Assuming that your spreadsheet was a big bulk statistical data collection, it sounds like you happened to get one of the statistical outliers. Distribution across all planet types is directly solvable in principle, I'm sure, but I'd certainly do some sort of empirical test.

Just T vs ST, though, is simple to do directly. Their populations are identical except for T and ST worlds, so we ignore those. T and ST worlds are always rocky liquid-water-zone planets, so we don't need to worry about the distribution there, either. The differentiation is when you roll for mass: 1-25 (one part) gives neither, 26-75 gives mass 2 T (2 parts), 76-100 gives mass 3 ST (1 part). So, exactly twice as many T as ST.

HI then divides the T set in half -- 1 part is within 2 of HI (call it A), and 1 part is not (B). Each of these parts is the same size as the 1 part ST (C) from the preceding distribution.

Thus a T race has A Benign, B Harsh, and C Hostile, while an ST race has C Benign, A Harsh, and B Hostile. But since A=B=C, they're the same numbers.

The last thing is considering twin planets. According to W7.05.1, "...the moon is treated as a planet, with the same mass as the other planet." If that's the case, then the ratio of T worlds to ST worlds is maintained (minus a very small shift to account for ST worlds having a slightly better moon quantity modifier than T worlds -- but note that this makes ST worlds slightly more common).

However, according to Table W6.04.2, note a (and matched in that table as found in EE), "except: 85% of T/ST twins are instead the opposite type world". If that's the correct interpretation, the balance tilts further towards ST (85% * 2 parts from T + 15% * 1 part from ST, vs 15% * 2 + 85% * 1 for T -- 1.6 ST twins for every T twin). That will be somewhat abated by the lesser likelihood of twins generated from T worlds, but not by enough to make T twins more numerous. Either way, though, ST-native real estate is not only not less common than T-native real estate, it's slightly more common (though not, I think, enough to matter).

Regardless of this particular discussion, I think it would be good to clear up the W7 vs Table W6 language.

---

Detailed moon math follows:

A T world is 10% 0 moons (0% twin), 55% 1 moon (1% twin), 30% 2 moon (2% twin), and 5% 3 moon (3% twin). Weighted, 1.3% chance.

An ST world is 55% 1 moon (1% twin), 30% 2 moon (2% twin), and 15% 3 moon (3% twin). Weighted, 1.6% chance.

So, under the W7 interpretation, ST worlds are 0.3% more common than either half of the T world divide (there are still ~twice as many T worlds, but the "within 2 HI" and "not within 2 HI" parts are each slightly smaller than the ST part).

Under the W6 interpretation, ST is .013 * .85 * 2 + .016 * .15 * 1 (chance of twin * chance of mass * parts), while T is .013 * .15 * 2 + .016 * .85 * 1. ST worlds in this scenario are 0.7% more common across the universe than each part of the T worlds.

For what these numbers say the GG5.06 table should look like, if different caps are desired, here's a ~0.5% difference:

- Code: Select all
`Hab T ST`

Benign 3015+252*EL 3000+250*EL

Harsh 1508+201*EL 1500+200*EL

Hostile 754+151*EL 750+150*EL