The assessment of a planet's potential to host life is a complex field, and as a result, several different models and indices are usually used to focus on different aspects of habitability. A truly universal formula is elusive because we only have one example of a planet that can sustain life: Earth.
In this game, the habitability score used is a weighted average of dozens of planetary and stellar characteristics based on real data and real models.
The individual scores range from 0.0 to 1.0. Because data availability is not constant across all celestial objects, this approach tries to derive missing data when possible, and null scores have a penalty, otherwise planets with missing data could outscore a well-rounded planet.
So, to try and calculate a fair score, I used the hypothesis explained below.
The Roche limit is the minimum distance a planet can orbit its star before being torn apart by its gravity.
I used the simplified formula for a rigid satellite based on the star's radius and density and planet orbit axis and density. A rigid satellite is an idealized object that is held together only by its own gravity and doesn't deform or stretch as it gets closer to the star.
The Roche limit is directly proportional to the star's radius. A bigger star naturally has a larger danger zone around it. Conversely, if the planet is very dense compared to the star, it's held together more tightly, and it can survive closer to the star, whereas if the planet is fluffy (has low density) compared to the star, it's easier to tear apart.
The Circumstellar Habitable Zone (CHZ) is the the region around a star where liquid water could exist on a planet's surface.
I used the Kopparapu model with a flat plateau of 1.0 across the entire conservative zone and then a smooth down slope through the optimistic zone, as a simple gradient peaked at the center unfairly penalizes planets like Earth, which is perfectly habitable but located near the inner edge of the Sun's conservative zone.
The host star's temperature is used to calculate the fluxes with the model's coefficients. If it is not available, the Kasting simple luminosity model is used instead.
A small planet is unlikely to have enough mass to sustain the geological activity and atmospheric pressure needed for surface liquid water, whilst a large planet is very likely to be a mini-Neptune, possessing a thick, crushing gas envelope that makes them uninhabitable on the surface.
The radius must be within 0.5 to 1.5 Earth radii to be considered rocky. Rocky planets are typically above 3 g/cm^3 (Earth's density is ~5.51 g/cm^3), so the telluricity is also factored in.
Mass is needed for gravity to hold onto a substantial atmosphere over billions of years, protecting it from being stripped away by stellar winds, and to retain enough internal heat to power long-term geological activity like plate tectonics, which is vital for cycling chemicals and nutrients. Mars is a classic example of a body that lost most of its early atmosphere.
Conversely, a planet with a very strong gravity will likely hold onto a very thick hydrogen and helium atmosphere from its formation, turning it into a gas-dominated mini-Neptune with no solid surface.
The mass must be within 0.1 to 5.0 Earth masses to be a viable candidate.