Animals in combat

 D&D combat is weird when it come to animals (and animal-like fantastic creatures like griffons and owlbears.) Actually, it's weird in several ways.

First is the breaking of combat abstraction in creatures with multiple attacks. It's pretty well established by now that, for a character, one attack roll does not equal one swing of a sword. It represents the cumulative effects of ten seconds (or a full minute, in the case of AD&D) of thrusts, jabs, feints, parries, and such. Even the damage rolled need not come from a single wound; it could be several small wounds. For a while I struggled with the idea of a single creature making multiple attack rolls per round, but I've come to the conclusion that multiple attack rolls, in and of themselves, don't violate abstraction. It's just a particular mathematical relationship between offense and defense and damage, which need not imply a particular number of specific strikes any more than a single roll would. So a bear with three attacks, being able to concentrate its fury on a single target or spread it around to three separate opponents is no big deal. I'm totally cool with the idea of a bear being able to put the hurt on more than one man-sized creature in a ten-second round, or to really lay into a single one. 

But then... then! D&D goes and explicitly designates those three attacks claw/claw/bite, and gives them different damage ranges, strongly implying that each of the bear's attack rolls represents a discrete use of one claw, the other claw, and its teeth, respectively. A similar conceptual problem arises with animals like elephants, rhinos, and dinosaurs, which are listed with an either/or choice of two different attacks, typically trampling and a horn or tusk attack. In reality, such a creature is not going to choose one or the other, but simply attack, inflicting a melange of damage from all its possible bodily weapons.

Next, there's the weird issue of scale in damage between different creatures. Consider, for example, the 3 HD giant crab (Cook Expert rulebook), which has two attacks ("2 pincers") doing 2d6 damage each, for a potential maximum of 24 points of damage. Contrast that with the mighty grizzly bear (Moldvay Basic rulebook,) a 5 HD monster with its three attacks (claw/claw/bite) doing 1d4/1d4/1d8 damage, for a potential maximum of 16 points. A larger (in terms of HD, at least) creature, famous for its brute strength and sharp claws and teeth, tops out at 2/3 of the damage potential of a pumped-up crustacean whose pincers are suited for grasping carrion rather than slicing and rending prey. (It's also a bit weird that the grizzly -- let's remind you that it is a notoriously strong creature with massive paws tipped with huge, hooked claws -- does with its claw attack the same damage as any average human with a dagger.)

I could go on. A 1+2 HD giant fire beetle, described as being 2 1/2 feet long, does the exact same damage (2d4) as a full-grown 6 HD rhinoceros. A 9 HD elephant, attacking with its tusks (somehow using them as separate attacks!) does 2d4 damage per attack, and even a 15 HD woolly mammoth does 2d6 damage with each of its tusks -- only on par with our giant scuttling beach janitors above! For a more fantastic example, compare the owlbear (generally depicted with a more eagle-like than owl-like beak) to the griffon and hippogriff, also described as having eagle heads; the 5 HD owlbear does 1d8, the 7 HD griffon does 2d8, and the 3 HD hippogriff does 1d10 -- a huge amount of variation! I think you get the picture by now that D&D's designers weren't overly concerned with consistency.

The final oddity of animals in D&D is that the standard initiative, attack roll, damage roll, rinse-and-repeat-as-needed sequence doesn't really model what real world animals actually do very well. A wolf doesn't just chomp and let go, chomp and let go. It will try to get a solid hold on its prey, on its throat if possible, shaking and twisting while trying to force it down to the ground with the assistance of its pack mates. A crocodile also doesn't just bite and release. It doesn't do damage primarily with its teeth, but by getting a grip on its prey with its teeth and then dragging it into the water for a "death roll," which would inflict terrible twisting on the prey while it drowns. 

What's a DM to do? You could shrug it off as just a game, and I can't say you'd be wrong to do so. If you're like me, though, you'd want at least to consider possibilities for increasing the verisimilitude of animals and their fantastic cousins.

The first problem could be solved simply by assigning a single damage range to all of a creature's attack rolls. The description may note what bodily weapons the animal possesses, but there would be no attempt to divide damage discretely among them.

The second problem requires a little rebalancing. It's likely that these creatures were designed and statted up by people paying little attention to what the others were doing, but we can look at them with an eye toward making apples-to-apples comparisons and determining what their strength and lethality should be relative to one another. This is probably the issue with the most important impact on actual play, as it's tougher for players to assess the threat of a creature when there's no reasonable proportionality between one and another. 

The final issue is the thorniest. A robust but simple and streamlined grappling rule would be most welcome, since what wolves and crocodiles do is essentially armed combat and grappling rolled into a single action. Interestingly, the rock python (listed under Snake, Giant in the Moldvay Basic rulebook) offers one possible fix: if the snakes bite attack succeeds, it automatically coils around its prey and constricts. The tactics of wolves or crocodiles could be modeled on the same basis, with perhaps penalties to counterattacks and a chance to escape each round by inflicting enough damage or something. 

That's my take on animal combat in D&D and where it goes wonky. Do these things bug you as much as they do me, or have you made your peace with them, or did they never trouble you in the first place? Do you have any thoughts on how to model things better without adding a lot of extra complexity? I'd love to hear your suggestions!