If you've been reading (or browsing, skimming, or offhandedly glancing at) this blog for a while, you might remember this, in which I posited that rather than continually escalating attack matrices vs. more or less static Armor Classes, we might extract a character or monster's attack bonus from the chart, and then, in a fight, apply the difference between the attack bonuses of the combatants to the attack rolls of the greater. In other words, opponents become easier to hit not based on the attacker's absolute skill, but on the difference in skill between attacker and defender. This means that a fighter with a THAC0 of 3 (an attack bonus of 17) dueling a fighter with a THAC0 of 5 (AB 15) doesn't hit his AC3 plate-and-mail clad opponent on a roll of 0 (i.e. missing only on a natural 1 as per the rules.) Instead, because his attack bonus is two points better than his opponent's, he'll hit on a roll of 14, and his opponent will hit with a 16. (We're using the line of the attack chart for 1st level characters to determine what number is needed to hit each AC.) If the fighter with the attack bonus of 17 is up against a fighter with a bonus of only 5, then he'll add the difference in his favor of +12 to his attack rolls, not simply because he's a high-level fighter in an absolute sense, but because he's a LOT better than his inexperienced opponent and capitalizes on every mistake the other fellow makes in battle.
Since composing that particular ramble, I've decided to drop the "attack bonus" terminology, and refer to this derived stat as Combat Rating, or CR. It also seems like a good idea to modify how ability scores affect combat. This is both to prevent extreme ACs right out of the starting gate (a level 1 halfling with plate, shield, and 18 Dex would be AC-1, and thus nearly unhittable to any opponent of similar level), and to tone down the massive advantages of high Strength (bonuses to hit AND damage make an 18 Strength fighter vastly more effective in melee than one with a 13.) The character's Dexterity adjustment should apply directly to his CR, and Strength adjustments apply only to damage rolls. This means that a Dex adjustment counts for both offense and defense, or more precisely, it affects the balance between the total (offensive and defensive) combat skill of both combatants. The 18 Dex, level 1 halfling has a CR of 4 (1 for his level, +3 for his Dex adjustment) and an AC of 2. The 18 Str, level 1 fighter has a CR of 1. He's raw and ungraceful in combat relative to the halfling, who with his net +3 CR advantage will be zipping in for quick slashes more often than Mongo is going to land a clumsy haymaker on him, but when Mongo DOES hit, he's going to mess the little guy up pretty badly with that +3 bonus to damage.
One interesting implication of this is that natural coordination and grace can to some extent make up for a lack of experience, and conversely, long experience and drilling of combat reflexes can make up for a lack of natural quickness and agility, but pure physical might is entirely its own thing. You just can't learn brute force. (As I used to hear a lot, back when I followed NBA basketball more closely and some team would take a chance on an awkward 7-footer, "You can't teach size.")
One option I've thought about is, instead of automatically applying the difference in CRs to the attack roll of the greater, how about the one with the higher CR gets to choose whether to apply that difference to his attack or to his defense? The more skilled dancer gets to lead, so to speak.
Of course, you still can have actual attack bonuses, whether those comes situationally (+1 for attacking from a position of superior height, say), or a magical bonus such as that of a sword +1. That would make these bonuses especially valuable to the lesser combatant. A CR 2 fighter against a CR 5 fighter who gets a +1 bonus to his CR will find that he can defend himself a little better. It affects the other fellow's hit rolls, not his own. If he gets a direct bonus to his attack roll, he actually stands a better chance of hitting.
CR and Climbing on the Dragon's Back
Another possibly interesting consequence of this combat system is that it makes the system of climbing onto really big monsters, first formulated by Scrap Princess and modified by Zak S., into a very valuable tactical option. I don't think it's too uncommon for parties of 4th-6th level characters to take on a red dragon, but using the CR system, none of them are going to have an advantage over teh dragon - in other words, no bonuses to hit vs. its AC-1. The dragon will have a little tougher time hitting the fighters than it otherwise would, but they're going to have a devil of a time dishing out any damage on it. What to do? Use those rules for climbing aboard and getting close to the vulnerable spots! (Zak's adaptation of the idea seems to fit a lot better with my underlying system, IMO.)
Weapons and CR
A while back, something I read on Charles's Spells and Steel blog got me thinking about the relative merits of different weapons. One of the points of the post was how bizarre it is that, in D&D, it's no more difficult to hit a trained fighter holding a sword and a peasant with nothing but his fists if both are similarly armored. Same goes for one fighter with a sword, and another with a dagger. In a real fight, it's quite as possible to kill a man in a single hit with a dagger as with a sword, but it's going to be tougher, all else being equal, for the dagger wielder to land a hit against the swordsman. Damage potential itself isn't so much an issue; it's the fact that the swordsman has a great advantage in reach, and he's got not only a sharp point but two or three feet of blade that has a chance to catch his opponent and deal some damage. The dagger wielder has to get in a lot closer, and to get a really good lick in, he must thrust with the weapon's point; slashing is far less effective. In other words, all else equal, bet on the swordsman.
Well now, what if, instead of variable weapon damage, different weapons added different bonuses to the wielder's CR, based on reach, ease of use, and defensive capabilities? A barehanded character fights with only his base CR, i.e. with a weapon bonus of zero. A dagger might add +1, a sword might grant +4. A two-handed greatsword might grant +5. Short swords, with less reach, might give CR +3. Maces and hammers, being heavier and less elegantly balanced, might be +2 to CR overall, but grant +2 to hit against medium and heavy armor, representing their purpose of defeating those armors with impact damage. (These numbers have been pulled out of the air with relatively little consideration. Perhaps a wider or narrower range might be more appropriate, but I haven't fully math-geeked out the details yet, much less consulted with folks more knowledgeable than I about the merits and demerits of medieval weaponry.)
Let's say a hypothetical rebellious peasant has gotten his hands on a sword, and caught a mercenary of the corrupt town alderman unarmed. Both are unarmored (AC9.) Under ordinary D&D rules, the 1st level mercenary fighter actually has a slightly better chance to land a blow against our sword-wielding peasant rebel (attack rolls for 1st level classed characters are 1 point better than those for Normal Humans.) Mercenary needs a 10 to hit AC9, peasant needs an 11.
Now let's try the CR system with weapon adjustments. Mercenary has a base CR 1, while the peasant's base CR is 0. Now put the sword in the peasant's hands, and add its +4 CR bonus. Peasant now has a total CR of 4 (0+4) and a net bonus of +3 over the unarmed mercenary, and the upper hand in combat. He can apply his bonus to attack, meaning he'll hit the mercenary on an 8 or better, or he can apply it defensively, and force the mercenary to roll a 13 or better to hit him. (But of course, not both at once.)
CR and Monsters
For monsters, the correlation of Hit Dice to escalating attack rolls in standard D&D wouldn't have to be modified at all. If a creature's Hit Dice are a function of its size, like a bear, dragon, or giant, then the corresponding CR represents an advantage in reach. If the creature's HD are more attributable to superlative skill at dodging or supernatural resistance, like a cockatrice or a wraith, there's no good reason we can't apply that same rationale to its ability to fight, too. Monsters who commonly use weapons can benefit from the standard CR bonus for weapons.
And finally, there's the issue of grappling, and of animals and other monsters with natural attacks mowing over weapon-wielding adventurers. As has been noted on another blog (which is utterly escaping me right now, so if it's yours or you know whose it is, please drop me a comment so I can give due credit!), it's pretty tough to bring a sword to bear on a wild boar that's goring you or a mountain lion that's jumped from a crag to pin you to the ground. But suppose that once you're grappled or successfully hit by an enemy in brawling range (i.e. occupying the same space on the combat grid, if applicable), only small weapons still grant a bonus to CR. Medium and large one-handed weapons convert their normal CR bonus to an equal penalty, and two-handed weapons are just unusable under those circumstances. A dagger is a lot more handy than a battle axe when a wolf has pounced on you and is going for your jugular. I'm thinking this just might be enough to counteract the advantage that creatures using weapons, and receiving a CR bonus for it, would seem to have over monsters that use natural weapons and get no CR bonus for them. The initial advantage will belong to the weapon-wielder, but once that wolf is tussling with you hand-to-paw, the advantage of holding a sword is less than nil, and you'd best let go of the sword and focus your efforts on throwing the wolf off, or stabbing it to death with your dagger.
I'm not at all sure that I'll actually use any of this in play, but it's fun to speculate and play with the numbers.