So in D&D-derivatives combat generally involves two rolls: an attack roll and a damage roll. The result of an attack roll is a simple success, failure, or a critical hit (which doubles the damage dealt). The result of the attack roll otherwise has no affect on the damage roll.

Why not remove the damage roll and have the damage be determined by the attack roll? Are there any popular optional or house rules to this effect?

Why not remove the damage roll and have the damage be determined by the attack roll?

In other games, the most common approach is to have the margin of success be the damage. You need 16 to hit, you roll 17, you do 1 damage. But then you get into, well what about the different weapons? Surely a margin of 1 with a greatsword and a margin of 1 with a dagger should give different results? So then you have to fiddle with the weapons.

I'm sure I have many examples of this sort of design on my shelves, its not a new idea, but it can prove surprisingly tricky to implement.

Heavy Gear, all the Dream Pod 9 games really, use the Margin of Success as a multiplier to the damage, which can produce some absurd looking numbers at the high end (being able to shoot holes in tanks...), and it sort of contributed to their major problem of having Dexterity being the be-all, end-all of combat stats.

The White Wolf storyteller system does this sort of thing as well, at least in the nWoD... not referencing the Dream Pod 9 but the OP, anyway. Again, you see an issue where the most deadly weapons are the most accurate, a .22 caliber pea shooter being far more dangerous than an elephant gun (this is not a made up example, btw), because successes from the dice pool translate directly, and accuracy increases those successes.

I was about to post the same thing, but I was thinking oWoD. MOS adding to damage makes Dex the Uber stat. I think mongTrav does it. Fudge does it.

For d20 or OSR you could use average of the damage die (round up) plus Str mod for quicker combats. Or max damage die for really quick combats.

As others have said, using degrees of success as damage makes it more difficult to simulate attacks from things like giants that have a low chance to hit but do a ton of damage of they do hit. But I've also found that many players just aren't that good at subtraction in their head (you need a 8 and roll a 14, what is the difference?) that it takes longer to calculate the degree of success than to just make another die roll.

A compromise would be to give attacks a fixed damage and have a simple trigger to increase that. So a sword might do 4 damage and do double damage if you roll 10 more than you needed.

Hey man, great idea for a discussion.

One attack roll that also determines damage... I have done this in AGE (Dragon Age, Fantasy Age, Modern Age) games with great success. However, despite being a d20 derivative, the AGE system is designed with enough changes in mind that make the implementation of these rules pretty easy.

So, here's how it works in AGE: rolls are 3d6 (not 1d20) + ability vs. Target Number (DC). Armor is damage reduction. One of the d6s is of a different color and the result determines how well you performed whatever check you tried. For example, you roll STR to jump a chasm and you are successful; if the different d6 is 1, it means you almost didn't make it and might be holding for your life; a 2 can mean you fell on the other side, safe, but will have to take a move to get up; a 3-4 is average, you did it ok; a 5-6 means you cleared the distance well and might get a +1 or +2 on your follow-up check to attack the dude on the other side, etc.

In combat, AGE has two rolls (attack and damage) as DnD. I always though that was stupid because... well, they have an amazing system to solve everything with one roll! So, I took the weapon damage (dagger d6+1, sword 2d6, two-handed axe 3d6) and converted to a bonus: dagger +5, sword +7, two-handed axe +10. Add your STR (AGE uses only the modifiers, not derived stats). When you attack, if you hit, add the damage bonus to that different d6 that determines how well you did and that's it! You made one roll and got the damage.

Now, in d20, admitting that you don't want to roll damage dice together with the d20 (I know, it doesn't solve the problem) my (very rough) idea is giving weapons a fixed damage (a dagger causes STR mod +2, a sword STR mod +4, etc. I'd say place it at half the die damage, rounding down as usual). When you hit, add +1 damage for every 2 points above the AC. So, my fighter (+6 attack and d8+4 damage) now deals 8 damage on each hit. If I hit the exact AC or around it, I deal average damage; on higher rolls, I deal much more damage. Hitting with a 18 against a poor AC 10 wizard would cause 12 damage; hitting with a 18 against a plate armored opponent would cause damage 8. A critical hit calculates damage as the total result was 10 points above; so, a critical hit against that wizard would cause 21 damage (26 attack, increased to 36 - AC 10 = 26/2 = 13+8 = 21).

This idea is untested and might benefit from some variants, but it is there. Hope there's something you can use. Happy gaming!

Since you mentioned WFRP 4e already, to continue with that, size matters much more in that game. Big creatures are not better at hitting, but it is much harder to parry (but not dodge) their blows and those blows do more damage.

Indeed. I just pointed out that you can simulate rare, but powerful hits with success level as long as there are other modifiers.

Rolling dice is fun! Also, some of the most suspenseful moments in my gaming career have been waiting to see how much damage was rolled against the big baddie. "Margin of success" sort of eliminates some of that.

I've never seen multiplication work, for the reason you describe.

I've made it work for my own generic system homebrew, but that has the advantage that I stepped away from the "hit point total" model almost completely, where the multiplied totals are compared to set Wound Level thresholds instead of subtracted from a running total. If the goal is to preserve the feel of most D20 games that's less of an option.

Instead of x1, x2, etc. Why not multiply by decimals?

best result = x1
good result = x .50
fair result = .25

That might make the numbers square up but I don't think it'd go over, playability-wise. I think most players wouldn't like the idea that their rarest, best result isn't any more than what the dice directly show them; most D20 gamers are too attached to that double-/triple-damage effect. (I certainly would miss it.)

Mathematically you're right; it would work better if the most variability was in weapon base damage plus STR modifier (1-10 plus -5 to +5), and the "to hit" roll was measured in set multiplier levels -- x0.5, x1, x2, x3 perhaps.  The problem there is that if you assume the most common hits are the x0.5s and the x1s, a lot of the hits start to feel much the same and some of the dynamism is lost.