TL;DR: Make number of attacks player choice by moving to a "per die" rather than "per attack" notion of damage. To make the damage scale nicely, let ability score increase the minimum possible damage on a die (ability mod > 0), or reduce maximum possible damage on a die (mod < 0) instead of simply adding to the total damage.
--------------------------------------------------------------------------
Along with many others I've been pondering the issue of weapon attacks, and particularly two-weapon fighting, after seeing the most recent playtest. I'm wondering if maybe we could revisit the basic D&D abstraction of weapon attacks to build a more flexible and balanced mechanic from the start.
The mechanical core I'm considering is moving away from damage determined on a "per attack" basis to a "per die" basis. In this conception, weapon damage dice are indistinguishable from expertise dice. Additional attacks can be made a long as one has at least one damage die to contribute, but as long as the effects scale per die and not per mechanical attack one can avoid many issues surrounding multiple attacks.
For example, in this scheme two-weapon fighting is simply wielding another weapon and gaining another expertise die, i.e. its damage die. (Presumably most one-handed weapons would have a single damage die, two-handed weapons could have a large single or two smaller damage dice, and "double weapons" would have two damage dice. One could also integrate things like shield bashing into this scheme quite easily -- it might simply be a special offhand weapon.) It grants another "potential attack" in that a character could choose to spend dice that way, but the player could choose to stick to only a single mechanical attack if desired. As another example, even a low-level fighter surrounded by goblins may choose to make a few attacks without needing fancy maneuvers, which I think is a pet peeve for more than a few players. As long as the mechanical incentives are such that making few attacks is usually more effective than many attacks, most of the time we can keep game pace more consistent and mitigate the traditional high-level slowdown introduced by take-it-or-leave-it iterative attacks.
Now, a potential problem is how to re-imagine the mechanical relationship between ability score bonuses and the like to avoid the math and balance issues surrounding the iterative attacks of yesteryear. It's pretty clear to me that simply adding an ability score modifier to each damage die would only lead back to hit point inflation and huge gaps between the average damage of different characters unless the possible number of ED were tightly limited. For example, if an 18-strength character could add +4 damage to each of his 3d6 ED he would average 22.5 damage per round. The same character adding that damage per attack, i.e. in the traditional way, would average 14.5 damage per round. And the gap between a character with few ED and a modest ability score and one with many ED and a high one would grow even larger.
So, I think we would need to look for alternate mechanical expressions for the ability mod. I've considered several, but so far my favorite is letting the mod increase the minimum damage roll on a die (if positive), or reducing the maximum damage roll on a die (if negative). For example, a character with 16 Str might increase the minimum damage per die by +3. If he rolls a d6 and gets a 1-4 the damage is 4, and on a 5 or 6 it is as shown on the die. A character with 8 Str might roll a d6 and get a 6, but because of the -1 modifier to max damage the attack would only do 5 damage. That is, if that character had rolled a 5 instead, it would still be 5 damage. The ability mod therefore makes it more likely the weapon does damage closer to its theoretical maximum (or minimum if negative) and keeps the overall range of damage in the game a bit more limited. In other words, the modifiers increase the reliability of weapon damage to be either good or bad, but because they only alter rolls at the extreme ranges of a die the damage difference between characters only turns on slowly. If a strength 18, 10, and 6 character all wield the same d12 weapon, the first will always get between 5-12 damage (avg=7.33), the second 1-12 damage (avg=6.5) and the latter 1-10 (avg=6.25). Over many dice these differences can add up, but they will inhibit one character from outdamaging another by multiples while doing the same thing.
Example in play: A 4th level fighter with 17 Str is dual wielding short swords. She thus has 4d6 ED. On the first round 4 goblin underlings rush her, and the fighter decides to attack each once. She hits on the first attack and rolls a 2 for damage, but because her ability mod is +3 the minimum damage on a d6 is 4 so she actually does 4 damage. She also hits on the second and third attacks, but rolls a 5 and 6 for damage, which is unmodified by her strength. That is enough to kill those three. She misses the 4th goblin, however. On the next round the goblin boss shows up as well. She attacks the last underling and hits, rolling a 1 for damage but still doing 4 points of damage. Then she attacks the boss and hits, and her damage roll is 5,2,4 for a total of 13 damage. The next round all the other enemies are dead, so until the boss is dead she puts all her dice into a single attack on it each round. Whether she makes 1 attack or 4, however, her total expected damage remains in the same ballpark.
The major downsides I see are that it is different from the normal and by now second-nature way of calculating damage simply by adding. It may be a little slower, although I think that would disappear with practice. In addition, for extreme ability scores the range of possible damage is so low that rolling may feel like a waste of time. Whether the much lower variance in weapon damage (given a particular character) is a feature or a bug probably depends on the player.
A lesser downside, in my opinion, is that high ability mods eventually surpass the size of the die, and that seems not quite elegant. For example, if a giant has 22 str for a +6 modifier, does a d4 damage die cause 4 damage (capped) or 7 damage (because minimum damage is raised by 6). I'd lean toward the latter, but I don't know. Either way, if ability scores for characters max out at 20 this will only happen for d4s and d6s. Only a mod of +11 (32) would saturate a d12, and that is clearly up in "god of strength" territory if the giants article is any indication.
In order to keep play moving swiftly it is probably best if multiple attacks are not always the best option. For example, if Deadly Strike allows one to add ED as damage after finding out an attack is successful, then the most efficient use of ED even against a single opponent would be to make attacks with a single ED one at a time until one hits. Even a decreasing attack bonus per attack wouldn't change that tactic. Instead, the player must have an incentive to spend ED on attacks before the outcome is known. Perhaps adding damage after-the-fact does not include the ability mod, for example. Likewise, many maneuvers should probably work better if the dice are spent or risked up front. Whether or not additional attacks should be penalized in terms of attack bonus I'm not sure, but at least a single rule would suffice instead of the rainbow of penalties for additional attacks in 3/3.5.
Other areas that would need to be considered are how to handle magic items, especially since special weapons have generally had per hit effects that may not be appropriate with this design.
Overall, I think this conception embraces the abstraction of attacks in D&D more fully than past systems and could reduce cases where the abstraction leaks. It has always been the case that a mechanical attack and narrative attack do not necessarily correspond, but non-correspondence can cause conflicts in scenarios where the player often would like them to correspond, as with two-weapon fighting granting an extra attack. Likewise, when iterative attacks were essentially required to get full effect in a round some narrative elements (e.g. a single power lunge) were more difficult to buy. Moving to a per die system from a per attack system embraces the existing abstraction by letting the player or DM choose their own correspondence more freely, and from round-to-round, with far fewer mechanical side effects.
Let me know what you think.
--------------------------------------------------------------------------
Along with many others I've been pondering the issue of weapon attacks, and particularly two-weapon fighting, after seeing the most recent playtest. I'm wondering if maybe we could revisit the basic D&D abstraction of weapon attacks to build a more flexible and balanced mechanic from the start.
The mechanical core I'm considering is moving away from damage determined on a "per attack" basis to a "per die" basis. In this conception, weapon damage dice are indistinguishable from expertise dice. Additional attacks can be made a long as one has at least one damage die to contribute, but as long as the effects scale per die and not per mechanical attack one can avoid many issues surrounding multiple attacks.
For example, in this scheme two-weapon fighting is simply wielding another weapon and gaining another expertise die, i.e. its damage die. (Presumably most one-handed weapons would have a single damage die, two-handed weapons could have a large single or two smaller damage dice, and "double weapons" would have two damage dice. One could also integrate things like shield bashing into this scheme quite easily -- it might simply be a special offhand weapon.) It grants another "potential attack" in that a character could choose to spend dice that way, but the player could choose to stick to only a single mechanical attack if desired. As another example, even a low-level fighter surrounded by goblins may choose to make a few attacks without needing fancy maneuvers, which I think is a pet peeve for more than a few players. As long as the mechanical incentives are such that making few attacks is usually more effective than many attacks, most of the time we can keep game pace more consistent and mitigate the traditional high-level slowdown introduced by take-it-or-leave-it iterative attacks.
Now, a potential problem is how to re-imagine the mechanical relationship between ability score bonuses and the like to avoid the math and balance issues surrounding the iterative attacks of yesteryear. It's pretty clear to me that simply adding an ability score modifier to each damage die would only lead back to hit point inflation and huge gaps between the average damage of different characters unless the possible number of ED were tightly limited. For example, if an 18-strength character could add +4 damage to each of his 3d6 ED he would average 22.5 damage per round. The same character adding that damage per attack, i.e. in the traditional way, would average 14.5 damage per round. And the gap between a character with few ED and a modest ability score and one with many ED and a high one would grow even larger.
So, I think we would need to look for alternate mechanical expressions for the ability mod. I've considered several, but so far my favorite is letting the mod increase the minimum damage roll on a die (if positive), or reducing the maximum damage roll on a die (if negative). For example, a character with 16 Str might increase the minimum damage per die by +3. If he rolls a d6 and gets a 1-4 the damage is 4, and on a 5 or 6 it is as shown on the die. A character with 8 Str might roll a d6 and get a 6, but because of the -1 modifier to max damage the attack would only do 5 damage. That is, if that character had rolled a 5 instead, it would still be 5 damage. The ability mod therefore makes it more likely the weapon does damage closer to its theoretical maximum (or minimum if negative) and keeps the overall range of damage in the game a bit more limited. In other words, the modifiers increase the reliability of weapon damage to be either good or bad, but because they only alter rolls at the extreme ranges of a die the damage difference between characters only turns on slowly. If a strength 18, 10, and 6 character all wield the same d12 weapon, the first will always get between 5-12 damage (avg=7.33), the second 1-12 damage (avg=6.5) and the latter 1-10 (avg=6.25). Over many dice these differences can add up, but they will inhibit one character from outdamaging another by multiples while doing the same thing.
Example in play: A 4th level fighter with 17 Str is dual wielding short swords. She thus has 4d6 ED. On the first round 4 goblin underlings rush her, and the fighter decides to attack each once. She hits on the first attack and rolls a 2 for damage, but because her ability mod is +3 the minimum damage on a d6 is 4 so she actually does 4 damage. She also hits on the second and third attacks, but rolls a 5 and 6 for damage, which is unmodified by her strength. That is enough to kill those three. She misses the 4th goblin, however. On the next round the goblin boss shows up as well. She attacks the last underling and hits, rolling a 1 for damage but still doing 4 points of damage. Then she attacks the boss and hits, and her damage roll is 5,2,4 for a total of 13 damage. The next round all the other enemies are dead, so until the boss is dead she puts all her dice into a single attack on it each round. Whether she makes 1 attack or 4, however, her total expected damage remains in the same ballpark.
The major downsides I see are that it is different from the normal and by now second-nature way of calculating damage simply by adding. It may be a little slower, although I think that would disappear with practice. In addition, for extreme ability scores the range of possible damage is so low that rolling may feel like a waste of time. Whether the much lower variance in weapon damage (given a particular character) is a feature or a bug probably depends on the player.
A lesser downside, in my opinion, is that high ability mods eventually surpass the size of the die, and that seems not quite elegant. For example, if a giant has 22 str for a +6 modifier, does a d4 damage die cause 4 damage (capped) or 7 damage (because minimum damage is raised by 6). I'd lean toward the latter, but I don't know. Either way, if ability scores for characters max out at 20 this will only happen for d4s and d6s. Only a mod of +11 (32) would saturate a d12, and that is clearly up in "god of strength" territory if the giants article is any indication.
In order to keep play moving swiftly it is probably best if multiple attacks are not always the best option. For example, if Deadly Strike allows one to add ED as damage after finding out an attack is successful, then the most efficient use of ED even against a single opponent would be to make attacks with a single ED one at a time until one hits. Even a decreasing attack bonus per attack wouldn't change that tactic. Instead, the player must have an incentive to spend ED on attacks before the outcome is known. Perhaps adding damage after-the-fact does not include the ability mod, for example. Likewise, many maneuvers should probably work better if the dice are spent or risked up front. Whether or not additional attacks should be penalized in terms of attack bonus I'm not sure, but at least a single rule would suffice instead of the rainbow of penalties for additional attacks in 3/3.5.
Other areas that would need to be considered are how to handle magic items, especially since special weapons have generally had per hit effects that may not be appropriate with this design.
Overall, I think this conception embraces the abstraction of attacks in D&D more fully than past systems and could reduce cases where the abstraction leaks. It has always been the case that a mechanical attack and narrative attack do not necessarily correspond, but non-correspondence can cause conflicts in scenarios where the player often would like them to correspond, as with two-weapon fighting granting an extra attack. Likewise, when iterative attacks were essentially required to get full effect in a round some narrative elements (e.g. a single power lunge) were more difficult to buy. Moving to a per die system from a per attack system embraces the existing abstraction by letting the player or DM choose their own correspondence more freely, and from round-to-round, with far fewer mechanical side effects.
Let me know what you think.
Last edited: