Type rebuild method: poll

[Vebyast]

Back in 2006, in The Tome of Necromancy, Frank and K modified the Undead type to make more sense by breaking it down into several subtypes. In doing so, they made the following claim:

The obvious, and slickest, way to handle the excesses of the Undead type would be to simply rewrite the Undead type with a lot less in it and throw down a number of subtypes (mindless for skeletons, amorphous for shadows, and ponderous for zombies) to put in the abilities that each type of undead is supposed to have. But polls have shown that people aren't willing to play with optional rules that do that – but perversely they are willing to add new subtypes to monsters to remove rules instituted by the base template. I don't know why, but DMs are honestly more likely to use an additional subtype that removes an inappropriate game effect from a monster than they are to use a modified base type that doesn't have the inappropriate effect in the first place. So that's how we're going to do it here.

I'm currently doing something pretty similar to Constructs, but it's been 4 years and I want to do the same research again.

Benefits of subtypes subtracting properties from a full-blown type:

• Don't have to modify the base type

Benefits of subtypes adding properties to a basic type:

• Makes much more sense
• Much, much easier to do the cyborg/bicentennial man thing

[CatharzGodfoot]

"Undead" should be a subtype rather than having Undead (augmented human), Undead (augmented animal), Undead (augmented watermelon), etc.

Also, body plans should be separated as much as possible from other things. So humans, elves, giants, goblins, trolls, angels, and zombies are all "humanoids"; mice, some dragons, buffalo, lizards, cats, catoblepases, and badgers are all "beasts"; whales, fish, anacondas, some dragons, and eels are all "serpents"; bats, birds, wyverns, and pterodactyls are all "avians"; oozes, jellies, and swarms are all "amorphous"; and really strange things are "aberrations".

A red dragon would be a Beast (dragon, fire, reptilian). A red dracolich would be a beast (dragon, fire, reptilian, undead). A zombie red dragon would be a beast (dragon, fire, reptilian, mindless, undead).

This would help in adjudicating stuff like called shots, swords of sharpness, and equipment slots.

[BearsAreBrown]

The benefit of subtypes subtracting abilities is simplicity. I realize this sounds reverse but you need to remember that 'ability subtracting' subtypes are rare. All constructs work like constructs, except Living ones. This is easy to remember. All undead are X, but skeletons are XY and vampires are XZ and shadows are XC but Wights are XZC is complex. But it makes a helluva lot more sense.

Also, Catharz's system is great. You could take it a step further and remove Type and Subtype all together and just hand out 'tags' for a creature. Red Dragons wouldn't be Dragon(Fire), or Beast (dragon, fire, reptilian), they would be "dragon, quadrapedal, spellcaster, fire, reptilian"

[Username17]

4e's typing system is actually better. The "types" are things like "Soldier" and "Lurker" that actually determine combat roles, and everything about whether it's a living being or from the outer planes or whatever is all just a series of subtypes. Now, 4e fucking sucks and they didn't let their subtypes do nearly enough (I don't think it actually bothers to tell you anywhere whether monsters have to eat or sleep because the monsters aren't supposed to "exist" outside the encounters you fight them in), but the basic setup is superior.

The question is thus not what what would be best while making a new system, it's how to integrate a change in the rules to 3.5 so that you can actually still use the monster manuals.

-Username17

[Vebyast]

The benefit of subtypes subtracting abilities is simplicity. I realize this sounds reverse but you need to remember that 'ability subtracting' subtypes are rare. All constructs work like constructs, except Living ones. This is easy to remember. All undead are X, but skeletons are XY and vampires are XZ and shadows are XC but Wights are XZC is complex. But it makes a helluva lot more sense.

The question is thus not what what would be best while making a new system, it's how to integrate a change in the rules to 3.5 so that you can actually still use the monster manuals.

I'm currently concerned with fixing one thing and exactly one thing: types that have too much stuff in them and need to be broken down. More specifically, I'm thinking about rewriting Constructs, and that type has so much stuff in it that it's going to break into (at my analysis) four subtypes.

Ability-subtracting subtypes may be relatively rare among undead, but they would be incredibly common among Constructs. Anything with distinct anatomy (I'm looking at you, Inevitables) should not be immune to critical hits. Lots of things should "wind down" after a while (clockworks), and so they need con scores, they need to be vulnerable to fatigue and ability damage and so on. Lots of things have minds (nimblewrights, inevidrons), and so are vulnerable to mind-affecting. Lots of things have souls (warforged, robot girls), and so can be rezzed and are vulnerable to necromantic effects. Even more, a bunch of constructs have multiple of these: Nimblewrights wind down, they have brains, and they're vulnerable to crits. Inevidrons have brains and they're vulnerable to crits. Warforged and Robot Girls are constructs in nothing but name.

Additive subtypes take one less step to figure out. With subtractive subtypes, the process for determining vulnerabilty goes like this:
[*]vulnerable to mind affecting -> but wait! Undead type -> immune to mind affecting -> but wait! Dark Minded -> vulnerable to mind-affecting.
Not only is that's bass-ackwards relative to the rest of DND, it's two steps to think about. With additive subtypes, it looks more like this:
[*]vulnerable to mind affecting -> but wait! mindless -> immune to mind-affecting.
Needless to say, that's much, much easier.

4e's typing system is actually better.

I agree with Frank here. I've only played one 4e game, but I did like having a monster's combat role and basic properties described right in its declaration.




BTW: Koumei, if you're reading this, can I stick the Robot Girl and associated stuff in the Construct Operator's Manual (if it ever progresses beyond the concept stage, that is)?

[Username17]

If you were going to use Constitution at all, about 100% of Constructs should have it. But writing in a Con score for every Construct is a pain in the ass. First because you have to go through all of them and do it, but secondly because when actually playing the game you'd have to go through the revised list and recalc everything based on a lookup table. It's frustrating.

In almost every case however, you could fix the Construct hitpoint problem with a blanket rule: Constructs add their Strength Bonus to hit points each hit die. That change severely mitigates the weird bullshit tiny hit points that supposedly high end Constructs end up with in 3.5.

-Username17

[Vebyast]

A bunch of constructs are going to need constitution scores again. Not having a con score is fine if you're immune to everything that requires a roll against con, but if you can be forced to make checks for exhaustion or starvation you need to have a con score to use on those checks.

There are definitely some constructs that should be immune to everything that even remotely interacts with constitution; inevidrons, animated objects, and golems, to start with. Those things should basically be unstoppable forces of nature. I can see your point about constitution scores making hit points easier to deal with, though. Are there any disadvantages to giving every construct a con score and only making them immune to things that check against it?

As for difficulties figuring out con scores, the obvious answer is to make the con score dependent entirely on the stats already given for the construct. If the algorithm is simple enough that you can do it in your head in a few seconds, you can reconstruct a construct's constitution score just by looking at it. A few options I was considering:

• Copy the Str score into the box for the Con score. This indirectly does the same thing you're suggesting with strmod to HP.
• Use a really simple table based on size and HD. Big, high-level constructs have big con scores, smaller and lower-level constructs have smaller con scores.
• Start at 10 and give a scaling con bonus for having more HD or more levels in [Construct] classes.