How big would Mercury need to be to retain an atmosphere? How close could Venus orbit the Sun before its atmosphere erodes away? Are habitable Earth-like atmospheres even possible around the smallest stars? In this talk, I describe how exoplanet observations are starting to provide insight on what environments permit terrestrial planet atmospheres to thrive. I present a new probabilistic model for a boundary between planets with and without atmospheres, one that is rooted in remotely measurable quantities, accounts for how drivers of atmospheric loss can change strongly with stellar type, and quantifies the charming, complicated, crinkly, chaotic fuzziness of planets’ diverse individual life histories. I use this model to predict where we might most productively search for rocky exoplanet atmospheres in the near future, I highlight work we are doing to mitigate the astrophysical challenges currently limiting us from observing some of those planets, and I outline goals for continuing to train generations of Colorado Researchers in Observing Chromatic Exoplanet Time-series (CROChET) with curiosity, creativity, coding, and kindness.