Note that I'm coming at this from the perspective of a long time Terraria player, so most of my comparisons and examples will come from that game. Also please pardon the technical language. If anything I write is too convoluted for its purposes, let me know. Something I keep in mind when interacting with procedurally generated elements is how the distribution of variety contributes to a sense of identity and the perception of unique elements. A large world made up of many distinct elements will still feel repetitive and boring if they are completely randombly distributed into a homogeneous mixture. It will not be perceived as unique, and will have only one identity, that of the overall mix. Terraria addresses this issue by breaking up the world into specific biomes, each of which has a unique identity with uniquely themed aesthetics, loot, and enemies. Furthermore the introduction new biomes and monster types is distributed fairly well over the course of a play through, with certain stages of the game focusing on different biomes and whatnot, allowing for each identity to also hold a unique place in time with respect to any given playthrough. Of course, each identity is completely static, which is something that Starbound is working to address. Starbound also does this with the different types of worlds and stars. The moons, ocean worlds, desert worlds, ice worlds, alien worlds, and so on create a lateral diversity, while the progression through different star types via tech create a vertical diversity (especially with respect to the introduction of new world types as you go to different stars, allowing for the game to focus on different worlds over the course of a given playthrough). Each world is also broken up into mini-biomes, which can hold specific enemy and and special loot. The problem is that since the monsters that any given world holds are also procedurally generated, there is potential for the different world types and biomes to lose much of their sense of identity unless the monster types are similarly broken up. If each monster part is flagged to show up in only specific overlapping sets, based on factors including and most notably: planet type star type biome type as well as perhaps: quests the player has active and or has finished (for quest-specific triggers like penguin assassins, or bounty hunters if you have Vash the Typhooned through too many of their settlements) progression within the main quest (for those invader type scenarios) what race if any has colonised the planet (if they might bring invading species with them) and so on Regardless, the point is that if creature body parts are distributed among well-defined overlapping sets that interact well with the diversity and intended sense of identity uniting specific procedurally generated elements, then a more diverse and interesting universe can be created by achieving the intended design purpose of the procedurally generated monster parts in the first place. A couple notes: I really like the idea of the individual monster parts each tied to a specific behaviour and attack type. But if they are also or are at least better tied to the identities of the worlds they populate, then once a suitably large number of monster parts have been created, this should work quite well. The same of course goes for world and biome-specific loot, which currently has relatively little variety due to there being too few objects populating the game world at this stage of development. It will come. There are other factors that should be used to better define identities for different worlds, especially star types, based on their threat level, is not just how much damage each monster does, but how many are spawned, how large they are, and how aggressive they are. It is a gaming convention that harder worlds would have a larger population of hazards, and that might help with how empty they currently feel. Since there are still around as many monster encounters as on earlier level worlds, it doesn't feel much different. Also larger monsters would work to communicate their threat level. I know this is already done with mini bosses. EDIT: This kind of thing is undoubtedly implemented in the game to some degree, but I would like to see more dynamic, or drastic implementation. Something that could also be used to communicate more effectively the identities of planets and biomes through monster parts, in the same vein as the size / spawn frequency argument, would be visual style, which might be easier to get across than attack types and traits / statuses. -- - - - - - - - -- - - - - - - - - - - - - - - - - - - - -- - - - - - - - - - - - - - EDIT2: This thread ties in very well with a post I made on a recent thread about higher order monster behaviour patterns.
I agree with the fact that currently even thought the procedural generation of monster parts tries to create a unique feeling to the various environments but in a larger scale it do seem to become repetitive especially after encountering many variations of the procedural generation after longer explorations and reaching towards the current "end-game". I think that providing special encounters based on a player's progression would may be possible in singleplayer but would be inappropriate in multiplayer so i don't think that is a way to go. I have a thread i created about collecting ideas how certain wholes the game has as stated by the developer's summerize could be filled efficiently here : http://community.playstarbound.com/index.php?threads/re-to-the-developers-return.88543/ and i have an idea about how could be the old temperature system be brought back and be implemented to the current state of the game where i have a planet generation system which provides deeper and more diverse planet natures. I think if the monster's procedural generation could be linked with these more various environments it could encrease the diversity of the monsters as well. However simply forcing the generation to be more distinctive in generation related to specific properties of different worlds would be a more elegant and efficient solution indeed. These ideas would actually create quite a difficult generation algorithm but they might as well provide a more standout result. I would like to create a link to this thread from my thread if you don't mind.
Please do. Anything that can generate more interest / exposure to any positively contributing discussion of the game is a good thing to me.