Anyone else notice that there were more identifications than observations this year on iNaturalist’s year in review? 60 someodd million identifications compared to almost 50 million observations. I feel like that’s a healthy ratio, though obviously there is a huge ID backlog. Any info on how that ratio has changed over the years? Also, many thanks to all the people who contributed both observations and identifications.
8 Likes
Does that count include the IDs people put on their own observations? (Edit: evidence seems to indicate the initial observer’s ID is not counted.)
You can look at past years by changing the year in the URL:
https://www.inaturalist.org/stats/2023
2024: 48 million observations, 65 million IDs.
2023: 41 million observations, 59 million IDs.
2022: 34 million observations, 49 million IDs.
2021: 29 million observations, 40 million IDs.
2020: 24 million observations, 31 million IDs.
So, it looks like this year continues the normal trend.
7 Likes
I did the math on the ratios of IDs to observations, to make it easier to compare years.
2024: 1.35
2023: 1.44
2022: 1.44
2021: 1.38
2020: 1.29
Worth noting that 2024 is not over yet, and there’s still a month for both IDs and observations to roll in. I know I’ve got a huge backlog of observations from the summer I’m chipping away at, and when the cold gets dreary I’m planning to ID as a form of virtual exploration.
8 Likes
Pretty sure it includes the identifications from when a user uploads. So if there’s 48 million observations and everyone added a starting ID so that it isn’t unknown. No There is not enough 2nd identifications to make even half of all observations uploaded in that time RG.
1 Like
I don’t think that’s the case. There are 26 million research grade observations from 2024: https://www.inaturalist.org/observations?d1=2024-01-01&quality_grade=research
That means a minimum of 52m IDs
Plus 21 million needs ID: https://www.inaturalist.org/observations?d1=2024-01-01&quality_grade=needs_id
That brings the ID total to 73m IDs minimum. Given that this is more than the 65m in the year in review, and this is necessarily an underestimate (due to all the canada geese observations having a minimum of 8 IDs each), it must not include the initial ID made by the observer.
5 Likes
As others have mentioned, that is not enough IDs to cover every observation twice which would be the ideal situation.
However, the potential number of IDers is so much higher than what it is currently.
If we could get green+orange to be 1/4 of the donut that would be great. 1/3 even better. 1/2 utopian…
If everyone who made an observation this year made even just 1 ID every week that would be roughly 60,000,000 additional IDs. That would almost double the current number.
5 Likes
I analysed the UK Hoverflies a while back and I found that the average observation required just under 2.5 IDs to reach research grade including the original ID. I don’t known how that varies between taxa/location. I suspect that the 60,000,000 IDs quoted by the OP does not include the initial ID-on-upload (there can’t only have been 10,000,000 IDs-for-others), so it would need 75,000,000 such IDs to ‘keep pace’.
I also worked out that, in Animalia globally, 70% of IDs are made by 0.16% of identifiers (roughly the top 500). I wonder if more could be done to encourage a small amount more identifying by a larger cohort of people in the lean winter months. Perhaps some people feel there’s no point identifying if they can’t do a lot. But it’s actually a great way to keep in touch with nature - and with iNat.
Nevertheless - great work to everyone involved - both those observing, and those IDing, and those doing both.
2 Likes
Presuming we upload with our own ID, a ratio of 1.25 allows for the obs which need discussion and more than one more for CID. 2 and a bit seems reasonable.
I agree with what’s been said here that the expertise of iNat’s top 500 identifiers is an extremely valuable resource that’s probably not worth risking burn out with this task. I just committed a change that makes the panel appear not to the top 500 identifiers (~>100k IDs) but rather to identifiers with between 25k and 50k IDs who have been active in the last month (there are currently about 1k candidates).
@loarie (https://www.inaturalist.org/posts/100580-identification-pilot-to-onboard-new-users)