Mob Taxonomy

[thyrr]

You're right about generalizing the code, but I don't think it should be done entirely through attributes. If you had 30 different things that could happen when you eat something, and each could be modified in a dozen ways, you're looking at a lot of code, possibly all stuffed into the on_eat() function. And maybe "exploding" is extremely uncommon; perhaps only loompa-oompas explode in that particular fashion. I don't want to add a body-part-blow-distance=4 and smellslike=duck attribute just to differentiate between normal explosions and loompa-oompa explosions, and an I-am-a-loompa-oompa check in on_eat() is plain ugly and unmaintainable. What's wrong with custom code on the relevant class?

So ideally, if I made smorfs grow magical warts that produce confusion-inducing pus (useable in potions) after eating poisonberries, the code would be localized to smorfs and automatically inherited (or not, according to the rules) by smorf descendents. Maybe there could even be a generic illness-that-produces-magical-body-fluids function that looks at the attributes, and would be attached to the smorf class and the one or two other classes with this odd trait. But it'd be a little silly to have on_eat check for say, "poisonberries:poisonous:warts:confusion:can-be-harvested" and call the code there, IMO. Or even to have a table of effects+callbacks outside the on_eat function.

Now yes, for the numerous common cases, it makes sense to have a centralized table of effects. But for something like sunlight-damage, which pretty much only affects vampires and vampire-lings, it just seems to harken back to the MERC days (or maybe ROM) of flags, flags, and more flags for every custom behavior. Or a table of spec_progs. And there has to be a better way.

[KaVir]

[thyrr]

[KaVir]

[Kjartan]

I suspect that the lion's share of the effort of implementing generic mix-and-match mobs is independent of whether it's done via inheritance or not: you have to figure out how to mix each trait, and what has precedence over what. KaVir's color-mixing example illustrates this: he had to think up a fairly complicated system just to cover one trait. Imagine the effort in coming up with a mixing system for every single possible characteristic of mobs! And yet, to actually mix mobs, this is what would be needed.

I think maybe there's not much call for it, because while there's a very large number of base mobs you might want to mix things in to, the number of things to mix in is much smaller. Nobody wants a half-wolf-half-bull; it's just not a compelling mob, it's too random. (Although it would be called the "bulf".) One of the pieces needs to be something with a lot of social weight, like a human or a dragon or a demon. I like the D20 approach of templates: have a zillion base mobs, and some templates for mixing in "half" or "a little bit" of various important things like demon, angel, dragon. Then all you really have to design for the system to work is the templates.

I think that vampire and werewolf are not best treated as mixtures, they would be much better handled as something like D20 templates. You can't really have a pure vampire that isn't part something else; "vampiric" isn't associated with a complete set of traits, for example there is no characteristic number of eyes. (But I guess you could use KaVir's intensity system and just put the intensity of its number of eyes at zero, and add the requirement that any legal mixture must have some nonzero intensity for every trait from one or the other piece).

[thyrr]

[Vopisk]

[thyrr]

Well, this topic *is* called Mob Taxonomy, which I suppose by definition of "taxonomy" should be about using attributes/flags to describe traits. So we're probably slightly off-topic anyway.

I use the term "flags" in a broad sense: if you have an attribute like damageFromSunlight=20, you're describing the behavior in a rigid numerical way, and somewhere you're going to have to write a check that says if(npc.damageFromSunlight) etc. Kavir's system is not particularly object oriented (whether that's good or bad is up for debate.)

So you'd have something like

[Massaria]

Spurred on by this and other threads, I’ve been considering how to mix together a system of mutating mobs/characters and a combat system that has strategic elements while still being automated (no action needed during).

It’s far more simplitic than what I had in mind to start with, but I’d like to get this on the thread while rigor mortis is still present. As I think, and write, about this, I’m seeing a strategy game, whether graphic or not – but not really close to any mud I’ve seen. Still I believe there might be something in it for a mud’ish system.

There would be a number of basic animal types, templates if you will, and these would form the basis of any player’s “breeding program” (the players, naturally, are dreadful aliens who breed creatures for sport or war).
This list of animals probably shouldn’t be very long, perhaps no longer than:

- Primate
- Quardroped

Horse
Lion
Goat
…

- Mollusk
- Crustacea
- Insecta
- Aves
- Reptilia

Now all animals (in this system), have at least two parts, but usually three: Head, Body and Tail. The player starts by choosing two animals to form the stem of his pedigree. Half of it, anyway, as the result of those two could be coupled with the result of two other basic forms, or a pure form, or perhaps even itself (in which case the alien sports manager has decided that this particular animal is fit for mass-production. Just in time for the first batch to mature before the next game. Whew).

The basic forms would need information such as:

- Name
- Head (type)
- # Body segments / # supporting segments
- Limbs on each segment (type)
- Tail (type)

Thus, a horse would read:

- Horse
- Grazer
- 2/2
- 2 * Legs = 4
- Flyswatter

With each trait that might carry over to a new animal, there would be a ‘dominance rating’, which would indicate the likelyhood of that trait being passed on (when matched against the rating of the other parent, if that parent has the trait at all)

The types of head (predator, grazer, nibler, piercing, …), would have different biting-capability, perhaps affect a diet (thereby maintenance), but otherwise have little impact. Their one big point might be that biting is a very powerful, piercing attack on organs which most other natural attacks cannot hope to equal. See below on types of damage.

Tails would probably play an even lesser part, with most results of the ‘flyswatter’ type, but a scorpion’s tail might be interesting, and tail-feathers a necesity for decent flight-capability.

Then, if two animals were to cross, they would simply produce a head and a tail based on the dominance rating of that trait in the parents. The segmented body (two limbs to each segment. A human has two segments, four limbs – so does a horse or bird), would also produce results based on dominance (I’m thinking that pairing up the limbs will make life easier, so no left lion-paw and right chicken-wing), but if we introduce a couple of variations we might have the groundwork for a strategic set-up. We might say that sometimes, the segment is left completely limb-less, that both pair of limbs from the parents are present in the same segment (one pair growing from the back, like an angel (wings+arms)), or we might <em>add</em> a new segment. Thus we’d grow centaurs, driders and basillisks (in some traditions).
The addition of a new segment should be a rare and prized event, as it will obviously allow for more attacks, or a more secure footing, during fights. To keep things from growing too fantastic, there would be a rule to the effect of needing at least half your segments for support, if the remaining half should remain at full capacity – imbalances in this regard might lead to penalties and immobilization.

The players would probably end up with creatures that are pretty hopeless more frequently than not, but that’s alright, because that would be a major part of it all: To weed out the undesired and culture the powerful.

The combat that I’ve considered is quite unsubstantial, and, I believe, better fitted for a mud than the large-scale battles of a war-game.
In a one-on-one fight, I imagine two L-shapes (back-to-back: _| |_ ) of arbitrary length in both height and width, but always proportional: The height of the “L” never longer than the width. The width of the “L” is the supporting segments of the animal, and the height represents the segments that are free to do combat.
Each pair of limbs would attack once (possibly parry once too), at the limb directly in front of it (default) or perhaps out of criteria set before the fight (e.g. attack pinchers before paws). Limbs would have reach too; a simple 0-3 number which represents the number of segments that that limb can reach (a limb in segment 3 with a reach of 2 could attack segments 1 through 5 on a foe of equal length (segment 1 = first after head, thus segment 3 would be the hind legs of a centaur).
Limbs could be associated with additional effects upon successful attacks, like pinning a limb (squid tentacle), poison/irritant, continual (most bites), etc.
If limbs grew in dominance over time (a creature who’s ancestors has had the scorpion’s tail for 10 generations, would be more likely to impart that characteristic to the offspring than a tail with a 5th generation history), some creatures will naturally become more disirable and might earn the owner large sums for breeding or selling it.
In the plans for my mud, I’m trying to make a destinction between loss of blood and fatigue on one side, and tissue damage on the other (flayed muscle, broken bone, pierced vital organs), but a simple implementation of Hit Points would probably be to prefer here, allowing severance and incapacitation of limbs, possibly whole segments (collapsing any segments resting on that segment, forcing the foe to ‘come down a notch’ and incurring manoeuvring penalties).

As vague as this sounds, it can get worse. I’m still missing a good way of doing size, and thereby speed (I’m sure I’m missing a lot of other things too, but these worry me the most). While having a measure of limb size would be nice, it’s a distant feature even if a proper system for total mass is done. I’d rather like if there’d be something to the idea of breeding a horde of smaller creatures at a low individual cost, and have them effectively gang up on larger types of animals. That would add invaluably variation to the strategies at hand, but it’s so damn hard to account for the relative mass between two combatants.

I’d like to divide damage into the three standard categories (Slash/Piercing/Bludgeoning). Not because I want to make armor/weapon penetration values (but that would be nice too), but moreso because they represent different kinds of damage to an organism.
Of course, by doing this, I’m moving away from the standard HP system, where only one factor determines if a limb or creature can function (how many HPs it’s got left).
As hinted above, there’d be a distinction between the structural integrity of a limb/creature, and the ‘vitality’ or energy left in the organism. Hit Points, therefore, could be seen as an arbitrary measure of the blood or energy left in the organism, while Structural Integrity Points would represent the state of the foundation through which this energy flows.
The aim, naturally, is to have the three types of damage do different amounts of damage against HP and SIP.
While HP would be more of an over-all measure of the creature’s state, SIP would be used to measure if individual limbs are working – allowing us to penalize the cripple justly, to the extend of collapsing the structure and life functions completely.
Piercing damage would be a fairly rare occurrence, limited to bites only (I can’t think of any animal who actively uses a piercing technique for attack or defense (discounting passive defences, such a the spikes of a porchupine (and bites, obviously)), even horns seem to be used as a butting force rather than to penetrate flesh). It would cause significant organ/HP damage, and may occasionally hit a bone (SIP-dam), but usually won’t.
Slashing damage would be much more normal, inflicting a roughly even amount of damage to HP and SIP.
Bludgeoning force seems prevalent in the natural world, and inflicts a lot of structural damage, but has little effect on energy levels.

Hm. I think that’s most of it.
Sorry if it came out all jumbled up, but it’s my two cents, something to choke on

Mass.

PS: Found your post very informative thyrr, but it's a bit too techy for me to have anything to add.