Editing Openai/6946e998-79f0-8007-9a65-19828450838d (section)

==== You also suggested: “Take a random highball estimate for the number of mini-Neptunes… and thus not overcount.” ====

In practice:
* In the 1–4 R⊕ range, a lot of planets turn out to be low-density (volatile-rich) once you get mass measurements.
* In the 1–2 R⊕ range, there’s a known radius valley: planets below ~1.5–1.7 R⊕ are more likely rocky; above that, more likely to have gas envelopes. So if you cut off at ~1.3–1.4 R⊕, you’re already rejecting many mini-Neptunes.
* A “highball” correction might be: - Assume that, in the 1.0–1.4 R⊕ bin, maybe 30–50% are actually volatile-rich “stealth mini-Neptunes” you can’t yet identify individually. - Multiply your current η⊕ by ~0.5–0.7 to reflect that.

So if your conservative-HZ, tight-radius FGK η⊕ came out around 0.2 per star, and you then assume “maybe half remain after we eliminate hidden mini-Neptunes,” you get:
* Effective η⊕,rocky,conservative ≈ 0.1 per Sun-like star as a plausibly cautious number.

For M dwarfs, you’d likely apply even harsher penalties, both for volatile envelopes and for stellar activity stripping.