone second the person who is correcting me assures me this might be true - I might then withdraw my species level ID).... as well as for me as an identifier ( to know that another identifier agrees with me )."
@sbushes I can definitely relate to that. There's a balance between notifications/confirmations where having a couple more votes of support is especially reassuring for difficult taxa, but too many, or for easy taxa, it starts to get excessive.

Posted by upupa-epops over 3 years ago

@upupa-epops - Sure. This is a rare thing in Diptera! But I found one to add a 5th ID to last night to experience! Haha... it still seemed acceptable to me...as a way to cross the t's and dot the i's, and note that I'd reviewed it. I remain overall fairly ambivalent though, as I´m not a high level identifier...nevertheless, I definitely sympathise with a lot of the larger points here about recognising data quality and expertise.
I think taking the button away after 3 IDs in agreement though would certainly make more sense and be a better compromise....and/or a shift in leaderboard metric to count IDs which improve accuracy instead of more general, supporting IDs.

@myelaphus - yes - a fundamental disconnect maybe between the sort of data quality actually visible in many "RG" observations ...and the need/want for expertise and/or multiple voices to weigh in?

Posted by sbushes over 3 years ago

@loarie - @sbushes explains clearly above the Ceanothus problem
@tonyrebelo comment in your point 2 was regarding the comment above his that suggested copy and pasting the name, which does not work when you are working through a whole genus/family, as South African experts often do. We go through identifications for the whole genus, regardless of whether they have achieved research grade or not, and add a confirming ID to those who appreciate and understand who is giving such an ID. Tony's confirmation of my Proteaceae IDs are invaluable to me, but lots of others often have already put in their own IDs before he gets around to adding his.

Posted by chris_whitehouse over 3 years ago

Please don't scrap the ID leaderboards. Like @humanbyweight I also use them at times to identify those with knowledge of a taxon. Would it improve matters to have the leaderboards count leading IDs only?

Posted by nancyasquith over 3 years ago

I gave up reading all of this thread looking for answers to my question so it may be covered somewhere....I have been trying to use the Table , Bold and Italic markdown codes in Journal posts, without success. I will now look for answers on the Forum. Any links to answers or wiki or tuition on that would be appreciated.

Posted by kaipatiki_naturew... over 3 years ago

Update....looking again at the Table feature above, which is different to the instructions I saw elsewhere, I may have the answer to that.

Posted by kaipatiki_naturew... over 3 years ago

@loarie I think others basically answered your question to me, but: in the example with Romalea I am adding IDs to observations that are already RG, for both the purpose of individually seeing that I already looked at a particular observation, and for others to see that I have checked it as well. I wrote a more detailed comment addressing the same case in the Forum post since this commenting format is a little awkward.

Posted by brandonwoo over 3 years ago

@myelaphus summed up what I was going to reply, perfectly. :-)

Posted by calebcam over 3 years ago

My side of "apparent disconnect between iNat staff and the identifier community":

I quit IDing (more-or-less) a few weeks ago because I joined the forum and saw an amazing amount of nasty comments about 'agreeing' identifiers, even from well-respected Curators. I belong to this 'gaming' (rolls eyes) community apparently, since I study and go through Agamidae and Pelomedusidae observations, and add agreeing identifications on those correctly IDed and conflicting IDs on those wrongly IDed, and therefore didn't want to face this kind of criticism.

I see now that I evacuated before the dam gave way.

Posted by calebcam over 3 years ago

Without having read all of this - if you really want to improve something about agreeing, remove the "Agree" button for the original observer in all cases. The vast majority of people uses that as a "like" or "thank you" instead of actually being able to confirm an ID based on facts. From my observations, that's a much bigger and more common issue that "redundant" confirmations by additional users.

Posted by teadrivendev over 3 years ago

@calebcam - I too make lots of identifications of certain taxa and therefore I've gotten high ranks on some leaderboards. But I didn't (and don't) think those people complaining about "gaming" the leaderboards would complain about me, or you. They're talking about things they've seen some students do, just agreeing and agreeing to push up numbers. I've only seen it on a small scale, but my students certainly did something similar when I required 100 identifications from each of them spring term. I'm sure some people do it, though you don't and I don't, and I don't know how much of a problem it really is.

Posted by sedgequeen over 3 years ago

'Without having read all of this - if you really want to improve something about agreeing, remove the "Agree" button for the original observer in all cases.' The one challenge I have with this is it will force people who are trying to be diligent and conservative in their id's on their records to be more aggressive. It is not uncommon for people to research something, think they have identified it but intentionally ID at genus or higher (often even saying what they think it is) to ensure someone has to actually enter the species and not blindly agree.

Please note I am taking this to mean that once an observer has entered an ID, they can't change to a different one, whether by an agree button, manually typing it etc.

Posted by cmcheatle over 3 years ago

I agree this is a more pressing issue. For me, it wouldn't stop me from agreeing the ID manually though I think, as one would still want to shift the main observation ID if it was conflicting. Alternatively... instead of removing further practical functionality, why not just add the missing buttons that are needed?
1 x Thanks button
1 x Withdraw button.

At the moment, there is only one visible option:
1 x Agree button.
The lack of visible options here, is surely, one of the root causes?

Posted by sbushes over 3 years ago

FURTHER TO THIS. How this impacts apon IDing of Vespula wasps. Another thing i do besides checking research grade of nz plant obs (by the thousands) is to check the research grade obs of Vespula in NZ, Ausi, South America, and Africa which are a very problematic genus and 99% of people have no idea what they are doing in them. As it stands now when people pull up the research grade Vespula wasps of these places there is a gallery, some of the obs have an agree button and some dont.


there is nothing to stop some overzealous IDer from running along the gallery page agreeing with everything that has an agree button! But that is a great button! I use that button to clear out the obviously right ones. then go in and look carefully at the others. on obs where there was some doubt and both germanica and vulgaris were given as IDs i could just work out which it was then push agree on the right one. but you have removed that very useful button (and not the other one go figure) now I have more typing to do grrrr.

so unfortunately you have not solved the problem of overzealous IDers at all.... you have just increased the amount of time and effort it takes us to check the research obs.

some thoughts on ID from me would be:

if someone posts an obs without an ID do not let them confirm someone elses ID.
if someone posts an obs and the ID proves wrong, only let them withdraw, dont let them add another one.

increase the number of IDs needed to flip an obs to Research

Do not let new users ID to species level. Make them clock up a certain number of IDs in a taxon group BEFORE letting them loose to ID whatever they want. (obviously this would have to be wavered for experts in their field but they could just apply to a curator)

Suspend the IDs and IDing capability of anyone that has a significant number of maverick IDs. Iv done tens of thousands of IDs and have only a handful of maverick IDs because I am extremely careful when IDing. I also go back and check and fix the maverick ones (or keep them if I am absolutely certain i am right and the other three IDers are blind). But there are mass IDers here that have hundreds and hundreds of maverick IDs and never bother to go back and fix them, this means they can be very careless in their IDs (which allows them to do them very fast), If they were forced to go back and address their mistakes or face suspension of their ID rights they would also be forced to slow down and be more careful.

Instead of looking at gaming as a negative thing, look at they way that they control and reward their communities, and then implement the same sort of things here! the ones that connect people like having a friends list and the ablity to see who of them is online and be able chat to them, compulsory training programmes on IDing and making OBs for new users, limitations on new users till they "level up" and "completion awards" eg an award for making confirmed obs of all species in a genus in ur country etc etc.

Posted by dave_holland over 3 years ago

@cmcheatle No, I really only mean removing the convenience of the "Agree" button for the original observer, because in the vast majority of cases people click that simply because it's there and looks like an easy way to say thanks for an ID. It should still be possible to change an ID, or manually enter an agreeing one as @sbushes mentions, but that is a much more deliberate action than simply clicking a button.

Posted by teadrivendev over 3 years ago

@dave_holland "Do not let new users ID to species level. Make them clock up a certain number of IDs in a taxon group BEFORE letting them loose to ID whatever they want."

This seems like an absurd suggestion. So if a new Australian user joined, they would be 'banned' initially from IDing magpies, rainbow lorikeets and other extremely easy to ID taxa? IDing charismatic taxa like birds is often a gateway to becoming more interested in other groups because they're easy to recognise/learn about, so you get a sense of accomplishment. Preventing new users from IDing to species would strongly discourage them and certainly wouldn't create the sense of community iNat fosters.

"if someone posts an obs and the ID proves wrong, only let them withdraw, dont let them add another one."

Also seems a bit bizarre? People make mistakes and learn over time, why would you remove the ability for them to rectify their mistakes and add a new ID based on their new knowledge?

Posted by thebeachcomber over 3 years ago

@dave_holland Such a change would be likely much more negatively impactful for the community than this change will ever be. Just my opinion.

Posted by calebcam over 3 years ago

I don't know about you, but I find that this time of COVID-19, weird politics, and relative isolation is bringing some good things -- like hours and hours each day on iNaturalist -- but also some problems. I can be sailing along fine and the start sobbing over an inability to find a particular book. I can get angry over stupid stuff (and see a couple of my relatives reciprocating sometimes). Both my wife and I reason to stupid conclusions or fail to do simple math, then realize it's another example of what we're calling "COVID brain." Reading this thread, I'm happy to see that so many of us care passionately about iNaturalist, but I also wonder if this difficult year isn't getting to many of us. So I'm going to be kind enough to stop posting for a while. (At least eight hours!) You don't need my fevered concerns.

Posted by sedgequeen over 3 years ago

I have very mixed opinions on the "Agree?" change. I'm one who often doublecheck hundreds of "Research Graded" observations not because I'm in it for the identification count but to ensure each id is correct, especially in my field of expertise. So to manually type in the name of the species in these identifications will definitely be a pain on my part. On the other hand though, it ensures that I don't follow the flow and blindly agree with an observation that could very well be incorrect. Pros and cons for me.

Posted by birdwhisperer over 3 years ago

This is a real pain. Many "Research Grade" observations are wrong. It's very useful to be able to click agree to double check those that are correct. The loss of this feature is a retrograde step in my opinion.
The only way around the problem now is to enter a confirming ID, if that is case. It's much more time consuming to do that.

Anyway shouldn't the heading be "Fewer Agreeable Observations". It certain is less agreeable in my view, and retrograde step for iNaturalist!

Posted by mef over 3 years ago

Has anyone asked for this change?
If not then this seems to have been done by someone who has more interest in computers and figures rather than natural history.
What is a problem is someone who posts something unknown to them and then immediately agrees with the first suggestion; in doing so the observation becomes research grade. That is the issue that needs fixing and not barring someone who quickly wants to double check an observation.

Posted by mef over 3 years ago

For what it's worth, I agree with all of the points @sedgequeen has raised.

If someone knows a taxon very well, and I am aware of this fact, I look for their ID and disregard the rest as well as I can. So I get a lot of value out of those IDs, whether they have an influence on community ID and "research grade" or not. The identification summaries are not the only valuable parts of the identification history. An "agree" button as a substitute in such cases would seem to preserve the information, but make it harder to retrieve.

If I don't know who on iNaturalist has expertise in a given taxon, the leaderboards can be a good way to get that information. A couple of times, it's turned out to be people I know, but didn't know were on iNaturalist. I'm not sure if changing the leaderboard based on categories of agreement would affect the utility of leaderboards for this purpose. My intuition is that the number of "improving" identifications is well enough corrected with the raw number of identifications that it wouldn't make much difference, but my mental model of how the different categories play out in practice is not very good.

There is a functionality that I've wanted for a long time that seems like a natural fit for the identification leaderboards. Suppose that the number of identifications over at the right were also a link to the set of observations identified as that taxon by that user. This could address some of the perennial questions about taxonomic expertise on iNaturalist, by taking what I think is the path of least resistance--making it easier to access iNaturalist data in flexible and nuanced ways, rather than tacking a database of experts onto it.

Posted by aspidoscelis over 3 years ago

For what it's worth, I also support @sambiology's suggestion of renaming "research grade". At present, I think any discussion of using iNaturalist for research kind of has to start with, "Yes, there's a thing called 'research grade', but just pretend it doesn't exist and isn't called that." :-) It plays a useful role, but it is not a marker of observations that are appropriate for use in research, or are ready to incorporate into research without further review,

Also, @bouteloua, the server gods must be smiling on you. Back when I was doing high volume ID, I'd find that the step between typing a taxon name in and having the pull-down list populate was frequently the limiting step. Sometimes those lists populate in no time; sometimes you wait... and wait... and wait. I've wanted a "just type the name in and hit enter" function since I started using iNaturalist. It's reliably faster in aggregate and always requires less attention per instance. It becomes only a text entry task, rather than rapid switching between concurrent text entry and text recognition tasks.

By the way, I'm glad to see the html tags still seem to work, since I'm more used to using them.

Posted by aspidoscelis over 3 years ago

Yeah I used to have that search hang problem alllllllll the time, but thankfully haven't encountered it in a really long time. Nothing has changed about my setup so I think there were some improvements to the search dropdown.

Posted by bouteloua over 3 years ago

Removing the Agree button makes it harder to identify down to subspecies and variety. Many of us work hard on identifying to that level, and now we have to type in the subspecies/variety.

Posted by birgitknorr over 3 years ago

I had my comment, and then went away for three days to cool off, and decided to read through all the discussions, and see if my opinion had changed.
However I only got as far as @bouteloua before I my adrenalin levels hit the roof again.
The Identity Tool (I call it the curation tool), is in my mind the most important place that the I Agree button must be included. Not having the "I Agree" button changes the task of reviewing a genus or family from a pleasurable half hour catchup, to over an hours slog to accomplish the same thing.
It is especially irksome that this was chosen as an example of what we should do, when it is this tool that is most compromised by the removal of the "I agree" button (regular users {& gamers!] will have discovered by now that the Community Summary "I Agree" is still there and quite useful in these beleaguered times).

If the goal is to reduce gaming, then design something that accomplishes it without making the site a pain for everyone else. Just how many gamers are there? A handful? Do you even know who they are? How do you know? Then, design a tool that will stop gaming without inconveniencing your power users and identifiers.
The problem with removing the "I Agree" button is that it will slow down serious users far more than it will slow down gamers. Gamers will just change their approach:

select a filter for "needs ID" and go for it - thereby foreclosing the "I Agree" for more reputable identifiers.
select any species in the ID tool, and load the species name into the copy/paste memory and paste the name on all the observations not yet already reviewed.
The problem with this proposed "solution" is that the gamers will get an even bigger advantage. They will still be able to add IDs at a 10-15% penalty, but serious and useful identifications will incur a 20-30% penalty in time and effort.
Is this proposed solution supposed to stop gaming, or encourage it?

Posted by tonyrebelo over 3 years ago

@bouteloua - That's good to hear. I haven't done much binge IDing recently, so I guess my recollections are becoming obsolete. :-)

Posted by aspidoscelis over 3 years ago

Excellent update @kueda, thank you. I think every change feels uneasy at first.. but I believe it will be very positive.

Posted by carancho over 3 years ago

Another worrying issue is how the proposed changes leaderboards will be managed.

In order to try and get specialists and leading experts to work on iNaturalist, we have negotiated that the identification summaries on user profiles will be used to audit contributions by South African National Biodiversity Institute staff to iNaturalist - this amounts to many hours of expert identification in diverse taxonomic groups in a region with very few experts, a huge number of diverse species, and a lack of field guides (beyond birds, butterflies and trees). So a major investment of time for a major return.

This works well (the biggest gripe being that identifications on one own's observations are not included in the tally - but this is only relevant where the expert on some obscure group contributes most of the observations as well: for instance, if an earthworm specialist would add 80 observations and ID them, and ID 20 other observations, they would only get credit for the 20 other observations, a bizarre state of affairs).

Please make sure that any deprecation in the reporting of identifications, does not remove the Identification totals from user profiles. I realize that this might be an incentive for gamers to abuse the system, but it is the only way that we have been able to get staff at our major herbaria to be able to help with identifications on iNaturalist. Do you know how difficult it is to convince managers that making identifications for laymen on a website is actually useful work, rather than goofing off during work hours? Being able to quantify staff involvement has been the single most useful way to leverage professional support. Whereas the profiles are essential, the "Leaderboards" are useful but not vital. Professional identifiers are far more meticulous than most identifiers and seldom process the large numbers of IDs that enthusiastic volunteers without a professional reputation can make. However, ,it does help to state that x identified 60% of observations compared to the leaders, especially in groups were IDs are difficult and the proportion of observations ID'd to species is low.

So just a heads-up please: dont mess too much with the identification summaries. The leaderboards may be game, but hiding or deprecating the summaries on user profiles will have much more serious consequences.

Posted by tonyrebelo over 3 years ago

@roget What do you find positive about it? I see it doesn't affect you as much as it does those of us who do a lot more ID's. If that means you are spending more time in the field, I respect that! I see you have done 2,600 ID's in 5 years. By contrast Tony Rebelo has done about 144,000 IDs in less than 4 years.

Posted by stewartwechsler over 3 years ago

rolls eyes Must be "useless" identifications ... since it is higher than 10,000 ...

@tonyrebelo has also been on iSpot (before he and the SANBI came to iNat) for many many years, I may note ... so he really has experience in this kind of thing :-) Absolutely agreed about the leaderboards, and it is extremely worrying to see where iNat seems to be headed.

@sedgequeen Yes, we are all a bit 'under the weather', probably. 2020 has been unforgiving :-(

Posted by calebcam over 3 years ago

••• Re the removal of the AGREE button:
►A terrible mistake.

Others have stated the case against this change very well. I am in complete agreement that this change should be reversed, for numerous valid and even obvious reasons.
••• Re the possible removal of the leaderboards:
►It would be a terrible mistake.

(a) Is gaming really a problem and has it been quantified?
(b) In the absence of any other measure of expertise on iNat, leaderboards are critically necessary to guess-judge a stranger's expertise; and to know whom to 'call in' for ID help. It is really easy to look through the number chasers tofind the pool of probably true experts.

Please do not break iNat twice.

Posted by beetledude over 3 years ago

I have to say the comments of @beetledude and @calebcam (just above) ring true for me as well. Yes, 2020 has been a dumpster fire of a year and it is not getting any better. Given that, especially, please don't break iNat. It is the only thing keeping some of us half way sane....don't make it more difficult to use.

Posted by greglasley over 3 years ago

I'll be honest, I haven't read all of the comments in this thread but I can sympathize with some of the complaints. I think it's worthwhile to keep the Agree button, and the only problem I've encountered is the notification spamming. It could be useful to have certain users or scientists granted some iNaturalist Researcher status - perhaps based on seniority on a taxon and/or familiarity with the site - such that any 'Agree' or ID modification these users submit will generate a notification, and 'Agree' submitted by other regular users can just be muted (similar to the quiet thumbs up/thumbs down votes).
A lot of the comments are against this volume 'gaming' - the differentiation in user credibility will hopefully weigh against them, but honestly, you can't stop people from doing what they want to do. Frequent, long-time users on this site know each other's specialties well enough, and the feature would more likely benefit newcomers and learners. It would be a shame to lose the crowd-sourced community charm of iNaturalist to a small group of folks who don't know any better.

Posted by achang over 3 years ago

I'll add to the chorus of "this is a bad change that discourages experts from confirming IDs". Find a better solution for whatever issues there are with users gamifying the leaderboard. As an "expert" who spends quite a bit of time curating certain groups on here, I need a quick way of confirming identifications. The notion that once an observation reaches Research Grade it no longer benefits from additional IDs being added is ludicrous. Taxonomy as a science is built on the consensus of experts... this change fundamentally hamstrings such a consensus by making it a tedious chore to confirm IDs.

This is an awful decision and I'm amazed it would get implemented without input from the community.

Posted by joe_fish over 3 years ago

@joe_fish, absolutely agreed!

Posted by calebcam over 3 years ago

"Agree" with so many of the comments here. This change is pointless and makes the job of the expert more tedious. I haven't read all these comments but there is no "Agree" showing up when there are several suggestions and one wants to agree with one of them. You end up having to type in the name and choose it. Tedious!

Posted by fmr over 3 years ago

My 10c

Is gaming really a problem? If it is, then where? In what groups?

scholar/graduate projects. This is the only place where I have really encountered a problem. And this is where "gamers" are 'playing' for marks. It is localized to a few weeks, at one place, in one group (every year). Most of the "gamers" then vanish - never to been seen again, leaving a mess that needs cleaning. But as a proportion of observations it is trivial. Its annoyance value is also trivial (compared to the missing I Agree button). Two points here: (1) the lecturers/teachers need to be pressurized to prevent this - there is guidance online, but perhaps more thought needs to go into this. (2) the benefits of participation and learning probably far outweighs any inconvenience to other iNatters - despite the problems, I would love more schools and universities using iNat for course work.
real gamers. I have never encountered them. Perhaps we dont have enough users in southern Africa. I cannot relate to it. In the groups that I "curate" I have never detected any gamers. Might a better solution be to allow other users to flag users as potential gamers - with their reasons? Yes, it will be more work for curators, and will possibly allow personality clashes to arise, but might it not be less inconvenience all round? Gamers could have restrictions (such as no "I agree" button, or worse) imposed, if found to be true.
apparent gamers. Anyone making a concerted, intense, real effort at identifications in a group could be misconstrued as a gamer, if it was not possible to realize that they were specialists or experts in the field. (Un!!)Fortunately, in any group there are usually so few experts, that any mavericks are easily identified, and usually well known. These are our quality identifiers and we need to do everything and anything to make the process as smooth and as quick as possible. (and forget about keyboard shortcuts that are in fact convoluted roundbabouts for a one button click). Part of the problem with specialists is that their contribution will primarily be identifications, and their observations and other contributions will be minimal. So measures that restrict users for not contributing observations, penalize specialists. And in groups that are poorly known, specialists will run the risk of posting "maverick" IDs and IDs that are not agreed to.

If the problem is gamers, can we not make identifying and dealing with gamers the issue, rather than making the entire site less fun, easy and useful?

Can we not turn the issue on its head?: can we not find a way to encourage gaming and to incorporate it into the procedures and checks and allow it to make the IDs more accurate or refined? Perhaps a way of getting users who make IDs to improve the quality of their IDs and get rewarded for it?
Can it not be integrated with the AI, where each cross-checks the other? Where gamers can compete with other gamers and the AI? By necessity, this will be most helpful at the upper ranks, moving IDs from meaningless "Plants" to orders, families and genera. Such gaming will require penalties for too detailed identification. Alternatively it can be part of a background ghost system, where iNat "offers" finer level IDs on poorly identified observations based on gaming IDs, versus the basic user IDs. (unfortunately, iNat is not interested in a reputation system that will allow users to train both other users and the AI, but we have to work within this system: but the fact remains that a gaming system that trains gamers to be better identifiers is a win-win situation)
((It is worthwhile noting though that a reputation system will - short of a hack - solve the gaming issue!))

One possibility to limit gaming is to keep the leaderboards, but remove the actual numbers, so that gamers do not have a easily discernible target to aim at? Would that help to show who the "experts" are, while reducing the incentive to game? [I dont like this idea personally, but might it help?]

What about a restriction, that users with more than x maverick IDs (say 10) lose the "I Agree" button until they have fixed their maverick IDs. (so long as this does not result in a "game" of agreeing to maverick IDs). [My problem with this idea is simply: how many users know how to find their maverick IDs? - How many maverick IDs do you have? I have just discovered that I have 95]

Someone objected to new users having "restrictions". Is that the correct way to view it? In many other societies/sites/communities one has to prove oneself before being accepted as a full member. Can you post a Project? The correct level to allow full access is debatable (and will generate as much hot air as "research grade" - both level required and its name), but perhaps the bar for IDs should be as low as possible - but high enough to exclude school and graduate courses in their gaming.

Posted by tonyrebelo over 3 years ago

Oh dear? my html code is gone, and the bullets that showed in the preview (one for each type of gamer) are not there.

What has happened to my text formatting?

Posted by tonyrebelo over 3 years ago

@stewartwechsler it's pretty easy for me to see... I won't get +200 notifications in a day of someone who has been reviewing all the sparrows and domestic doves observed in inat and ID a RG obs that already had been identified 15 times, something that happens to me nearly every week. Just because of this is a great update to me, I have missed a lot of important notifications because of these overzealous identifiers. Also as an admin, I saw this has proven to be a problem because people would identify anything even without knowing what they were agreeing to, and as a result we got RG observations with incorrect identifications. I think those who are commited to reviewing observations will continue contributing. I would summarize this as:

The ID is ok? Then leave it as it is, is RG already.
The ID is wrong? Then correct it.
Still want to add an ID? Then enter it manually.
Any other thing, leave a comment.

Posted by carancho over 3 years ago

@roget there's an option to turn off notifications for agreements.

"The ID is ok? Then leave it as it is, is RG already."

Let's say an observation from several years ago has 2 IDs and is Research Grade... then someone comes along an incorrectly adds a Maverick ID. That observation now gets bumped to the tail end of the Needs ID queue, where it is unlikely to be seen. Therein lies the reason why we need as many confirming IDs as possible, from as many "experts" as possible. Making this onerous by removing the quick and easy Agree button is going to dissuade users from making the effort to do so.

Posted by joe_fish over 3 years ago

One last issue before I zip my lip and get back to loading observations and making identifications and doing what I enjoy, instead of being here, upset:

As a user of the data: for Red Lists, Modelling, Taxonomy, assisting taxonomists to get data from volunteers, conservation planning and other projects, the most important thing by far is the quality of the identifications. iNat does not have a reputation system, so one cannot evaluate the quality of the identifications. The only measures one has of trying to work out whether the identifications are meaningful or junk, is:

* how many identifications/agreements are there versus other identifications made.
* how many reputable users have made an ID.

The former is extractible in the downloads as a value, and no one in their right minds should extract data from iNaturalist without including it. Simply the more identifications the more likely the ID is to be reliable, all things considered. Ten agreements is far better than two and much better than three. It is not linear and not equal, but there is no other easy measure on the system to evaluate the reliability of the ID.

((iSpot had a reputation system, which allowed one to evaluate an identification directly in terms on its equivalent to a professional identification - paid for - in a museum or herbarium. Contrast that with what iNaturalist offers for an ID: and then apparently what we get is iffy because it has been rendered meaningless by gamers. The solution: to prevent agreements and make the IDs totally unquantifiable! [which his countered by some saying that agreements will still be made: it will be more difficult: but it wont improve the quality of the data??])).

The second check is possible if one is working with a taxonomic group. One can check which observations have been identified by known experts in the region, check that they are the current names, and compile an index of what proportion of the identifications are bad.
What we have chosen to do instead, is get our experts to use the identification tool, and curate the entire collection. That involves about 60-99% clicking the "I agree" button. Two or three experts need to do this. And in cases where the pre-expert ID was wrong, we get "disciples" (who trust the experts) to click "I agree".

Yes: we game the system! But our belief is that it is better to have the correct ID on iNaturalist, than to extract the data, correct the IDs offline, and leave the wrong identifications on iNaturalist.
We then extract the corrected data from iNaturalist for analyses.

This new development will cheese experts/specialists/reputable amaeturs off immensely. The quality of the IDs will suffer as a consequence. And observers will be less certain of the identifications they get. An "I agree" from one of the taxonomical experts in a group, from a known specialist, or a respected amateur is worth its weight in gold, even if it is only number 27 in the list of agreements.

We have a competing Citizen Science platform in South Africa to iNaturalist. Its major selling point is that the IDs are made only by experts, and are therefore far more reliable than those one can get from iNaturalist, where one has no idea of how good the ID is. They are going to crow about this!!

I welcome new ideas and modules and features on iNaturalist. It is a dynamic and fun site. But this new feature horrifies me to the core (and I have sat and thought about this for half a week: this is not a knee jerk reaction). We must have a solution (assuming that there actually is a problem) that makes Identifications more reliable, not less reliable! That allows us to identify our experts and reputable amateurs on leader boards, and make the site as pleasant and as efficient as possible for them.

Posted by tonyrebelo over 3 years ago

@joe_fish but I don't want to turn off notifications for agreements, I just don't want users reviewing all the house sparrows I've seen in my life.. why someone would do that? I think it's a great way to prevent such behavior.

What you mention seems to be a possible issue, but as I said there are also other issues related to the Agree button.

Let's say an observation from several years ago has been misidentified twice and is RG (for example, this happens a lot in bugs found in the southern hemisphere). The poster goes for the ID the computer vision gives, which for arthropods is usually wrong here, and then those who want to get more IDs, blindly agree with it, making a RG; then a third person who also wants to get more ID just clicks the agree button, and the incorrect RG is made almost perpetual. I have found this to be one of the most challenging issues as an admin of a gateway.

Hopefully we will be able to find a solution for everyone.

Posted by carancho over 3 years ago

@roget I really just don't understand your argument here. You seem to be telling users not to use the site or contribute in a way that they find valuable (and many users have given multiple reasons above why confirming IDs can be valuable to them or other members above) because you don't like getting notifications that you are given the ability to disable. Notification management is a common feature of pretty much every social networking site, and users are expected to tailor the notifications they receive to their preferences.
The observations are there for people to interact with; that seems to be the core mission of iNat! This argument seems to me like telling people not to like or comment on posts you've liked on Facebook or retweet tweets you've liked on Twitter because you've made your contribution and don't want to get notifications about anyone else's activity. The point of a social network is to encourage interactions amongst users, not tell the users to stop interacting with each other after one person has given their input.

Posted by cthawley over 3 years ago

@roget I recently curated ~2000 old observations of Amphiprion, a group I have considerable expertise in. I spotted several such Research Grade misidentifications that are now corrected because of my efforts. I only attempted this curation (which took many hours on my part) because I could simply click Agree on 99% of them. If I had to manually type out the scientific name on these, I would never have bothered and those misidentifications would still linger.

Any changes that discourage this sort of curation need to be avoided at all costs. There surely must be better solutions for whatever problems this change sought to address. Ultimately, if this isn't reversed, I'll simply spend less time making IDs on here, which will ultimately degrade the quality of this site in the areas that I curate. Judging by the mass of negative comments on here, I'm likely not the only one who feels this way.

Again, I am truly baffled that such a ham-fisted change to this site's functionality was added without so much as a peep of discussion with the broader community of users. Terrible communication by the iNaturalist team.

Posted by joe_fish over 3 years ago

Mixed signals from iNat - On May 15, 2020, iNat posted a blog praising the users that contribute identifications (see https://www.inaturalist.org/blog/35758-we-ve-reached-1-000-000-observers). But now they change a feature that sends the message, "we don't want multiple confirming ID's". And now we are speaking of "leaderboard gamers" when iNat previously praised "super-identifiers." Why the mixed signals??? The removal of the agree button has apparantly disatisfied the identifying users they were hoping to attact. Some of the users praised in the blog have even spoken up about their displeasure in removing the agree button. And some of those users praised in iNat's previous blog are on the "evil gamer" leaderboards! Read this excerpt from iNat's 5/15/20 blog and tell me if it lines up with this new policy of making ID more difficult (emphasis below was added):

"Are you an identifier yet? One vulnerability that this very long identifier tail reveals is that even though iNaturalist has over 2.5 million users, the site is extremely dependent on a much, much smaller group of super-identifers. You can see the top 500 identifiers here and I’ll name just the top 10: @aguilita, @sambiology, @greglasley, @johnascher, @maxallen, @john8, @graysquirrel, @maractwin, @joshuagsmith, and @thebirdnerd. Please join me in thanking the small group of super-identifiers for literally making iNaturalist function. But there’s also an opportunity to take steps to try to grow this identifier community. Statistically speaking, if you’re reading this there’s a good chance you’re not one of these super-identifiers. We definitely encourage you to give identifying other people’s observations a try. Who knows, you might be our next super-identifier! Here’s a video on how to use the identify tool and here are some tips on how to dive into identifying. There’s also lots of good ideas on how to recruit more identifiers on the iNaturalist forum such as this thread."

Posted by williamwisephoto over 3 years ago

@cthawley I'm not much of a Facebook user, but actually if you try to RT a lot of tweets or follow a lot of users in a very short amount of time, your activity will be limited for a while and you won't be able to do so. I think it's healthy. But I don't want to convince anyone, and I my opinion may change later, but by now I just think is good.

As for experts, it's a tricky thing. I've been told by users who were top identifiers thanks to the agree button (and who had commited several wrong IDs) that they should be treated as authorities only because they were there.

Posted by carancho over 3 years ago

Since my name was mentioned by @williamwisephoto above as a "super identifier" among the top 10, I suppose I should mention why I spend a lot of time doing this. I am not a "gamer" and as a 70 year old geezer, I'm not even sure what a gamer is or what the point is. Sorry. I have been involved in wildlife and natural history since I was a small kid (so 60+ years) and have some degree of competence with birds in the U.S., Mexico and central America as well as odonates in the U.S. and butterflies and moths in Texas. An iNat staff member did an interview with me two+ years ago and I explain why I do the IDs:
I will do my best to keep at it. iNat staff folks know that I am currently dealing with a medical situation that keeps me home more than I would like and unless and until that changes I will continue to spend more time making IDs than I am able to spend making new observations. I hope this current iNat issue can be resolved to the satisfaction of everyone involved. I have previously expressed my opinions on it. Obviously many of us are passionate about iNat. I have read every comment here and I hope iNat folks will as well. My best regards to all who have commented here regardless of your stance on the issue.

Posted by greglasley over 3 years ago

Regardless of your views on this change, no one is being 'stopped' from doing anything, including adding ID's.

They may choose to stop, they may feel that taking a second or two to enter some text is too high a barrier for their time, but no one is being stopped from anything

Posted by cmcheatle over 3 years ago

I am surprised at the response here, to say the least. Let it be known as Agreegate 2020. From my perspective, removing some of the "agree" buttons was a minor change to discourage behavior we on staff view as unnecessary (at best; facilitating a culture of competition at worst). I wasn't imagining huge gains, just a small corrective. As such, I'm perfectly happy to roll it back since the cost to some of you so clearly outweighs the benefit, and will do so later today.

Thanks to those of you who have expressed your critiques in practical terms, describing how it affects your workflow, especially those who have taken the time to test it out for a day. I'm closing comments here because at this point I'm mostly seeing re-iteration of the same themes Scott outlined above. We on staff have read all the comments and while we don't agree with all of the opinions expressed, we are taking them into consideration.

Posted by kueda over 3 years ago
Comments have been disabled