In the spirit of my posts on racial diversity on Manhunt.net and Adam4Adam (see NYC; SF; Atlanta posts), I'll be posting new data about the variation in tops, bottoms, and versatile-identified (what I call "positional identity") profiles on these two websites. I was holding onto this data, thinking perhaps I might do something more formal with it, but I decided in the in to share it here on my blog.
As a recap, what I've done here is reconstructed the two website's database by doing repeated searches of profiles using narrow categories. For instance, if I want to know how many top-identified profiles exist among 18-21 year-olds, I did four searches (one for 18 y/o tops; 19 y/o tops; etc). Note that profiles ARE NOT people. And this technique for amassing data is somewhat vulnerable to human error. So take this data with a grain of salt. We need more research on this topic, to know how these identities are distributed in our communities. This is just a shot in the dark, really.
Also note that this data does not include anyone who did not identify as top/bottom/versatile. People who identified as "top/vers" were categorized as "top" here; "bottom/vers" also as "bottom." Without further ado, here's the data. There are 3,755 profiles from Manhunt documented here, and 4,177 from Adam4Adam.
Adam4Adam
As I did last time for race, I've gathered data on the distribution of positional identity categories by age group. Note that the last data point (58+) includes very few profiles and is incredibly vulnerable to bias because of the number of people who put "99" or "69" as their age instead of their real age (a rather common practice).
But what you see here is quite fascinating, in any case. Note the very high percentage of bottoms among 18-21 year-olds, and the comparatively low percentage of tops. This disparity quickly reverses -- by the 22-25 year-old group, they're stastically dead-even, and then for every older age group there are more top-identified profiles than bottom-identified profiles. Curiously, the percentage of versatile profiles stays about the same over time.

And here's the breakdown for all the profiles, overall. 1/3 of profiles identify as top; 40% as versatile; and just 27% as bottom:

Manhunt
Now let's turn to Manhunt. The important question: is there any obvious differences from Adam4Adam, as was found with racial diversity (A4A's profiles were consistently more racially diverse than Manhunt profiles)? What we see here is interesting, compared to Adam4Adam. While there are similarly many more bottom-identified profiles in the youngest age bracket, it takes until the 34-37 year-old age group for there to be less bottom-identified profiles than top-identified profiles.
Strange. Why the "delay" in the shift? Manhunt profiles tend to be more "white" than those on Adam4Adam, though I'm not quite how that could impact the data. Is "bottom" a more stigmatized category for men of color? That would be an interesting question for more research! Here's the data:

Overall, however, the data looks about the same. Almost exactly 1/3 tops; 38.5% vers; and 28.4% bottom-identified profiles:

What does this tell us? Several things. First, it appears there's a correlation between age and positional identity. We obviously need some "real" social science research here to explain this variation with age. Second, it appears that there is a difference in this variation between Adam4Adam and Manhunt. While comparing this with the race data I've previously presented, it's tempting to link the two differences. However, we can't even say this is a correlation yet. We need "real" data.
Hope that invigorates your curiosity! I'll be back soon with Atlanta and NYC data.
xoxo
T
Both of these data sets seem to have similar trends. Do you have the data wrt to racial identification? I wonder how recon.com data would compare.
Cool observations btw.
No the data is reconstructed via searches, and I didn't cross top/bottom searches with race searches. So I only know the outcome of searches of lump categories, not individual profile demos.
I wonder how wrapped up positional identity is with the coming out process? My a priori assumption would have been that 'versatile' would be most common among the younger guys, settling into the relatively more fixed roles as they get more of a sense of who they are and what they want.
Although, these websites may well draw in the guys who more or less have got that stuff figured out.
I also would have expected that the young guys would be more likely specify themselves as tops, though I'm not sure why I think that.
I'm curious about the relative proportion who do not define themselves as any of the above, my hunch would be that the undefined group would be largest in the young guys, but half of my intuitions were wrong already...
You're quite right to wonder about the number of profiles who don't readily identify -- many do not indicate their positional identity as top/btm/vers. On Manhunt, the alternative options are: "Ask Me"; "Into JO Only"; and "Into Oral Only". Guys who are in any of these categories may actually identify as top/bottom/vers offline -- but they don't wish to disclose that on their profile for any variety of reasons.
Great work Trevor, love it! Yeah, for positional identity research - we'll have to pursue this further . . . I've managed to include positional identity questions in some of our current SF work. Can't wait to start collecting data . . . hopefully you'll be able to correlate some of this with your bottom work too. Can't wait to see the other gayopoli - here's to datamining.