Jtdcftul

As far as I know, that kind of list is only found in Star Trek EU works, specifically the Star Charts, which shows a planetary classification from Class A to Class Y with some useful information.

It might be too coarse for you needs (as in "not detailed enough"), but you could use the Universal Planetary Profile from the RPG Traveller. (In the modern version of the game, it's called the Universal World Profile and is far more comprehensive than in the original game that I'm most familiar with.)

The format describes:

Starport (Sp)

Planetary Size (S)

Atmosphere (A)

Hydrosphere {H)

Population (P)

Government (G)

Law Level (L)

Tech Level (TL)

Check the wiki for planet types.

There are eight types by size, seventeen by orbital regime, seventeen by composition and three miscellaneous types. That does for 6936 possible combinations, so if you wish for one letter descriptors you may need to use chinese characters.

Otherwise, you can use any fictional classifications such as Star Trek's, as long as you don't do it commercially I guess.

I'd personally as a reader want to be able to differentiate between planets easier than that, and an easier way to do that might be to use a taxonomic system for classification.

Instead of having so much variation like animals, you can have a class A-B-C planet meaning different items, with further breakdowns in classification beneath that. Just like using the word "plant" versus "flower", we can use these indicators to provide ever increasing levels of specificity. It doesn't need to be anything completely complicated, but provide the full 5 bits, putting emphasis on certain pieces over others based on how important someone would use them in conversation.

First letter determines size and type.

• A: Gas planet, strong gravitational pull, tidally locked


• B: Gas planet, normal gravitational pull, tidally locked


• C: Gas planet, weak gravitational pull, tidally locked


• D: Gas planet, strong gravitational pull, tidally unlocked


• E: Gas planet, normal gravitational pull, tidally unlocked


• F: Gas planet, weak gravitational pull, tidally unlocked


• G: Rocky planet, strong gravitational pull, tidally locked


• H: Rocky planet, normal gravitational pull, tidally locked


• I: Rocky planet, weak gravitational pull, tidally locked


• J: Rocky planet, strong gravitational pull, tidally unlocked


• K: Rocky planet, normal gravitational pull, tidally unlocked


• L: Rocky planet, weak gravitational pull, tidally unlocked


• etc.

Then further elaborate into atmosphere type and hostility of location.

• A: Weak atmosphere, close to star, hot


• B: Weak atmosphere, close to star, temperate


• C: Weak atmosphere, close to star, cold


• D: Normal atmosphere, close to star, hot


• E: Normal atmosphere, close to star, temperate


• F: Normal atmosphere, close to star, cold


• G: Strong atmosphere, close to star, hot


• H: Strong atmosphere, close to star, temperate


• I: Strong atmosphere, close to star, cold


• J: Weak atmosphere, Goldilock's Zone, hot


• etc.

After that, you can get into habitability, resource availability, breathing possibility, and population with just numbers.

• 01: Not habitable, scant resources, breathing apparatus necessary, not populated


• 02: Habitable, scant resources, breathing apparatus necessary, not populated


• 03: Not habitable, normal resources, breathing apparatus necessary, not populated


• 04: Habitable, normal resources, breathing apparatus necessary, not populated


• 05: Not habitable, heavy resources, breathing apparatus necessary, not populated


• 06: Habitable, heavy resources, breathing apparatus necessary, not populated


• 07: Not habitable, scant resources, breathing apparatus not necessary, not populated


• 08: Habitable, scant resources, breathing apparatus not necessary, not populated


• etc.

Forming from this, you could classify any type of planet, and instantly know what sort of place you're visiting. For instance, Earth might be classified as a K class planet, but with further information on it bearing more fruit.

Earth: K-P17 planet

Population: 7.2b

Crust Compostion¹:

- oxygen, 46.6%

- silicon, 27.7%

- aluminum, 8.1%

- iron, 5.0%

- calcium, 3.6%

- sodium, 2.8%

- potassium, 2.6%

- magnesium, 2.1%

Intelligent Species Types: Humans, Dolphins, Mice

Prevalent Languages:

- Mandarin Chinese (1.1 billion speakers)

- English (983 million speakers)

- Hindustani (544 million speakers)

With this, you have a complete knowledge of a planet, and can further fill in more information on a longer format, but can refer to Earth as a K class planet, then further specify with more taxonomic indicators.

¹ - "Essentials of Geology" (7th Ed., Prentice Hall, 2000) by Frederick K. Lutgens and Edward J. Tarbuck

I don't think there is a good way to simply classify planets with the granularity you want without creating a barrier between you the creator and the audience you are presenting to.

For your system to work, the user needs to know and remember it. This is hard. You can't just present it in a table (I'm assuming your users are both gamers and readers) in the appendix or index because a user needs to keep flipping back to it. This breaks the flow of your story or game (sort of like tutorial prompts) and creates a disconnect for casual or first time users who might not want to invest in learning everything and just dive straight in (this is one of the reasons why casual and mobile games are such a huge market compared to knowledge intensive RTS games).

A very basic system would be something along the lines of:

• Planet/Asteroid/Star


• Livable/Unlivable

Which can give you the following combinations:

• Livable Planet


• Livable Asteroid


• Livable Star

Which you could then sort into a code that you want e.g. (LP, LA, LS).

But a casual user doesn't understand this, especially if you want to factor in all possible factors of a planet. The only thing you are doing with such a code is making it harder for a user to understand your game.

If you really want to... Unicode has 137,374 mappings which means you could, if you really really wanted to, present 137,374 unique combinations to the reader with a single character (some of them look exactly the same, but they are different). But at that point you might as well be using Webdings.

Weighted k-medoids clustering

k-medoids clustering is a method of dividing a set of points into clusters of similar points. You can use this to divide your worlds into clusters, and then give each cluster a label.

One algorithm to do this is the PAM algorithm:

1. Give each world a weight (representing how likely it is to come up in conversation). Also define a "dissimilarity" function that measures how dissimilar two worlds are. This should make the set of worlds a metric space.


2. Give each label a "metroid" world. This can either be done arbitrarily, or meaningfully (the meaning of the labels will be loosely based on these initial metroid worlds). The metroid world does not need to have a high weight; it could even have weight $0$.


3. Assign each world the label whose metroid is most similar to it.


4. Calculate the weighted average cost of the labelling, which is defined as the weighted average dissimilarity between a world and the world's label's metroid.


5. For each label $l$ and each non-metroid world $w$, change $l$'s metroid to $w$. Then do steps 3 and 4 again. If the weighted average cost went up, undo the change.


6. Keep doing step 5 until the weighted average cost stops decreasing.


7. The current labelling is the output of this algorithm. You may want to rename the labels if you want them to have meaningful names, as the algorithm will probably have relabelled worlds multiple times until it found a local minima.